100% found this document useful (2 votes)

12K views324 pages

Introduction To Environmental Economics

Q: How might non-uniform mixing of pollutants make a single pollution tax rate inefficient, and what solutions have been proposed to address this?

A single pollution tax rate is inefficient with non-uniformly mixed pollutants because it ignores the varying environmental impacts of emissions from different sources. A uniform tax does not address the differential damage caused by emissions in different locations or contexts. To manage this, proposals include creating 'banded tax rates' that adjust taxes based on a firm's location or impact severity, or considering unique rates for different polluters to better align economic incentives with environmental outcomes .

Q: What challenges arise from using expected utility theory in risk-benefit analysis for environmental risks, and how do heuristics affect this analysis?

Expected utility theory assumes well-defined preferences and logical risk evaluations, but in practice, this model is 'too thin' as it doesn't account for the systematic deviations observed in human behavior. People often rely on heuristics, which are simplified reasoning rules that can lead them to overvalue or undervalue risks based on context rather than objective probabilities. These heuristics introduce variability into the stated value for risk reduction, complicating the accuracy of risk-benefit analyses in environmental policy-making .

Q: How does the value of statistical life (VSL) reflect the private ability to reduce risk, and what issues may arise when applying a fixed VSL across different environmental contexts?

The value of statistical life (VSL) reflects the private ability to reduce risk as it is partly dependent on an individual's access to private risk reduction markets and their willingness to pay for risk reductions. People's ability to privately reduce risk through markets for self-protection, like purchasing smoke detectors or air filters, affects their baseline risk and thus influences their willingness to pay for additional risk reduction . Applying a fixed VSL across different environmental contexts may lead to inefficiencies because the ability to reduce risks privately varies between different risks and individuals. For example, markets for reducing water contamination risk may be different from those for reducing toxic air risk, leading to variations in perceived value of the same VSL in different contexts . Additionally, this can result in biased upward estimates of VSL because it fails to account for the role of private risk reduction actions . Inappropriately applying a fixed VSL can lead to misjudgments in policy-making, potentially resulting in inefficient levels of environmental protection and resource allocation .

Q: What implications does biodiversity loss have on ecosystem services and societal costs?

Biodiversity loss negatively impacts ecosystem services by disrupting the species that provide critical functions such as nutrient cycling and water regulation. This loss translates to societal costs because species offer direct value as genetic resources for agriculture and potential pharmaceutical compounds. Additionally, the disappearance of biodiversity carries aesthetic values and non-use values, which society may place on the preservation of ecosystems and 'charismatic' species .

Q: What are the economic challenges of applying uniform abatement standards across industries, and how does this relate to marginal abatement costs?

Uniform abatement standards across industries do not consider variations in marginal abatement costs (MAC) among firms. Implementing a uniform policy can lead to inefficient resource allocation as firms with lower MACs over-invest in abatement while firms with higher MACs under-invest. A more cost-effective approach allows firms to meet targets by trading permits or implementing different standards based on specific cost conditions, optimizing overall economic efficiency and achieving environmental goals at reduced costs .

Q: Discuss the economic rationale behind tradable permits and how they can lead to increased industry-level profits as shown in the Swedish paper and pulp industry case.

Tradable permits provide an economic framework where emissions are capped and permits are allocated to industries, allowing firms to trade these permits if they can reduce emissions below their allotment. This creates an incentive for companies to invest in cleaner technologies if doing so is cheaper than buying permits . In the Swedish paper and pulp industry, tradable permits led to increased profits because firms that could reduce emissions at a low cost could sell their surplus permits to those facing higher abatement costs, thus minimizing overall industry compliance costs . This setup encourages efficient resource allocation, reducing overall compliance costs compared to traditional regulation, which in turn can lead to increased industry-level profits .

Q: Evaluate the implications of setting a zero-risk target in environmental regulation. Why might this approach be problematic?

Setting a zero-risk target in environmental regulation can be problematic due to scientific and economic constraints. As scientific methods become capable of detecting minute chemical traces, zero-risk goals become increasingly restrictive, as virtually all substances could present some level of carcinogenic risk. The economic costs of achieving zero-risk are excessively high, often leading to prohibitive measures. Thus, while appealing in theory, a zero-risk strategy is often impractical and inefficient in application .

Q: What are the potential cost-saving benefits of tradable pollution permit (TPP) systems compared to traditional regulation according to the case study of estuary DO levels?

Tradable pollution permit (TPP) systems offer significant cost-saving benefits over traditional regulation by allowing for flexibility in meeting emissions targets. In the case study of estuary DO levels, a TPP system could achieve a 20% improvement in dissolved oxygen at one-ninth the cost of uniform regulation. This substantial saving is due to the variability in marginal abatement costs among firms, which TPP systems exploit by enabling firms with lower abatement costs to sell permits to those with higher costs, thereby minimizing overall costs . Additionally, TPP systems guarantee that total emissions stay below a predetermined cap, offering certainty over environmental outcomes compared to taxes, which may result in emissions exceeding targets if taxes are insufficiently stringent . However, savings may be reduced by uncertainties in water-quality impacts and transaction costs, which are associated with permit trading . Despite these challenges, TPP systems can still achieve target emissions reductions at a lower cost than standards, reflecting their potential for economic efficiency ."}

Q: In what way does the species-area curve inform conservation strategies, and why is prioritizing rare habitats significant for biodiversity conservation?

The species-area curve demonstrates that larger habitats typically sustain more species, guiding conservation towards protecting extensive habitats to maximize species preservation. Prioritizing rare habitats is crucial since the loss of a small area in these can lead to disproportionate species loss. Conservation strategies focusing on maintaining diverse and connected habitats are significant for sustaining ecosystem services and preventing extensive biodiversity loss, as they help maintain species' genetic diversity and ecosystem resilience .

Q: How does allowing permits to trade at a one-for-one rate affect water quality standards in the context of non-uniformly mixed pollutants?

Allowing permits to trade at a one-for-one rate in the context of non-uniformly mixed pollutants can result in local violations of water quality standards. If two firms trade permits without accounting for their different pollution impacts, the downstream environment can suffer increased damage. For instance, if Firm A, located upstream with more harmful emissions per unit, buys permits from Firm B, whose emissions are less harmful, the aggregate emissions remain unchanged, but local water quality downstream of A deteriorates. This situation illustrates the need for trading restrictions or adjusted rates to prevent localized harm in water pollution trading systems .

This document provides an introduction to and overview of the second edition of the textbook "Introduction to Environmental Economics" by Nick Hanley, Jason F. Shogren, and Ben White. The book covers key economic concepts and tools for analyzing environmental issues, including markets and market failure, methods for valuing the environment, environmental risk and behavior, economic growth and sustainability, strategic interactions related to the environment, and the relationship between trade and the environment. The document lists detailed chapter contents and provides brief biographies of the authors dedicating the book to their families.

Uploaded by

Aditi Sarkar

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

100% found this document useful (2 votes)

12K views324 pages

Introduction To Environmental Economics

Uploaded by

Aditi Sarkar

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

Nick Hanley | Jason Shogren | Ben White

| lsecondedition) = —’ | ge
Digitized by the Internet Archive
in 2022 with funding from
Kahle/Austin Foundation

[Link]
boo Introduction to Environmental Economics
Introduction to
Environmental
Economics

Professor Nick Hanley

ProfessorofEconomics, University of Stirling

Professor Jason F. Shogren

Stroock Professor of Natural Resource Conservation and
Management, University of Wyoming

Dr Ben White
Professor in the School of Agricultural and Resource Economics,
University of Western Australia

OXFORD
UNIVERSITY PRESS
OXFORD
UNIVERSITY PRESS

Great Clarendon Street, Oxford, OX2 6DP,

United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Nick Hanley, Jason F. Shogren, and Ben White 2013
The moral rights of the authors have been asserted
First Edition copyright 2001
Impression: 5
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope ofthe
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Data available
ISBN 978-0-19-956873-4
Printed in Great Britain by
Bell & Bain Ltd, Glasgow
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
N. Hanley
For Fiona, Rose, Charlie, and Finn

J. Shogren
For Maija, Riley, and Mikko
B. White
For Jane, Catherine, Steven, and Dominic
Acknowledgements

The authors thank staff at OUP for all their hard work and encouragement. Nick thanks Mik
Czajkowski and Dervla Brennan for help on several chapters, and Jack Pezzey for many discussions.
He also thanks the Crawford School, Australian National University, for their hospitality whilst
writing the final drafts. Ben would like to thank Michael Burton for his insights into environmental
economics and Deborah Swindells for administrative support. Jason thanks his teachers and students
over the years for their willingness to share ideas, agreeably and otherwise.
New to this Edition

Building on the success of the first edition, the second edition of an Introduction to Environmental
Economics features the following updated material:

e Substantial revision of all existing chapters

e Greater focus on environmental policy and management links to policy issues which have
recently emerged
e Updating of box sections with recent findings from the literature
® New pedagogical features including tutorial questions and definitions.

PRAISE FOR THE BOOK

‘Tt bases heavily on empirical evidence and also provides an extensive number of examples, which explain
well definitions mentioned in the text.’

Leon Vinokur, Queen Mary, University of London

‘The writing style is of a high quality and is pitched at the appropriate level for an introduction to
Environmental Economics. Economic concepts are introduced using easy-to-understand language and
examples.’

Professor Mike Christie, Aberystwyth University

Detailed Contents

List of Boxes xiii

List of Figures XV
List of Tables xvii

PART |: Economic Tools for the Environment

SSHSHHSHHHSHEHTHESHESEHESEHEHHSSHEHEEHHOHHHHEHEEESSHESHEHTOSOHSOHHATEHHOHSOSESEHOSHHEESGEOEH ENS eseoeeee

1 Introduction: Economics for the Environment 3

1.1 The Economy and the Environment 4
1.2 Key Insights from Economics of which Environmental Scientists,
Environmental Managers, and Politicians Should be Aware 6
1.3 The Rest of this Book: An Overview 8
1.4 Using this Book for Teaching and Learning 8
2 Markets and the Environment 11
2.1 The Power of Markets 12
2.2 Market Failure 15
2.3 Markets for the Environment 23
3 Valuing the Environment: Concepts 39
3.1 What Does Economic Value Mean? 39
3.2 In What Sense Does the Environment Have Economic Value? 44
3.3. Why Place Economic Values on the Environment? 49
3.3.1 Cost-benefit analysis 49
3.3.2 Other uses of environmental valuation 55

4 Valuing the Environment: Methods 59

4.1 An Overview of the Methods 59
42 Stated-preference Approaches 60
42.1 Contingent valuation 60

42.2 The choice experiment method 63

43 Revealed-preference Approaches 67
43.1 The hedonic pricing method 67
43.2 Travel-cost models 70

44 Production-function Approaches 76
45 Benefits Transfer 78
DETAILED CONTENTS

5 Environmental Risk and Behaviour

oe Rational Benaviour under Risk
a2 Behavioural Economics and Risk
De Valuing Risks to Life and Limb
5.4 Regulating Risk
6 Economic Growth, the Environment, and Sustainable Development
6.1 Economic Growth and Development
6.1.1 Why do economies grow?
6.1.2 Growth versus development
6.2 Predictions from the Past
6.3 Growth and the Environment: The Environmental Kuznets Curve
6.4 Broadening
the Issue: The Economics ofSustainable Development
6.4.1 Definitions of sustainable development
6.4.2 Sustainability rules?
Go Measuring ‘Sustainability’
6.5.1 Green net national product
6.5.2 Genuine savings
6.5.3 Other sustainability indicators

7 Strategic Interactions and the Environment

pe Introduction
ea Game Theory
7.2.1 Basic concepts
7.2.2 The prisoner's dilemma
7.2.3 Common property—the fisher's dilemma
73 Self-governance—Escaping the Tragedy of the Commons
7A Repeated Fishing Games
Tos: Co-operative Games
7.6 Game Theory and Transboundary Pollution Control
7.6.) Introduction
7.6.2 The acid rain game

8 Trade and the Environment

8.1 Introduction
8.2 Why do Countries Gain from Trade?
8.2.1 Extending the basic trade model to include the environment
8.2.2 Trade and the environment
Empirical Evidence on the Trade Effects of Environmental Regulation
8.4 International Trade Agreements and the Environment
8.4.1 International trade agreements
8.4.2 Multilateral environmental agreements and trade
DETAILED CONTENTS xl

PART II: Applying the Tools

CSP eee eeeeesoseessoreseees Peeeeseeeeeeeesoeseses e@eoeeecee SCeoeeseosesereoseseesessse eeesece eee

9 The Economics of Climate Change

oa Introduction
9.2 A Global Environmental Risk
93 The Challenge of International Coordination
9.4 The Benefits and Costs of International Co-operation
She Economic Issues Underlying Benefit and Cost Estimates
9.6 A Flexibility-Stringency Trade-off
10 Forests
10.1 Introduction
10.1.1 Local benefits
10.1.2 National benefits
10.1.3 Global benefits
10.2 Forest Distribution and Losses
10 Why is Rainforest Lost? Economic Theories of Deforestation
10.4 Tropical Deforestation and Poverty
10.5 Forestry Management
10.6 How Much Natural Forest Should be Preserved?
10.6.1 The optimal area of natural forest for a region 205
hon, Policies for Rainforest Conservation 206
10.7.1 International policy 206
10.7.2 The Global Environment Facility (GEF), REDD, and REDD+ 207
10.7.3 How effective are international policies? 208
0.7.4 National and local policies 209

11 The Economics of Water Pollution IN

Introduction AV
11.1.1 Trends over time 213
The Costs of Water Pollution Control 214
Non-point Source Water Pollution: A Difficult Problem to Solve 217
Measuring Water-quality Benefits 223
Problems for Cost-benefit Analysis of Water-quality Improvements 227
11.5.1 Identifying the water-quality change to be valued 228
11.5.2 Problems in pricing the environmental change, once identified 228

12 Biodiversity 3)
21 Introduction JENS
a2 What to Conserve? 236
12.2.1 Economic insights into conservation objectives 236
12.2.2 Allocating resources between species and ecosystems 238
xii DETAILED CONTENTS

12.3 Values of Biodiversity Ao)

2.3.1 Direct values 240
2.3.2 Bioprospecting
for drugs 241
12.3.3 Ecosystem service values 243
2.3.4 Aesthetic use and non-use values 249
12.4 Biodiversity and Conservation Policy
24.1 |nternational policy
12.4.2 National policies are more challenging

13. Non-renewable Natural Resources and Energy

13) Natural Resource Types
eZ The Extraction of aNon-renewable Resource
3.2.1 The baseline case—an introduction to Hotelling’s rule
13.2.2 Resource extraction and the monopolist
3.2.3 Model including extraction costs
iiss Measuring Resource Scarcity
3.3.1 Resource lifetime

13.3.2 Unit cost measures

3.3.3 Real prices

13.3.4 Economic rent

13.3.5 Hotelling and the oil market

13.4 Global Energy Demand and Supply
135 Global Issues in Energy Policy

Glossary
Index
List of Boxes

2.1 Market Equilibrium and the Gains from Trade

22 An Example of an Externality: Air Pollution in Ecuador
2.3 The Tragedy of the Commons
2.4 Government Intervention Failure: The Common Agricultural Policy
2:5 Trading Pollution Permits: What's the Evidence?
SL What Role Should CBA Play in the Policy Process?
3.2 An Illustrative CBA
3.3 Discounting and the Discount Rate
3.4 An Example of Cost-benefit Analysis of Hydropower Regulation
4.1 Valuing Heritage Sites: Stonehenge
4.2 Hypothetical Bias in Choice Experiments
4.3 Hedonic Price Measures of Noise Externalities
4.4 Using the Travel-cost Model to Value Outdoor Recreation
5.1 The Allais Paradox
5.2 Hyperbolic Discounting
5.3 Valuing Reduced Risk from Food-borne Pathogens in the Laboratory
6.1 ls GNP a Good Measure of Well-being?
6.2 Testing the Environmental Kuznets Curve 115
6.3 Social Capital and the EKC
6.4 The Costs of Not Following the Hartwick Rule
6.5 Green Net National Product and Genuine Savings for Scotland
7.1 Cod Wars
Ihe Derivation of the Nash Equilibrium for a Fishery
7.3 An Example of Self-organization ina Common-property Resource
7.4 Revision on Discounting
ie Self-enforcing International Environmental Agreements
7.6 The Montreal Protocol
8.1 The California Effect
8.2 The Porter Hypothesis
8.3 Testing the Pollution Haven Hypothesis
8.4 European Environmental Regulation and Trade
9.1 The Kyoto Protocol on Climate Change
9.2 The EU Emissions Trading Scheme
9.3 Stated Preference Evidence from the USA
10.1 Forest Ecosystem Services to Shifting Cultivation in Brazil
10.2 Deforestation in Rondonia: An Example ofthe Frontier Model
10.3 Deforestation in the Ivory Coast and the Discount Rate: An Example of Immiserization
10.4 Do Protected Forests in the Neighbourhood Make People Poorer?
xiv LIST OF BOXES

10.5 The Example of the GEF Project for Tropical Rainforest Conservation in Brazil 207
10.6 The Noel Kempff Mercado National Park, Bolivia: A Case Study of a Carbon Offset Project 208
11.1 Changes in River Quality over Time in Scotland 214
11.2 Do Tradable Permits Save Money in Controlling Water Pollution? EMles
11.3 Water-quality Trading in Practice aay
11.4 National Water-quality Benefits Assessment in the USA 224
11.5 Comparing Contingent Valuation and Choice Experimeht Values for
Water-quality Improvements 226
11.6 Who Faces the Costs and Who Gets the Benefits of Water-quality Improvements? 230
12.1 ow Many Species Are There and How Many Have Been Lost? 234
12.2 oah’s Library 236
12.3 ow Hot Are Biodiversity Hotspots? 240
12.4 Bioprospecting in Costa Rica 241
12.5 angrove Fisheries Linkages in the Campeche of Mexico 245
12.6 The Value of Biodiversity in Poland 250
12.7 Valuable Viper Venom 2353
12.8 Saving Elephants 254
13.1 Testing Hotelling’s Rule 267
13.2 Why Is Resource Scarcity so Difficult to Measure? 293
13.3 Energy Facts Ns)
13.4 Renewable Energy in the UK: Promotion and Impacts 281
List of Figures

ils. Economy-environment interactions

2.1. Marginal benefits and marginal costs: private and social
2.2(a) A MAC for the Jones Company, for reducing emissions by installing end-of-pipe treatment
2.2(b) The MAC curve for the Jones Company defined across all emission reduction options
2.3 From firm's emissions to aggregate emissions
2.4 Pollution taxes as the least-cost solution
2.5 The impact of awater pollution tax
2.6 Tradable permits as the least-cost solution
2.7 Prices in the permit market
2.8 Permit trading in an estuary
3.1(a) WTP forwildlife protection
3.1(b) Marginal WTP for wildlife protection
3.1(c) Marginal WTP for two people
3.2 Indifference curves and the value of an increase in environmental quality
33 Indirect environmental values
3.4 The economic value of a wetland
4.1 The value in the housing market of an improvement in air quality
4.2(a) The relationship between visits and travel costs
4.2(b) A derivation of consumers’ surplus from the visit-cost relationship
4.3 Predicted changes in agricultural outputs and market values for a 1 degree rise in mean
temperatures in the UK
5.1. = Diminishing marginal returns
5.2 Objective and subjective risk perception
6.1. =The environmental Kuznets curve We
6.2. Critical thresholds and planetary boundaries 128
6.3 How the World Bank calculates genuine savings 126
7.1 Fishery revenue and cost curves 138
7.2. Fishery Nash equilibrium 138
7.3 The co-operative solution: individual rationality 44
7.4 The co-operative solution: individual and group rationality 145
7.5. Transboundary pollution: the acid rain game 149
8.1 World gross domestic product (GDP), trade volume, and value 52
8.2 Trade and comparative advantage 185
8.3 The effects of domestic environmental policy on trade 157
8.4 = The production possibility frontier and national income 158
8.5 Scale, composition, and technique effects (sg
8.6 Optimal pollution 160
8.7. Possible causal links between trade and the environment 161
9.1. Voluntary contributions to abatement 181
9.2. The flexibility-stringency trade-off eh
XvI IST OF FIGURES

The deforestation environmental Kuznets curve 201

The optimal timber rotation 204
The equilibrium natural forest area 206
The effect of a pollution tax (t) on emissions from two firms, A and B 216
Permit trading in an estuary 2\i
Using input taxes to reduce non-point pollution ale
\
Explaining varying levels of an attribute
Examples of distance-decay functions for English water-bodies
The species-area curve
The trade-off between cost and probability of survival
Ecosystem cycles
Marine ecosystem services in Alaska
The Hotelling model of resource extraction
Resource extraction, technological innovation, and exploration through time
The mineralogical threshold
Dubai oil prices, 1972-2010
List of Tables

HL Some of the ecosystem service flows from moorlands and heaths

3.2 Values provided by tropical coastal and marine ecosystems
4.1 Choice experiment for valuing Australian wetlands
4.2 Costs of mangrove losses in Thailand, 1996-2004
4.3 Valuing ecosystem service flows in moorlands
6.1 The ranking of global economies by GDP in purchasing power parity (PPP) terms, 2008
6.2 Per capita gross national income (GNI) for ten highest-ranked and ten lowest-ranked
countries, 2008
6.3 Economic growth rates for a selection of countries
6.4 Human Development Index scores, 2007 data
6.5 Shares of capital in total wealth, 2000
6.6 Genuine savings estimates for a range of countries, 2000 (all figures are as percentages
of gross national income)
7.1 The prisoner's dilemma, strategic form
1? The fisher’s dilemma
Tee The groundwater co-operation game
7.4 The acid rain game
7.5 Transboundary pollution—the acid rain game
8.1 Trade and absolute advantage
8.2 Trade and comparative advantage
8.3 Elasticities of SO, emissions
9.1 Total carbon dioxide emissions of Annex 1 parties in 1990, for the purposes of
Article 25 of the Kyoto Protocol
OZ Kyoto Protocol Annex 2. Quantified emission limitation or reduction commitment
(percentage of base year or period) 180
10.1 Regional forest area and change 198
11.1 Sources at risk of not achieving Good Ecological Status for different water-body types,
Scotland, 2005 25
11.2 Ranking of options to achieve a water-quality target according to cost 221
11.3 An example of a choice card 224
11.4 Example outputs from a choice experiment for two rivers in Scotland 205
11.5 Economic benefits to anglers in England from water-quality improvements 227.
13.1 Abundance, regional concentration, and recycling for metallic minerals 270
13.2 World primary energy production (quadrillion Btu) Zag
13.3 World primary energy consumption (quadrillion Btu) 277
13.4 World production of primary energy by energy type (quadrillion Btu), 1980-2006 278
13.5 World estimated recoverable coal, 2008 278
13.6 World crude oil and natural gas reserves (March 2009) 280
‘ . — a ee aieny » 96g% ere ee
ie an itn eos iin) sane “ies a
- - . pom sp Visits!)
ov | Sik! br-O8 tomes: s:
a
"Game + ; « _ ;

i] oan

a> ot ¥ an a eerie i Ry a
we wie Ot deg i:
4
- —=@ eon is

imps! TE
ae = eer
=—9-—-. tw af
smi Meu = 4 =o at
_ =e =e.
@ Fé: =
ee ‘saw tf
1% : f ~—> 2 & Ld
ann ake ae
ig) cam . Os dag
nee) eee ete.)
set (a ~~ a |e
- e6 +52 @¢or 5 LP pl f:
ae A
- =a ae: eee ti)
> <t_ ° aaa’ ats
= 9? £861 ae or?
es A -_ ee
. ' D _ = @ @b=*s ee
~ ae OS Ep manglerqene t+
rl ie 0 Gerard ive@ ai
. SS FO Mb Shure’ acyl! ‘ar
=“ 6 Siptommareent “a
a ho o==p
8 ae
Part |

Economic Tools for

the Environment
Introduction: Economics
for the Environment

Welcome! The title of this chapter is ‘Economics for the Environment’. Why? Because we
believe that economics has an important contribution to make in helping us understand and
solve the many environmental problems facing people throughout the world today. People
often equate ‘economics’ with ‘financial and commercial’, yet economics is as much about
Main Street as it is about Wall Street. Economic arguments can often be used to help protect the
environment, rather than harm it. In Part I of this book, we set out the principal insights that
economics has to offer. Then, in Part II, we take a range of important environmental problems,
and show how these insights can improve how people respond to these problems. We believe
strongly that an environmental policy that does not address the economic behaviour of con-
sumers and firms is likely to get things badly wrong. Often, economic actors can be induced to
take account of their impacts on the environment, by getting prices on environmental goods
and services right. Economic and environmental systems are closely interlinked: by not address-
ing economic insights, we are unlikely to produce cost-effective outcomes for either system.
People are paying more attention to the environmental consequences of economic activ-
ity, and to the economic value of the environment. Partly, this is due to greater public aware-
ness of environmental issues such as climate change or the loss of treasured local landscapes.
Partly, it is a consequence of the increasing interest of policy-makers to understand the
benefits and costs of environmental regulation. People have started to care and know more
about environmental degradation and the benefits of ecosystem services. Sustainable devel-
opment and green growth have become key concepts within many public policy pronounce-
ments, yet the usefulness of these concepts is unclear—whilst predictions of impending
doom due to world population increases and climate change continue to circulate. Given
our belief that economics has an important, indeed vital, contribution to make in under-
standing these issues and the trade-offs that lie behind them, we think that a book that tries
to get the basic ideas across to a wide audience seems a good idea.
This is the second edition ofabook first published in 2001. Since then, public debates over
climate change (the evidence for it, and what to do about it) have greatly increased, whilst
there has been a growing awareness since the Millennium Ecosystem Assessment of the
benefits that ecosystems provide to people. Many countries have greatly expanded their
renewable energy capacities, whilst global targets on reducing biodiversity losses have not
ECONOMIC TOOLS FOR THE ENVIRONMENT

been achieved. The links between the increasing scale of global economic activity, trade, and
environmental quality continue to be debated. We thus felt that there was still a need for an
introductory-level textbook that explains the contributions that economics can make to
understanding and resolving these problems. In revising the text, we have taken account of
both the evolution of environmental issues and the advances in environmental economics
since the first edition appeared. We have also tried to take into account feedback from peo-
ple who have used our textbook, both students and faculty.
In the rest of this chapter, we:
e Discuss the connections between the economy and the environment.
e Review ten key insights from environmental and resource economics that
environmental scientists, managers, and politicians ought to be aware of.
e Explain how this book is best used.
e Provide an overview of what happens in the rest of the book.

1.1 The Economy and the Environment

This book explores why economics matters more to environmental and natural resource
policy than many people think. To begin with, it is important to say that economics is not
just about financial comings and goings within markets: the ‘unpriced’ or non-market ser-
vices that the natural environment provides us with are equally its concern. The value of
protecting wetlands for their biodiversity, flood defence, and pollution treatment functions
is just as much an economic value as this week’s production of oil from a Texas oilfield.
Next, people can benefit from a better appreciation of how the economy and the environ-
ment are interlinked. The economy operates from inside the environmental system, with
conditions in the two systems being simultaneously determined in an evolving, dynamic
way. By ‘the economy’, we mean all the firms that make up industry; households (people) in
their twin roles as consumers and suppliers of labour; governments; the institutions that
govern interactions between these groups, such as markets; the state of technology; and our
stocks of produced capital (such as roads and space stations). By ‘the environment’, we
mean all natural resources found in the biosphere, including land, land cover, and ecosys-
tems (flora and fauna); resource deposits under the land surface; the world’s oceans and
atmosphere; and the natural climate and nutrient cycles. As Figure 1.1 shows, there are
many links between these two systems.
First, the environment provides the economic system with inputs of raw material and
energy resources, including minerals, metals, food, hydrocarbons, and fibres such as wood
and cotton. These resources may be either non-renewable—such as coal or iron ore—or
renewable, such as fisheries or forests. Inputs are transformed by the economic system into
outputs that consumers demand—wood into paper and oil into petrol, for example.
Second, the economy uses the environment as a waste sink. Wastes may originate from
either production processes, such as CO, from electricity generation; or from consumption
activities, for instance, when households put out the garbage for collection and disposal.
Wastes come in a number ofbasic types—solid, air-, or water borne—and the environment
has a limited assimilative capacity to absorb and transform some wastes into harmless
INTRODUCTION: ECONOMICS FOR THE ENVIRONMENT

Environmental system

Global life-support Impacts on

biodiversity

Economic system

Resource H Amenity values

inputs
Wastes (global,
regional, and
local pollution)

Figure 1.1 Economy-environment interactions.

substances. Pollution is said to occur when emissions exceed assimilative capacity, and pro-
duce some undesirable impact.
Third, the environment provides households with a direct source of amenity. People
derive utility (happiness, satisfaction) from the contemplation of scenic beauty and wildlife,
and from hiking and fishing. As Chapter 3 makes clear, these direct utility impacts are both
important and relevant from an economics viewpoint.
Finally, the environment provides the economic system with basic life-support services.
Since the Millennium Ecosystem Assessment, these services have increasingly been referred
to as ecosystem services. These include climate regulation, the operation of the water cycle,
the regulation of atmospheric composition, and nutrient cycling.
One obvious point is that if the economy increases its demand on the environment with
regard to any one of these four service flows, this can impact on the environment’s ability to
provide other services. For example:

e Anincrease in the use of the environment as a waste sink due to increased emissions of
pollutants may reduce the environment’s ability to supply basic life-support by
interfering with climate regulation; or may reduce the amenity value of the
environment by degrading wildlife populations.
e Anincrease in demand on the environment for resource inputs may mean a reduction
in amenity flows; for example, if quarries are developed in national parks or as logging
reduces the area of rainforest.
e Wealso showa link in Figure 1.1 between the economy and biodiversity. This shows that
economic activity can affect natural diversity, most notably by taking over habitats (e.g.
when a rainforest is turned into a gold mine). Diversity is thought to be an important
property of natural systems, especially with regard to their ability to withstand shocks such
as drought and fire (this property is sometimes known as resilience). Thus, ultimately,
reductions in biodiversity can have direct and indirect effects on human well-being.

An important feature of the interdependent economic-environmental system is co-

evolution. This means that the way in which the economic subsystem evolves over time
depends on the changing conditions of the environmental subsystem, and vice versa. A
good illustration of this, in terms of impacts of the changing environment on the evolution
of the economy, concerns climate change in Neolithic Scotland. In around 4000 Bc,
Neolithic farmers settled many parts of the west coast of Scotland, including islands such as
Arran. The prosperity of this society may be seen from the large stone monuments they
erected. Substantial land-cover alterations occurred, notably the onset of clearing land of
trees.| However, as the climate became cooler, crop’ growth rates declined and the land
became harder to subsist in. Areas of peat bog began to expand. Settlements became aban-
doned as people moved east and south: the evolution of economic activity was changed by
exogenous alterations in environmental conditions (Whittington and Edwards, 1997).
Environmental systems may change for reasons endogenous to the economic system as
well as in response to exogenous factors such as climate change. For instance, in nineteenth-
century Egypt, a change to irrigated agriculture in the Nile valley to produce cotton for
exporting, away from the system of natural flooding that had been in use since 5000 Bc,
resulted in the increasing salinity of farmland, and its eventual abandonment.
Overexploitation of farmlands in areas as diverse as Mesopotamia, Easter Island, the Indus
valley, and the Mayan civilization of Central America caused food production crashes that
changed the course of the development of entire societies (Ponting, 1991).
Economic changes also impact on ecosystem evolution. Examples include the following:
e Introductions of invasive species. For instance, the introduction of possums from
Australia to New Zealand, and earlier introductions of rats and stoats, have changed
the types and abundance of flora and fauna there. Zebra mussels invading the Great
Lakes of the United States have transformed the basic nature of freshwater ecosystem
services.
e Changes in aquatic ecosystems. Changes took place as the Industrial Revolution
unfolded in nineteenth-century Britain. The resultant increased sulphur deposition via
acid rain reduced pH values in lakes and lochs, changing the composition ofvertebrate
and invertebrate fauna over time.

1.2 Key Insights from Economics of which Environmental

Scientists, Environmental Managers, and Politicians
Should be Aware
The previous section set out the interactions between the economy and the environment.
What do economists have to say about these interactions? In the rest of this book, we will
explain the contribution that economics can make to understanding and solving environ-
mental problems. But suppose someone was too busy to read the rest of this book. What, at
a minimum, would we wish to leave them with? The following list is one possibility:
1. Economic and environmental systems are determined simultaneously. This means that
to fully understand these systems, economics must incorporate the mechanical

' There were in fact many phases of change in the woodland area up to 2,000 years ago, whilst the exact causes of
change are still much debated. See Whittington and Edwards (1997),
INTRODUCTION: ECONOMICS FOR THE ENVIRONMENT

underpinnings of the natural sciences, and the natural sciences must incorporate the
behavioural underpinnings of economics.
. The behavioural underpinnings of economics matter for environmental policy. First,
people respond to incentives, as do firms. The most important incentives tend to be
prices. Second, people make decisions ‘at the margin’: in other words, they try to
balance out the costs and benefits of going one step further. Finally, expect firms and
households to usually act in their own best interests. For firms, this typically means
maximizing profits and for households it means maximizing their well-being (their
utility). This implies that we should not be surprised when either behaves strategically:
for example, when someone free-rides when asked to make a donation for an
environmental good cause, or when a farmer threatens to destroy a wetland in return
for a compensation payment not to do so. Institutions need to be designed that take
these kinds of responses into account. Recent advances in mainstream economics that
take a broader view of the motivations behind human behaviour than the standard
model of ‘rational choice’ have become increasingly important within environmental
economics (Shogren and Taylor, 2008).
. Environmental resources are scarce, and using them in one way has an opportunity
cost. By ‘scarce’, we mean that there are not enough environmental resources around to
simultaneously meet every possible demand on them. By ‘opportunity cost’, we mean
the net benefits forgone from the next-best use. For example, suppose that a piece of
land has three possible uses, namely agriculture, forestry, and recreation, which have
returns of £2,000/ha, £3,000/ha, and £4,000/ha. These activities we will assume to be
mutually exclusive, in that the land cannot be used for more than one purpose at the
same time. Deciding to use the land for recreation purposes forgoes a return from
either agriculture or forestry, and the opportunity cost is the next-best return forgone,
namely £3,000/ha. This cost should be taken into account when evaluating the net
benefits of using the land for recreation.
. The free market system can generate the ‘wrong’ level of environmental quality. Too
many environmental bads (e.g. too much pollution) and too few environmental goods
(such as beautiful landscapes) will result from the point of view of social optimality.
Why should this be?—Because the system of property rights in existence means that no
market price exists either to discourage economic agents from polluting or to
encourage them to produce environmental benefits. This problem is known in
economics as market failure: Chapter 2 investigates this issue in detail, and suggests
alternative ways of solving such problems. Another way of thinking about this is to say
that the environment is valuable in many ways, but not all of these show up in market
values or prices. For example, it is hard for a private landowner to charge for the
landscape benefits that his farm ‘produces’, and no market price exists for many aspects
of landscape beauty.
. However, markets have proved to be the best way of allocating a vast range of resources:
Adam Smith’s ‘invisible hand’ still has much to recommend it. Markets are good at
coordinating actions and at transmitting information. For many resources, the market
system is also good at responding to changes in relative scarcity. For example, the
significant hikes in oil prices in the 1970s produced automatic adjustments in supply
and demand. Finally, markets allow people the opportunity to trade, which turns out to
ECO

be a good way of increasing social welfare on the whole. Markets can also be made to
work for the environment: see, for example, the discussion of the idea of tradable
pollution permits in Chapter 2.
. Government intervention does not always make things better, and can make
things worse. The Common Agricultural Policy of the European Union (EU)
has been frequently criticized as having given farmers an incentive to damage
the environment: this might be called ‘government failure’. When
governments interfere with the free operation of markets, they need to be aware
that they are likely to bring about coordination and information problems.
Government intervention may well hinder the responsiveness of markets to
changes in relative scarcity; for example, if they keep prices at levels other than the
market clearing rate (Chapter 2).
. Environmental protection costs money. Scarcity means that opportunity costs exist
for all choices, even those driven by moral imperatives. Protecting endangered species
costs money, both directly (e.g. in monitoring) and indirectly (in that land can no
longer be used for development). Spending more public money on public transport
systems to reduce air pollution may mean that less money is available to spend on
schools. What is more, the costs of protecting the environment typically increase at
the margin. As emissions from industrial sources are progressively cleaned up, each
extra reduction gets more and more expensive to achieve.
. When managing renewable resources such as fish and forests, choosing the maximum
sustainable yield as the best level at which to harvest is rarely optimal. This is because
this rule ignores the economic costs and benefits of renewable resource management.
Catching at the maximum sustainable yield usually means too many boats chasing too
few fish.
. Whilst economic growth may contribute to current environmental problems, few
people would swap their position today with their equivalent 200 years ago, due to the
huge increases in real incomes per capita and improvements in life expectancy. It is
hard to think of economic growth as a bad thing, but it is something that gives rise to a
series of environmental consequences that need to be dealt with.
10. Many ofthe world’s most serious environmental problems are global in nature;
however, economics predicts that getting countries to agree to do something about
these problems together is going to be tough (Chapter 7). This is because game theory
shows us that countries have an incentive to ‘free-ride’ on the actions of others and so,
for example, to avoid signing up for international agreements to cut global pollutant
emissions. However, economists can help in designing institutional arrangements to
reduce these problems.

1.3 The Rest of this Book: An Overview

We divide the book into two parts. Part I explores some important concepts in economics,
and illustrates why they matter for environmental issues. Chapter 2 explores the role of
markets in determining the level of environmental quality, what ‘market failure’ means, and
INTRODUCTION: ECONOMICS FOR THE ENVIRONMENT

how markets can be used to work for the environment. The practice of placing monetary
values on the environment has become controversial, but is an important component of the
economists toolbox: Chapters 3 and 4 deal with this issue, and with the methodology of
cost-benefit analysis. Many environmental decision-making problems are characterized by
high levels of risk and uncertainty, so Chapter 5 introduces economic approaches to these
issues. Sustainable development and green growth are fashionable buzzwords in environ-
mental and development debates at present: Chapter 6 investigates what economics has to
contribute to this debate, and to understanding the relationship between growth and indi-
cators of environmental quality. In Chapter 7, game theory techniques are introduced as a
useful tool to help understand situations in which people behave strategically, be they coun-
tries arguing over climate change conventions or fishermen competing over harvests.
Finally, Chapter 8 lays out some basic economics of the debate over free trade: Does free
trade always increase people’s well-being? Can trade restrictions be justified on environ-
mental grounds?
In Part II, we explore how economics can help us understand the causes of a series of
important environmental problems and, more importantly, provide more environmental
protection at lower cost. The issues studied are:
e Climate change
e Forests and forest loss
e Water pollution
I Biodiversity loss
e Energy policy

These chapters are brief considerations of these important and complex issues, so we only
have space to highlight how economics can be useful in thinking about our impacts on the
natural world. Yet the material in Part II does show (we hope!) the richness of analysis that
economics can bring to these problems.

1.4 Using this Book for Teaching and Learning

This book is aimed at a wide audience. Little economics background is assumed, and the use
of mathematical explanations is downplayed as much as possible. The book is suitable for
introductory courses in environmental and natural resource economics, both for econom-
ics students and non-economists. It should also be suitable for undergraduate- and MSc-
level interdisciplinary courses in which many students lack a background in economics.
Instructors will see that we do not follow a traditional path in explaining the subject: for
example, there is no chapter on the economics of renewable resources per se, nor on the
economics of the mine. The key ideas may all be found either in Part I (e.g. in Chapter 7 on
game theory, for renewables) or in the case studies in Part II (e.g. in the energy chapter for
non-renewables and the forest chapter for forestry economics). We recognize that there are
costs to presenting things this way, but we feel these are outweighed by the benefits: an ini-
tial introduction to important economic concepts; and then a look at important environ-
mental issues, which draws on these concepts and develops others. At the end of each
10 ECONOMIC TOOLS FOR THE ENVIRONMENT

chapter, we include a list of suggested questions for use in tutorials or seminars, and a glos-
sary of key terms. References are provided at the end of each chapter, rather than being
collected together at the end of the book.

References and Additional Readings

Ponting, C. (1991). A Green History of the World Whittington, G., and Edwards, K. (1997). ‘Climate
(London: Penguin). change; in K. Edwards and I. Ralston (eds.),
Shogren, J., and Taylor, L. (2008). ‘On behavioral Scotland: Environment and Archaeology, 8000 Bc-
environmental economics, Review of AD 1000 (Chichester: Wiley).
Environmental Economics and Policy 2(1): 26-44.
Markets and the
Environment

Weall use markets every day. Markets allow us to trade—buying and selling goods allows us
to create economic value. We appreciate the choices and opportunities that markets provide
to us, because we like choice and the freedom to choose. But what is a market to an econo-
mist? To non-economists, a market is a tangible place in which to spend money; such as a
neighbourhood mall or a village market in France, or a global website on which you buy and
sell apples or armchairs. But to economists, the idea of a market is more conceptual—they
see the market as a way to create value through the voluntary exchange of goods and services
regulated by competition.
Economists champion the marketplace as the most useful way to organize economic
activity. Markets arise spontaneously because people can create value in trade. Markets cre-
ate wealth through voluntary exchange of scarce resources, in which prices guide how peo-
ple decide to trade. Wealth is created when resources move from low-value to high-value
uses. Markets create value—the benefits exceed the costs—rather than just redistributing
wealth between people.
Markets are powerful for another reason than just trade. Markets send signals; they are a
channel of communication. Many scientists dedicated to protecting the environment
believe that markets are the most effective tool humans have ‘discovered’ to organize and
coordinate the diffuse set of information spread throughout society; for example, informa-
tion on what people want, how much of the good is in supply, given drought or rainy condi-
tions. Markets use prices to communicate scarcity—as defined by both the laws of nature
and the laws of humanity. Prices send signals to coordinate decentralized economic deci-
sions. Markets succeed when prices define the trade-offs we face such that resources are
allocated to their highest-valued use in society.
But markets fail too. Markets fail when prices send bad information about the true nature
of scarcity, or when markets fail to exist at all because they are too costly to construct. For
example, we suffer from over-pollution or overdevelopment when a market price is too low
to communicate our preferences for environmental protection. Prices might misstate the
economic value of a reduction in health risk from an environmental threat, or prices might
not even exist to signal the value. Left alone, a market might produce too few or too many
goods or services. A wedge is driven between what people want as individuals and what
society wants as a collective.
12 ECONOMIC TOOLS FOR THE ENVIRONMENT

The protection of endangered species on private land is a classic example of market fail-
ure. By one estimate, about half of the listed endangered species in the United States have
80 per cent of their habitat on private land. The challenge is that the benefits of the protec-
tion of endangered species extend to everyone—the entire world—but the costs of protection
fall on private landowners. Since the market price of private land does not capture the social
benefits of species protection, landowners have more incentive to protect their own private
investment (e.g. their capacity to produce cattle) than to protect endangered species. The
market fails when private decisions generate a less-protected habitat than society desires.
But even when markets fail, they can be the cornerstone of the solution. Rather than turning
to more government regulation or stakeholder-participation processes, society can fix existing
markets or create new markets to manage our environment and natural resources; for exam-
ple, when we create markets in rights to emit pollution. Since a market is a tool, its precision
depends on how society defines the rules to regulate its behaviour; that is, property rights,
liability, and information. People who dislike the prices that a market produces have to rethink
the connection between price signals and market rules. We can work together to change these
rules. We should view markets as having the potential to work for us, not against us.
This chapter explores the nature of markets, market failure, and market redemption. We
discuss the power of markets, and why economists continue to promote markets despite
their flaws. We examine how markets can fail the environment due to externalities, public
goods, common property, thresholds, and hidden information. We end by exploring how
we can use elements of markets to correct market failure, or policy failure due to govern-
ment subsidies.

2.1 The Power of Markets

A market serves society by creating value through free exchange. Markets use prices to com-
municate the wants and limits of a diffuse and diverse society so as to bring about coordinated
economic decisions efficiently. The power of a market system rests in the decentralized pro-
cess of decision-making and exchange. No omnipotent central planner is needed to allocate
resources. Rather, market prices ration resources to those who value them the most, and in
doing so under certain conditions, people are swept along by Adam Smith’s ‘invisible hand’ to
achieve what is best for society as a whole. Self-interest is the driving power, and competition
the regulator of that power—together, they work to improve the lives of common people.
A key idea behind the power of markets and free exchange is comparative advantage. One
person has a comparative advantage over another person in one good relative to another
good if that person’s relative efficiency in the production ofthe first good is greater than that
of the other person. Alternatively, a person has a comparative advantage if his or her oppor-
tunity cost is less than that of the other person. The opportunity cost is the economic cost as
measured by what we have to give up to do something else.
Absolute advantage does not translate into comparative advantage. Everyone, skilled or
unskilled, has a comparative advantage at something, because opportunity costs are lower
for some people for different activities. Markets and trade benefit everyone in society
because they allow people to specialize in activities in which they havea comparative advan-
tage, and then trade for what they want to consume (for an extension of this idea to the
MARKETS AND THE ENVIRONMENT 13

benefit of trade between nations, see Chapter 8). A market succeeds when it allows for the
efficient allocation of resources. Efficiency gains exist in an economy if it is possible to trade
goods and services such that at least one person is better off and no one else is worse off. This
is called Pareto efficiency—the inability to reallocate resources without making at least one
person worse off.
For example, should you shovel your own snow? No, not necessarily. If you follow the
idea of opportunity cost and comparative advantage, it might be to your and to society’s
advantage to use the market and hire someone to shovel your snow. Everyone has a com-
parative advantage at something, based on his or her relative opportunity cost. Suppose you
can shovel snow better than the next person, but it might be that your opportunity costs are
too great. Say you could remove the snow in two hours; but in those two hours you could
have written a new essay on Napoleon’s wild youth and earned $5,000 selling it to a blog on
historical French leadership. In contrast, your teenage neighbour could do the job in four
hours at $10 per hour, so it would cost you $40 to remove the snow. The teenager’s oppor-
tunity cost is that he or she would have to give up gaming on the PlayStation for four
hours—which we assume he or she values at less than $40. Both you and the teenager gain
from the trade: you write the essay and get paid, the teenager shovels your snow and gets
paid, and society is better off since you both gained—it is a Pareto-efficient trade. Box 2.1
illustrates the idea of gains from trade, by means of an example.
What is the key to a successful market? Most economists agree that well-defined property
rights are the key. A well-defined property rights system represents a set of entitlements that
define the owners’ privileges and obligations for use of a resource or asset. A well-defined
property rights system is based on four characteristics:

e Comprehensive—all resources are either privately or collectively owned, and all

entitlements are defined, well known, and enforced.
e LExclusive—all benefits and costs from use of a resource accrue to the owner(s), and only
to the owner(s), either directly or by sale to others. This applies to both private and
common-property resources.
e Transferable—property rights should be transferable from one owner to another
through a voluntary exchange. The owner has an incentive to conserve the resource
beyond the time during which he or she expects to make use of it.
e Secure—property rights to resources should be secure from involuntary seizure or
encroachment by other people, firms, and governments. Security provides the owner
with an incentive to improve and preserve a resource while it is in his or her control,
rather than exploit the assets.

These four conditions represent an ideal scenario in which gains from trade are created
and captured by people who trade. The idea of a complete set of markets does not reflect
reality, however; rather, the notion of complete markets represents a theoretical benchmark
against which economists can judge the effectiveness of different plans to organize eco-
nomic activity. This orthodoxy of maximization and equilibrium to capture the interactions
of intelligent self-interested people is a necessary fiction rather than literal truth.
Markets also force people to make a distinction between rhetoric and action in the con-
text of environmental assets. We all have opinions. Markets help separate those opinions we
14 NOMIC TOOLS FOR THE ENVIRONMENT

Consider the following market. Suppose that a market exists for a classic Ansel Adams photo of the Grand
Tetons National Park, with eight buyers and eight sellers. Each buyer has his or her maximum willingness
to pay (WTP) for one photo; each seller has his or her minimum willingness to accept compensation
(WTA) for one photo:

Buyer ID WTP Seller ID WTA

Bl $300 S1 250
B2 $200 S2 350
B3 $50 53 150
B4 $500 S4 450
B5 $300 S5 250
B6 $250 S6 100
B/ $400 S7 200
B8 $100 $8 100

Using the WTP and WTA data, we can graph the market demand and supply curves, and then calculate
the equilibrium price and quantity, and the total gains from trade. First, we can rank-order the buyers
from highest to lowest WTP, and draw them as a set of downward steps. This is the market demand curve.
Now we rank-order the sellers from lowest to highest WTA, and graph them as a set of upward steps. This
is the market supply curve. When market demand intersects market supply, we find the market
equilibrium—the market clearing price and quantity sold. In this example, then, the equilibrium market
price equals $250 and the equilibrium quantity is 4 or 5.
How do economists measure the gains from trade? We have defined the total gain from trade as the
sum of two measures of welfare—the consumer surplus and the producer surplus. The consumer surplus
(CS) represents the gains to each buyer, and is represented as the difference between each buyer's WTP
and the market price:

CS = ($500 — 250) + ($400 — 250) + ($300— 250) + ($300—250) + ($250—250) = $500.

For example, buyer 7 was willing to pay $400, but since the market set the price at $250, that is all he had
to pay; his consumer surplus equals $150 (= $400—250). The consumer surplus represents his gain from
exchange. If we add up each buyer's consumer surplus we have measured the total benefits to all the
consumers in this market.
Similarly, the producer surplus (PS) represents what the sellers gain from this market. Here the
producer surplus is the difference between the market price and each seller's WTA:

PS = ($250 — 100) + ($250 — 100) + ($250 — 150) + ($250 — 200) + ($250 — 250) = $450.

For example, seller 6 was willing to accept $100 to sell, but since the market set the price at $250, her
producer surplus equals $150 (= $250 — 100). Again, if we add up each seller's producer surplus, we have
measured the total benefits to all the producers.
Finally, the total gains from trade in this market equal the sum of the consumer and the producer
surplus, which in our example is $950 = ($500 + 450). This is represented by the shaded area between the
demand and supply curves. Both the low-cost sellers and the high-value buyers gain in this market. The
market is efficient because it has helped move low-valued resources to high-valued uses (for more
details, Krugman and Wells, 2008).
MARKETS AND THE ENVIRONMENT 15

are willing to back up with real economic commitments from those we are not. The disci-
pline provided by the market forces people to relate their choices to the choices of others
and to the consequences ofthese choices.
People tend to overstate their real willingness to pay or to co-operate—say, to protect
threatened tree frogs or critical wetlands—when asked a hypothetical survey question. Good
intentions are just that; but markets do not sustain cheap talk. Markets force us to decide how
to use our scarce resources for development or conservation, or some combination of the two.

2.2 Market Failure

Market failure is a reality that we must confront. Market failure exists when resources do
not attain their highest social value. For environmental goods and natural resources, mar-
kets fail when benefits and costs cannot be allocated with precision across and within
nations and generations. Sometimes, the needed conditions of well-defined property rights
do not hold up for environmental goods. Markets fail when private self-interested actions
could still be improved on relative to collective goals.
Market failure comes about when people cannot define property rights clearly. Markets
fail when we cannot transfer rights; when we cannot exclude others from using the good; or
when we cannot protect our rights to use the good. Under these conditions, free exchange
does not lead to a socially desirable outcome, because private actions provide too many
‘bads’, such as pollution, or too few goods, such as open space. Since everyone ‘owns’ the
right to clean air and biodiversity, nobody owns the right. This makes it a challenge for a
market to operate effectively. The market system is said to be ‘incomplete’, and we have the
problem of ‘missing markets’.
For example, most societies do not currently have well-defined rights to produce or con-
sume pig odour (i.e. the smell of manure) from a large-scale pig-farming enterprise. Those
up- or downwind cannot buy or sell tickets for fragrant air. The pig farm upwind cannot sell
fresh air; those downwind cannot buy fresh air. Since the farm does not bear the downwind
costs, it can ignore these costs. With incomplete markets, the pig farmer lacks a motivating
economic incentive to control emissions or to switch to less-polluting practices. Similarly, if
no legal or institutional basis exists, people who use polluted river water cannot receive com-
pensation from upstream farmers whose sediments, pesticides, or fertilizers impose down-
stream costs in the form of contaminated drinking water, poor fishing, or reduced recreational
opportunities. Farmers can impose ‘external costs’ on these other users of the river.
A market can fail for several reasons. Economists use a taxonomy to help identify and cat-
egorize the different types of market failure. Understanding how and why a market fails is the
first step to correcting the problem. We examine four types of market failure for environmen-
tal resources—externalities, public goods, open-access common property, and hidden informa-
tion. Some overlap exists between these types of market failure, whilst there are other types of
market failure—such as market power or monopoly—that we do not discuss here.
Externalities. Externalities are the classic type of market failure for environmental prob-
lems. Pollution is an externality when a person or firm does not bear all the costs or receive
all the benefits of his or her actions. An externality can arise when the market price or cost
of production excludes its social impact, cost, or benefit. If you look around, you can see
16

externalities everywhere. Think of how your actions affect other’s well-being, for better or
worse, and how you do not pay or receive compensation for the extra costs or benefits. The
market fails because the market is missing—no exchange institution exists in which people
pay others for extra benefits or receive compensation for extra costs (see also Box 2.2).

Failing to put the correct ‘price’ on using the environment can lead to too much environmental
degradation. Factory owners who decide to increase output face no cost for any resultant increase in
pollution. This suggests that they have no economic incentive to cut down on emissions, and then society
has too much pollution. Another example is the air pollution arising from driving to work. When people
drive to work, they pay no immediate price for the pollution coming from their cars.
Jurado and Southgate (1999) explore these factory and automobile externalities in a study of air quality in
Quito, the capital of Ecuador. Quito lies high up in an Andean mountain valley, which exacerbates its air quality
problem. The table below shows that the major sources of pollutants are vehicles and factories. Neither factory
nor vehicle owners face any immediate economic price for the pollution for which they are responsible.

Total suspended Sulphur dioxide (SO,) Nitrous oxides (NO,)

particulates (TSP)
Vehicles 1,069 659 5,298
Factories HALO 18,707 5,023

These levels of air pollution exceed the maximum recommended levels set by the World Health
Organization (WHO). The concentrations of TSP in 1991 averaged 149.9 g/m3, compared with the WHO
standard of 60 g/m?.
The authors estimated the economic costs ofair pollution as measured by the impact on human health.
Pneumonia and other respiratory ailments are the leading cause of death in all age groups in the city, and
are exacerbated by the high TSP levels. The study used statistical relationships between TSP levels and
three standard measures of ill-health: restricted activity days, work days lost, and excess mortality.
The predicted increases in ill-health from pollution were valued in economic terms using several
approaches, including the value of working time, and the cost of illness (e.g. hospital resources used up in
treating patients; see rows 2 and 3 of the table below). Extra deaths were valued as the discounted value
of lifetime earnings, a somewhat dated approach. This yields a value of $16,887 per avoided death, which
then gives the values in row 4 of the table.
gaeee ee ee ce eee ee
Annual costs to citizens
of Quito (US$)
Restricted activity day costs (given 3,433,000 restricted days) $14,418,600
Working days lost costs (given 1,765,000 working days lost) $12,708,000
Excess mortality costs, as per study (given 94 extra deaths per year) $1,587,378
Excess mortality costs, revised $12,220,000

We have used ‘value of statistical life’ estimates based on how much people are willing to pay for risk
reductions (for details, see Chapter 5). The value of the reduced risk is increased by an order of magnitude
in row 4. These new estimates are based on the figure of $130,000 per avoided death from Fankhauser
et al. (1998) for developing countries. Whatever the values used, the message is clear: externalities due to
private air pollution decisions impose social costs on the citizens of Quito.
MARKETS AND THE ENVIRONMENT 17

For instance, driving your car around town creates numerous externalities. The exhaust
contributes to air pollution, driving at rush hour adds to congestion, road rage increases the
risk to others and yourself, and the beautiful racing flames painted on the side add to cul-
tural pride. No explicit markets exist to exchange the good for the bad. You live with it. But
from an efficiency viewpoint, in which society is trying to get the most out of its limited
resources, the lack of a market leads to too many exhaust fumes, too much congestion, too
much road rage, and so on.
Consider another example of a negative externality, or external cost, in a local environ-
ment. The village of Centennial sits at the base of the Snowy Range Mountains. Not many
people live in Centennial, and development of new houses is slow. The houses and yurts
(originally, a Central Asian dwelling type) that exist sit below the ridgeline, so they do not
stick out as people look towards the horizon. Citizens of Centennial and people driving
through enjoy the wide-open space that the current development promotes. But suppose
that a newcomer called Riley moves into the valley and wants to build a new house on his
private property. This property sits on a prominent ridge, a ‘hogback; and the proposed
three-storey house is to be built on top of the ridge, visible to everyone in the valley.
Suppose that Ole lives below the ridge and has an undisturbed view of Centennial valley
and the mountains beyond. If Riley, the newcomer, builds his house on the ridge, Ole’s
view of the wide open spaces in Centennial valley will disappear. Riley’s actions reduce
Ole’s well-being. But Riley does not have to pay compensation to Ole or anyone else, since
no one owns the right to the vista. Riley’s private decision to build his house on his own
land does not account for the losses suffered by the rest of the community who valued the
open space—a wedge is driven between the private and socially optimal allocation of
resources.
Figure 2.1 illustrates this example. Riley accounts for his own marginal benefits and costs,
and chooses to develop his house at level H’. If he had accounted for the social costs of lost
open space suffered by Ole and the rest of the community, Riley would have developed at
level H*—a smaller house built below the ridge. Since H’ > H*, the market is said to have
failed to allocate resources efficiently—there is too much housing on the ridge in Centennial
valley.
This example involves a loss of aesthetics; other externalities affect matters of life and
death. Many externalities increase the risks to human life and limb. Toxic wastes that leach
into drinking water are an example; urban air pollution due to transportation, which cur-
tails the activities of young children, is another. All actions that alter the health risk to others
in which no compensation is paid or received create externalities that are left unattended by
the market.
These examples of pollution reflect a direct externality—you breathe or drink polluted air
or water and there is a direct impact on your health. But over the past century tracking the
direct effects of an externality has become more complicated, as humanity has developed
new technologies to separate itself from the whims of nature. In many cases, it is less obvi-
ous to determine the cause and potential effects: the effects are less direct and more rounda-
bout. The idea of ecosystem externalities captures these indirect impacts. An action affects an
environmental system at one obvious point, at which no harm seems to be done. But the
action is working its way through the system, and can show up somewhere else as an unex-
pected and unwelcome surprise. For example, citizens who kill predators in order to protect
18 ECONOMIC TOOLS FOR THE ENVIRONMENT

file H’ H

Figure 2.1 Marginal benefits and marginal costs: private and social.

children and domestic animals can generate large rodent infestations that affect crops and
sanity. DDT does not kill birds but, rather, it thins the shells of their eggs.
Ecosystem externalities highlight the fact that the economist cannot presume that the
cause and effect of production and consumption decisions are obvious; and that even
though there is no effect ‘under the streetlamp’, there might be an unanticipated effect
somewhere else (Finnoff and Tschirhart, 2008). Economists have to work with natural sci-
entists in the life sciences to better anticipate the potential for indirect externalities, feed-
backs between the economic and ecological systems, and the subsequent unintended and
unpleasant surprises.
Public goods. Public goods are a second form of market failure. A public good exists when
a person cannot be excluded from its benefits or costs—non-excludability; and when one
person’s consumption ofthe good does not reduce its availability to anyone else—non-rival
consumption. Together, non-excludability and non-rival consumption are what separates a
public good from a private good, which is excludable and rival in consumption.
Economists make use ofthe terms pure and impure public goods. The difference is that a
pure public good is both non-excludable and non-rival; whereas an impure public good
might be either non-excludable or non-rival, but not both. Climate-change protection, the
ozone layer, and biodiversity are examples of pure public goods in which the benefits accrue
to all those around the globe. Common-property and club goods such as rivers, local parks,
and lakes are impure public goods because the benefits can be excluded from non-members
of the group that owns the resource. We focus first on pure public goods, and then turn to
common property in the next section.
Non-exclusion depends on the physical characteristics of the good and the property
rights regime. Climate-change protection is the most obvious form of public good in the
MARKETS AND THE ENVIRONMENT 19

environment. No nation can be excluded from the benefits of the emission reduction
efforts of another nation. Biodiversity preservation is another example of a public good.
No person can be excluded from the public benefits created from a stable ecosystem cre-
ated by preserving species. For wildlife viewing, unless the species exist on private reserves,
it is difficult to exclude potential beneficiaries from enjoying the prospect of seeing them.
If people derive value from the existence of a species, exclusion is impossible. This holds for
many environmental resources—if air quality in a city improves, nobody who lives in or
visits the city can be excluded from the benefits. Other environmental resources are exclud-
able in consumption. For example, market prices control recreational access to downhill
skiing at a resort.
Non-rivalry depends on the characteristics of the good. Non-rivalry means that the ben-
efits gained from the resource are independent of the number of other people who wish to
use it. Better air quality, for instance, increases one person’s well-being, no matter how
many other citizens there are. If an endangered species is protected, the number of people
affected does not reduce the benefits to each person. Some environmental resources do not
possess this property. If designating a wilderness area as a recreational resource protects it
from future development and yet also increases its profile, then the presence of more people
implies more congestion and fewer benefits per trip for many visitors. Many people visiting
Yellowstone National Park, for example, feel that congestion is a ‘bad’: too many people, too
little solitude.
What is the relevance of these properties for the socially optimal provision of public
goods and market failure? Let’s revisit the Centennial open space issue as a public goods
problem. Open spaces are a public good if defined as an aesthetic available to all people in
the same amount—my visual consumption of open space does not reduce your access, nor
reduce the amount of open space. The potential problem with using the market to provide
this public good voluntarily is free-riding—since he or she cannot be excluded from the
same amount of the good, each person has an incentive to let someone else provide the
public good. Everyone has an incentive to free-ride off the efforts of others.
Free-riding can lead to the well-known case of the prisoner’s dilemma, also called the
social trap or the tragedy of the commons (for more details, see Box 2.3 and Chapter 7). The
dilemma exists when people find that individual incentives lead them to the worst outcome
possible—for themselves and society. The idea is compelling: a person looks out for his or
her own self-interest and knows that other people are doing the same. Given the incentives,
he or she cannot avoid the worst outcome, because the private incentives always push the
person towards the non-cooperative outcome. In public good terms, this means that a per-
son knows he or she should not free-ride, because it is better to be a good citizen. Yet the
incentives still exist to free-ride. If he does not free-ride but everyone else does, he ends up
in the worst position.
In reality, people do contribute to the provision of some public good voluntarily. Whether
people contribute to the public good is not the question. The question is whether they vol-
untarily contribute ‘enough’—do people pay enough to create the social optimal level of the
public good? If so, a decentralized market approach has succeeded in allocating a public
good. If not, by definition the market fails. The market fails because people have undersup-
plied the social optimal level of the public good, or oversupplied a public ‘bad’ such as pol-
lution. Free-riding can be partial or complete; the market might fail by a little or a lot. But
20 ECONOMIC TOOLS FOR THE ENVIRONMENT

Biologist Garrett Hardin coined the phrase ‘the tragedy of the commons’ in the journal Science in 1968.
The word ‘tragedy’ meant a ‘remorseless working of things’. Hardin was concerned with unchecked global
population growth, which he saw as unsustainable for common property. His example of a pasture open
to all graziers to let their cattle roam over leads to overgrazing. Each grazier gets the full benefit of adding
one more animal to the common (the profit per beast). But as Eattle numbers rise, overgrazing occurs,
which then causes everyone to lose. Each person rushes headlong down a path that leads to social ruin,
destroying the idea of Adam Smith’s beneficial invisible hand. This tragedy gets serious, Hardin argues, as
population rises, since rising population equates to more people wanting to graze their animals.
Hardin's analysis has been praised, criticized, and extended. We now know that overgrazing of
common areas is not new—it happened in ancient Mayan civilization and ancient Egypt. We also know
that the price of the good produced on the commons matters more to the rate of exploitation than
Hardin thought. We also know that many commons throughout the world are regulated by rules set by
those who have access to the area.
Yet Hardin’s message is powerful. We can find many examples of open-access resources damaged by
overuse. An example is a study of common grazing areas in the Northern Isles off the Scottish coast. On
Shetland, common grazings seem to suffer higher levels of environmental damage than comparable
privately owned areas.

technically, the market still fails. Market failure occurs with voluntarily public good provi-
sion when people contribute any amount less than their true benefits for the good.
Consider the incentives to free-ride on open space provision. Suppose that Ole and the
citizens of Centennial are considering pooling their resources and using the market to buy
out Riley, to stop him from building his new mansion on the hogback. Since no one can be
excluded from the potential benefits of open space, the marginal benefit to society from one
more unit of visibility equals the summed marginal benefits from each person who gains
enjoyment from the view. Marginal social benefits reflect how society values each extra unit
of the public good. The ‘optimal level’ of the public good is the level that maximizes the net
benefits to society. The optimal level is determined by comparing the marginal social ben-
efits to the marginal social cost to provide one more unit of the good. The public good
should be provided up to the point at which the marginal social benefits equal the marginal
social costs. This optimal level is called the Lindahl equilibrium, named after the Swedish
economist who first identified it.
If the market fails to generate the optimal level of the public good, how can society attain
the optimal level? One possibility is to build a club: everyone who benefits from the good
gets together to provide the good, and the benefits of this provision are then restricted to
those people belonging to the club. This seems attractive since the beneficiaries of the good
pay for it, and government is not forced to intervene. Organizations such as the Royal
Society for the Protection of Birds (RSPB) in the United Kingdom and the Sierra Club in the
USA, are partial examples of this. The RSPB buys and safeguards nature reserves using
funds generated by club members.
But while club members get reduced fees for entry to sites, non-members also benefit
from environmental preservation. The non-excludability of benefits means that fewer
nature reserves would be provided than optimal if their provision was left to the RSPB.
MARKETS AND THE ENVIRONMENT 21

What is more, club financing is likely to be less than optimal, in that people do not have to
join and pay, even though they know they will benefit anyway. The strategic incentive to
free-ride still exists.
Common property. Common property, or the commons, can be managed in many ways
depending on how property rights are defined and enforced. Open-access commons exist
when people cannot be excluded from accessing a resource, such that one person’s use
rivals another’s use. If a person’s use reduces the total available to all, everyone has an
incentive to capture the benefits before someone else gets them. This open-access free-for-
all leads to inefficient use of the resource. The moratorium on fishing for cod and witch
flounder off the Grand Banks in the North Atlantic is a prime example. By inefficiency, we
mean that the fishers harvest to the point at which the marginal costs exceed the marginal
revenue (i.e. the market price) of harvesting. Overuse implies that the market price has
failed to signal the true scarcity of the asset. Since a fish caught by one fisher is one fewer
fish for all others to catch, fishers have no private incentive to account for the scarcity value
of the resource. They expend too much effort and end up overharvesting the fish stock rela-
tive to what is economically efficient and potentially biologically inefficient. Again, we
return to this problem in Chapter 7.
What happens in open-access commons? Each fisher has an incentive to catch as many
fish as possible before someone else catches the same fish. He has no incentive to value the
scarcity of the fish, because if he does not catch them, someone else will. His decision to let
the fish be is not respected by others, because they have as much right to the fish as he does.
He starts expending effort and in doing so his effort is such that his marginal costs end up
exceeding his marginal revenue. This violates the standard efficiency condition that says
that net benefits are maximized when marginal revenue equals marginal costs.
Why? Efficiency is breached because each fisher imposes an external cost on all the other
fishers. This external cost arises because each time a fish is caught, it makes it more costly for
everyone else to catch the next one. This is like a variant of the infamous 90-10 rule of dimin-
ishing returns to effort—you get 90 per cent of the fish with 10 per cent of the effort, but catch-
ing the remaining 10 per cent takes 90 per cent of the effort because they are harder to find.
When do the fishers stop their effort in open access? A fisher stops when his marginal
costs equal the average revenue, or when his economic profits are zero. In this case, even
though net profits to fishers are zero, the net social value is negative, because the scarcity
value of the fish has been ignored in the free-for-all. In this case, each fisher does not earn
the economic rent that would reflect the scarcity value of the fish to society.
Open access is the classic case of a ‘tragedy of the commons’. Since everyone has access, all
have the rights to the resource and scarcity value is ignored. But the reality is that most com-
mons havea property-right scheme, either formal or informal, that works to allocate resources
in a more efficient manner. There are numerous documented examples of self-governing
commons in which people work as a collective unit and respect the scarcity value of the
resource. These groups succeed when they establish common-property rights that include
sharing rules, exclusion principles, and enforcement and punishment schemes (see Ostrom’s
classic 1990 book of examples of alternative management schemes for common property).
Market failure need not always occur with the commons. People have defined—and con-
tinue to define—rules to capture the scarcity value of a resource shared by many people. Be
aware that market failure is associated with the commons in situations in which people use
22 ECONOMIC TOOLS FOR THE ENVIRONMENT

the term ‘the commons or the ‘global commons’ when they mean a global or regional public
good that is both non-rival and non-excludable. Pure public goods are much more of a chal-
lenge to handle because of the need for a larger coalition of affected parties. We expand on
this in Chapter 9, on the economics of climate change.
Hidden information. Market failure occurs when people cannot observe either the actions
or types of other people. Moral hazard confounds market operations because one person
cannot observe the hidden actions of others. Adverse'selection frustrates markets because a
person cannot observe the hidden type of a person (e.g. whether a farmer has high or low
opportunity costs for increasing biodiversity on his or her land) or the hidden quality of
some good or service (e.g. whether or not a loaf of bread is organic). Both types of hidden
information slow the creation of markets that could be used to allocate resources to more
efficient use, such as the reduction of environmental risk.
Moral hazard implies that a regulator cannot perfectly monitor pollution abatement, and
a firm could shirk on pollution control. The firm has an incentive to shirk if it bears all the
control costs in return for a fraction of the benefits. The result is again too much pollution
relative to the efficient level. Also, moral hazard can lead to an inefficient ‘pooling’ of envi-
ronmental risk, a topic that we take up in Chapter 5.
Environmental risks are part of life, and it would be better to find markets to allow those
who are less willing to bear risk to sell the risk to those who are willing to buy it. But moral
hazard reduces the ability to reallocate risk among different economic agents. When the
private market cannot monitor actions, an insurer withdraws from the pollution liability
market because insurance affects the individual’s incentives to take precautions.
Given that accidental spills or storage of pollution can create potential financial liabilities
(e.g. clean-up costs, medical expenses), a firm would like to pay to pass these risks on to a
less risk-averse agent such as an insurer. But since there is a trade-off between risk-bearing
and incentives, the market for pollution liability insurance is incomplete, as insurers reduce
the information rents of the better-informed individual. The market produces an inefficient
allocation of risk.
Adverse selection affects the environment too. One key recommendation of environ-
mental policy is sustainable production of products. These products should be produced
using methods that protect the environment. The challenge facing these sustainable prod-
ucts is that these products are more expensive and demand a higher market price. But if
buyers cannot be guaranteed that they are getting their ‘sustainability’ bang for their buck,
they will not buy them. Reporting that a product is sustainable is not enough for many
consumers—they need a warranty or label.
If a consumer cannot distinguish the sustainable product from a similar product pro-
duced using standard practices, he has no incentive to pay the price premium. Why should
he or she pay more than average when he or she cannot distinguish high-quality from low-
quality products? The problem is that if the buyers disappear, the sellers have to withdraw
from the market. ‘The sellers with an above-average quality product who cannot get a price
premium have no incentive to stay in the market, because they could do better investing
elsewhere with greater returns. And if enough buyers who are willing to pay a high price
disappear, the sellers disappear too, and the market collapses. Preventing this collapse
requires a voluntary approach to certification (e.g. the AB scheme in France, or the Soil
Association scheme in the UK) or a government-sponsored certification scheme.
MARKETS AND THE ENVIRONMENT 23

2.3 Markets for the Environment

Market failure can create environmental problems. But even if markets fail, we can still use
the ideas behind markets to address the problems that might exist. We do not need to turn
to command-and-control government intervention or collaborative stakeholder processes.
Rather, people can rewrite the rules to create new markets to address the failings of existing
markets. They can use market-based policy as a substitute for technocratic or stakeholder
processes, which have their own successes and failures (see Box 2.4).
Think about other kinds of real issues—such as financial assets, for instance—in which we
are much less willing to delegate decision-making authority to the government or stakehold-
ers. That people have been creating and using markets to manage many assets for the past
three centuries signals their power. For example, the relative stakes per percentage risk are
larger in financial assets than in risks from environmental dilemmas. But we do not ask the

In this chapter, we are discussing how market failure can lead to undesirable consequences for the
environment. Governments have a case for some level of intervention to do something about it.
Sometimes, however, government intervention can make things worse for the environment. A good
recent historical example of this is the Common Agricultural Policy of the European Union, known as the
CAP.
Through a complex system of import levies, export subsidies, and intervention buying, the CAP offered
farmers throughout the EU higher prices for their outputs than the free market would generate. The
official reasons given were to support farm incomes, stabilize crop prices, and increase self-sufficiency
rates for major food products. Casting aside the dubious economic arguments that could be made in
support ofthese goals, let us focus instead on what the implications were for the environment. Farmers
responded to higher prices by increasing output, both by cultivating more land and increasing the
intensity of production on all farmland. Coupled with technological progress, this policy altered the
countryside, especially in a country such as the United Kingdom, where farmland accounts for about
80 per cent of the total land area. The most important impacts on farming historically are as follows:

© The removal of hedgerows, the ploughing of field margins, and the loss of farm woodlands
@ Replacement of permanent pasture with temporary pastures and arable
@ Land drainage
© Significant increases in the use of pesticides, herbicides, and fertilizers

The effects of these changes in farm practices on wildlife and landscape were extensive.
Figures from the Nature Conservancy Council published in 1983 showed, for example:

e A50-60 per cent loss in lowland heaths since 1949

e A95 percent loss in herb-rich flower meadows
@ An80 per cent loss in chalk grasslands of high ecological value

The CAP has since been reformed, with the removal of many aspects of direct commodity price support
and the introduction of a range of agri-environmental schemes that offer farmers payments to undertake
actions likely to produce environmental benefits, such as reducing grazing pressures or cutting fertilizer
or pesticide application rates. For an analysis of the effects of agricultural land use on one indicator of
biodiversity—birds—see Dallimer et al. (2009).
24 ECONOMIC TOOLS FOR THE ENVIRONMENT

government to dictate the price of stocks and bonds. We ask the government to help estab-
lish, monitor, and enforce the trading rules of the market, but not the market price itself.
We now consider three ways to create new markets to address market failure associated
with the environment. A market has a supply side and a demand side, which together pro-
duce quantity exchange at a market price that reflects the value of the asset. Working from
this basic construction, we consider three options. First, we can assign property rights for
environmental assets and let people negotiate over ‘the price and quantity of the good.
Second, we can work through regulators to set a market price per unit of the environmental
asset and let people decide how much of the asset they want to buy. Third, we can use regula-
tors to set the quantity of the asset that can be bought and let the people decide what price
they are willing to pay for the fixed quantity. Let us consider each in turn.
Assign property rights and bargain over price and quantity. In 1960, economist Ronald
Coase argued that we can create new markets for non-market goods such as the environ-
ment as long as we are willing to remove institutional constraints to assigning well-defined
property rights. Coase noted that two parties have an incentive to negotiate an economically
efficient and mutually advantageous solution to a dispute provided that one party is given
unilateral property rights to the asset in question. The key point for efficiency is that prop-
erty rights are assigned, not to whom they are assigned (that is an equity issue). The outcome
is the same—an efficient allocation of resources. This is the Coase theorem as traditionally
defined. The theorem only holds if transaction costs are low and legal entitlements can be
exchanged and enforced. The transaction costs are the price paid to organize the economic
activity, including information, negotiation, writing and enforcing contracts, specifying
property rights, and changing institutional designs.
Consider our open space example. Riley and Ole disagree about the amount of open
space in Centennial valley. Riley owns the right to build his house on the hogback and Ole
owns his own land below. All the citizens own the right to the view of the Centennial valley
and ridge around the Snowy Range Mountains. But again, the rub is that if everyone owns
the right to the view, no one owns the right.
One solution to the conflict is for the government to intervene and restrict Riley’s devel-
opment, or to tell Ole to live with it. This command-and-control approach allows a third
party to select the winner and loser(s). In contrast, the citizens or government could work
with the community to create a dispute resolution process in which everyone could sit down
at the bargaining table, and try to come to some agreement. This would require the stake-
holders to agree to some solution. It would promote collaboration, and by avoiding polar-
ization it would produce creative solutions with political momentum. This could support
local leadership and collaborative efforts to help Riley and the others enhance the environ-
ment and achieve economic productivity—or it could collapse into a bitter confrontation,
with no resolution.
Alternatively, we could promote a Coasean solution that would create a market for open
space by assigning property rights to either Riley or Ole. First, suppose that a third party
assigns the property rights to Ole. Ole now has the right to keep the open spaces as he sees
fit. This means that he could choose to keep the open spaces to his liking, and have the legal
power to ask Riley to build a house that would not disturb the open spaces. This would make
Ole the supplier of development, and he would have to decide how much open space
to surrender to development under a specific price schedule. Ole’s supply schedule of
MARKETS AND THE ENVIRONMENT 2

development space reflects his marginal costs from development, which increases as devel-
opment increases: the more development, the greater loss of well-being as measured by
more marginal costs.
Riley would now have to come to Ole to buy the use of open space for development. Riley
would have to decide how much open space he would demand for a specific price. His
demand schedule for development space would reflect his marginal benefits from develop-
ment, which are assumed to be decreasing in development—more development implies
lower marginal benefits. The market for development space would clear at the market price
at which all the space demanded would equal all supplied. Both Ole and Riley would benefit
from the trade: Ole would earn benefits pay through receiving a price for open space that
exceeded his opportunity cost; Riley would benefit from paying for space that is less than his
benefits for developing. The market leads to an efficient outcome—the marginal benefits
from development equal the marginal costs.
Now suppose instead that the third party assigned the property rights to Riley. The roles
are reversed—Riley has the right to use up open space as he sees fit; Ole has the right to try
to buy open space. The power of the Coase theorem is that the outcome would be the same
as before—same market price, same market quantity, same efficiency result. The economic
efficiency is the same. Note, however, that the distribution of wealth differs. Now Riley
receives the payment from Ole. Whether this is preferable to Ole receiving the wealth is a
question of ethics.
What happens is that we redefine Ole and Riley’s schedules to reflect the new property-
right structure. Ole’s supply curve for development now becomes his demand curve for
open space. Ole now demands open space—different levels for specific prices. Riley’s
demand curve for development now becomes his supply curve for open space. He chooses
to sell different amounts of open space at specific prices.
The market for open space would again clear at the identical market price at which all the
space demanded would equal all supplied. Ole and Riley would capture the same benefit
from the trade: Ole would earn benefits by paying a price for open space that was less than his
opportunity cost of development; Riley benefits from selling open space at a price that
exceeds his costs for forgoing developing. Again the market is efhcient—the marginal bene-
fits from open-space preservation equal the marginal costs. The smaller the number of people
involved in the dispute, the more likely the Coase theorem is to work. More people increase
the transaction costs necessary to come to an agreement, and make the market less efficient.
It is also hard to identify responsible parties in some cases, and people have more incentive
to free-ride in larger groups. In the case of larger numbers, we can still use the market. Now,
we can set either the price or the quantity traded, which will allow the market to work.
Set the price of social damage—Pigovian taxes or green taxes. For over a century, econo-
mists have promoted the idea that we can adjust market prices to fix environmental dilem-
mas. We create new economic incentives in the market by altering the relative price of
pollution or an otherwise unpriced environmental asset. The economist Alfred Pigou sug-
gested in the early twentieth century that an effective solution to pollution problems is to add
a tax on to the market price. This tax, now called a Pigovian tax or a green tax, would be set to
equal the external cost, or marginal damage, suffered by those affected by the pollution.
In principle, society can alter a person’s choices by imposing a green tax. The person con-
tinues to produce and pay the tax as long as the marginal benefits he or she receives from the
26 ECONOMIC TOOLS FOR THE ENVIRONMENT

output exceed the tax. Once the green tax exceeds the marginal benefits, he or she cuts off
production. Ideally, ifthe green tax is set such that it reflects the equilibrium level of marginal
damage, the person voluntarily selects the level of output that is the social optimum.
For Riley and Ole, a regulator could set the green tax at the equilibrium level of Ole’s
marginal damage suffered from the loss of open space. Now Riley would develop the space
up to the point at which the tax exceeded his marginal benefits, and then he would stop. This
would result in the same level of development as the Coasean solution.
Economists have debated the idea of whether a double dividend exists with green taxes. A
double dividend exists: (i) when a green tax reduces the amount of pollution emitted; and
(ii) when the revenues raised by the taxes are used to offset other distortionary taxes, such as
income or capital gains taxes. A distortionary tax is a charge on activities that society wants
to promote rather than discourage, such as working and investing. If we can use green taxes
to reduce pollution and then use the extra revenue to reduce income taxes, society has two
hits with one shot.
Unfortunately, if you compare it to the optimal benchmark, the double dividend story has
a twist. We know that environmental protection benefits society. But this protection also
raises prices and the overall cost of living, which cuts into a person’s wages. Environmental
protection is like a second labour tax added on to the already distortionary income tax. Now,
our worker has even more reason to decrease his labour supply, which is an additional cost
to society—less labour, less wealth created. The open question remains whether the environ-
mental benefits of the green tax offset the losses from the decreased labour supply. The dou-
ble dividend moral is this: correcting one mistake can make a second even worse.
In addition, green taxes have historically been set to raise small amounts of revenue, not
induce big changes in behaviour. Taxes have been set too low to induce people to increase
significantly either pollution abatement or environmental protection (Hanley et al., 2007).
This reality is in part affected by the lack of information required to successfully implement
an incentive to approach some social goal. Setting an efficient green tax requires informa-
tion on the marginal costs and benefits schedules. We also need information on the envi-
ronmental fate and transport systems and the monetary value of risks to life and limb. The
information required to find this marginal damage function is not free, as had been pre-
sumed in the original green tax models.
An important extension to the idea ofan optimal green tax was suggested by Baumol and
Oates four decades ago. They showed that if society wants to achieve a given target level of
emissions reduction, a tax on emissions could be the lowest-cost way of achieving this
reduction, compared with a command-and-control approach such as design standards (the
government tells firms how to control pollution) or a performance standard (the govern-
ment tells them how much pollution they can emit). To understand the Baumol and Oates
approach, we introduce the idea of a marginal abatement cost curve, or MAC.
Abatement costs are the costs of reducing emissions. Polluters can reduce emissions by
several alternative means:
e installing ‘end-of-pipe’ treatment plants;
e changing their production processes—for example, by using cleaner inputs or recycling
waste; and
e reducing output.
MARKETS AND THE ENVIRONMENT 27

We assume that polluters know the range of options, and that they always choose the
lowest-cost means. This may vary as emissions are reduced. A useful way of thinking about
how abatement costs vary with the level of emission reduction is shown in Figure 2.2(a).
This shows a MAC for a firm, the Jones Company, for reducing emissions by installing end-
of-pipe treatment. The graph is read right to left, since this shows falling emissions. As can
be seen, the MAC rises as the firm progressively cuts back on its emissions.
What does this mean?—That as emissions fall, the additional cost of reducing emissions
increases still further: in other words, that the marginal abatement costs are rising. This ris-
ing MAC curve is an almost-universal empirical finding. Marginal abatement costs ‘take off
at a 75 per cent cut, since end-of-pipe technology cannot cut emissions by more than this.
The area under the MAC curve at any point gives the total abatement cost (e.g. area ‘a’ shows
the total abatement cost of going from 100 per cent emissions to 75 per cent emissions).
Figure 2.2(b) shows the MAC curve for the firm defined across all emission reduction
options; this is now flatter past the 75 per cent reduction level, as the firm can choose to use
other methods, such as changing its inputs. The curve takes off at extreme levels of emission
reduction (95%), as emission reductions become costly at high levels of pollutant removal.
Another important empirical finding in the literature is that MAC curves can vary across
firms for the same pollutant. For example, for Biological Oxygen demand (BOD) discharges
into the Forth Estuary in Scotland, the actual abatement costs varied by as much as 2,500
times per kilogram of BOD removed. This may be due to firms that emit the same pollutant
having different production processes and plant; having different levels of managerial skills;
or being located in different areas (e.g. facing different transportation costs for bringing in
cleaner inputs).
In Figure 2.2(a), we show the MAC curves for Jones PLC and another firm (Bloggs) that
both emit BOD into a river. Jones has higher abatement costs than Bloggs, since it operates
a different production process. Assume for simplicity that, in the absence of spending any
money on pollution control, both firms discharge the same level of emissions, which we
show as e? in the figure, equal to 10,000 tonnes/BOD/week each. The total unregulated dis-
charge is 20,000 tonnes/week.! Figure 2.2(b) shows the aggregate MAC curve, which is just
the horizontal summation of the MAC curves of Jones and Bloggs, referred to as MAC.
Suppose that a control authority, the Environmental Protection Agency (EPA), wants to get
an overall reduction of 10,000 tonnes/week. How could it use economic incentives to
achieve this cut, and what might the outcomes be?
Suppose that the EPA sets a tax of t on every unit of emissions from Jones. This means that
if Jones emits level of emissions e’, it pays (e’-f) in taxes. What should the managers of Jones
do? Imagine that they are emitting at e°. The best they can do is to reduce emissions to e’, since
above e! the marginal benefits of cutting emissions are greater than the marginal costs
(t > MAC); whereas below e! the marginal benefits of increasing emissions (savings in abate-
ment costs, measured by MAC) are greater than the marginal costs (increased tax payments of
t per unit). Setting emissions equal to e! is the firm’s best response: it implies an equilibrium of
t = MAC.

| In practice, it is unlikely that firms would engage in zero levels of pollution control in the absence of regulation,
since some control might result from changes in operations motivated by cost saving, such as the recycling of waste
streams.
28 ECONOMIC TOOLS FOR THE ENVIRONMENT

MAC by end-of-pipe treatment

(£)
abatement
Marginal
costs

Areaa

0 75% 50% 25% 0%

<— Emission reduction

Figure 2.2a A MAC for the Jones Company, for reducing emissions by installing end-of-pipe treatment.

MAC (all options)

abatement
Marginal
costs

0 95% 75% 25% 0%

—=— Emission reduction

Figure 2.2b The MAC curve for the Jones Company defined across all emission reduction options.

At e’, the firm pays tax revenues equal to area ‘b’ and, relative to the pre-tax position, has
increased abatement spending by area ‘a’. How does the control authority know what tax
rate to set? Suppose that it knows the marginal abatement cost schedule MAC, from
Figure 2.3. The target level of aggregate emissions is shown at E*, equal to 10,000 tonnes of
BOD. At this emission level, the aggregate MAC schedule has a value of t. This is the correct
level at which to set the tax to achieve the target reduction.
But what if the EPA does not have this information? Information, or the lack of the
needed information, is a likely scenario in reality. The EPA has to guess the tax rate and
observe firms’ reactions. If it sets the tax rate too high, the firms cut emissions by too much
and the target is overshot;? if the tax rate is set too low, the opposite occurs; emission
MARKETS AND THE ENVIRONMENT Zo

MAC

20
10,000
Emissions
(tonnes/BOD/week)

(b)

MAG,

g
=

10,000 tonnes 20,000

E, aggregate emissions
(tonnes/BOD/week)

Figure 2.3 From firm's emissions to aggregate emissions.

reductions are undersupplied. The EPA iterates on to the correct tax rate. Firms dislike this
approach, however, since the future tax rate is uncertain, which makes planning for invest-
ment more difficult.
The main attraction of taxes over regulation can now be explained. Under a tax, and as
already shown, each firm’s best response is to adjust its emissions so that we get

b= MAG

2 While it might seem odd to talk about achieving too much pollution control, remember that each extra
reduction in pollution that we aim for imposes a cost on society.
30 ECONOMIC TOOLS FOR THE ENVIRONMENT

MAC,
tax

— j=)So

2,500 e* 7,500 e
Emissions, tonnes
a
Emission reduction by Jones
Emission reduction by Bloggs

Figure 2.4 Pollution taxes as the least-cost solution.

for each firm. This means that for Jones and Bloggs, the tax produces the following outcome:

t=MAC,..., Jones = MAG Bloggs *

This is the least-cost solution, which minimizes the cost of achieving the target reduction.
That is important: it implies that taxes can satisfy the efficiency criterion. Why is it the least-
cost solution? Because if the marginal abatement costs of each firm were not equal, society
could always save money by reallocating emission control responsibility away from the
higher-MAC firm and towards the lower-MAC firm. For example, in Figure 2.4, suppose
that the EPA imposed a performance standard equal to e* on each firm. This means that
Jones has marginal costs of £100/tonne at this point, and Bloggs has marginal costs of £50/
tonne. If we allowed Jones to increase emissions by one tonne and persuaded Bloggs to cut
emissions by one tonne more, total emissions would be unchanged, but we would have
saved (100 — 50) = £50. Such gains can be made whenever marginal abatement costs are not
equal. But how could Bloggs be persuaded to cut emissions by more than Jones?—By setting
the tax of t= 75. This gives the desired reduction in emissions (Jones emits 7,500 and Bloggs
2,500, so that new emissions are 10,000 tonnes), but at the lowest cost, since the tax results
in marginal abatement costs being equalized.
The least-cost property of pollution taxes (also known as the static efficiency property) is
their most important attribute, and it was first set down by Baumol and Oates in 1971 (see
their 1988 book). Pollution taxes have one other major advantage. Since each unit of
emissions now costs the firm money in tax payments, firms have more incentive to invest in
cleaner, greener technology than under regulation. This is known as the ‘dynamic efficiency’
property of taxes, and has been argued by some to be their most important feature in the
long run. What are the problems with pollution taxes as a policy solution?
MARKETS AND THE ENVIRONMENT 3]

abatement
Marginal
costs

0 et e?
Emission reduction

Figure 2.5 The impact of a water pollution tax.

e When pollutants are non-uniformly mixed, then a single tax rate is not efficient. This
happens because the tax is levied on emissions rather than on their environmental impact.
Firms that cause more damage per unit of emissions should be taxed at a higher rate than
firms that cause lower per-unit-of-emission damage. At the limit, correcting this problem
involves a unique tax rate for each firm. More pragmatically, suggestions have been made
for banded tax rates to try to control for non-uniform mixing to at least some degree.
e Taxes minimize the total abatement costs of hitting the target from society’s point of
view. But taxes can be more expensive than regulation for firms themselves. This is
because firms pay for both abatement and the pollution tax on their remaining
emissions. This tax payment could be greater than the abatement costs, and can result
in the total financial burden of taxes (areas ‘a’ +‘b’ in Figure 2.5) exceeding that under
regulation. This aspect of pollution taxes has, unsurprisingly, resulted in industry
lobbying against their wider use.
e The EPA has insufficient information with which to set the tax rate correctly.
Additionally, these taxes have to be updated based on new information; for example, as
firms’ abatement costs change. This makes the uncertainty problem more serious, since
it means that the EPA has to keep on re-guessing what the tax rate should be.

Set the quantity of social damages—tradable pollution permit systems, or cap-and-trade.

An alternative to a fixed tax is a fixed quantity of pollution that can be traded in a market.
Pollution permit markets work by assigning the property rights to pollute to firms, govern-
ments, and people (see Gayer, 2008). These rights create and add value to something that
was otherwise a free good; for example, clean air or water. This tradable permits, or cap-and-
trade, system was first introduced by Professors Thomas Crocker and J. Dales in the mid-
1960s. The idea was controversial then because many people did not think that we should
‘put a price’ on nature. Today, many environmental groups promote the idea of cap-and-
trade as the most cost-effective way to protect nature.
Tradable permits focus on the quantity side of the market equation. A regulator selects a
fixed quantity of pollution or development, and then sets the number of permits available
32 ECONOMIC TOOLS FOR THE ENVIRONMENT

for trade accordingly. These permits are then allocated to firms and people in the affected
area. People then buy and sell these permits on the open market. People who keep pollution
or development below their allotted permit level can sell or lease their surplus permits.
People who exceed their allocation must buy permits from those who either produce less or
find less-polluting technologies.
What makes a tradable permit system effective? Economists have identified the condi-
tions under which such a system is more likely to work. Permits must be well defined and
scarce, so that their value can be estimated accurately. Free trade should dominate the per-
mit market. Government intervention, bottlenecks, and transaction costs that limit the
scope of trading should be minimal. Less friction increases the odds that people who value
the permits the most will buy or keep them. Permits should be ‘bankable’, such that people
have the flexibility to save and spend permits as the market conditions fluctuate. People
should be allowed to keep any profits they earn from the trade of permits. The penalties for
violating a permit must exceed the permit price, to help make sure people play by the rules.
Again consider the case of Riley and Ole. With a tradable permit system, the regulator
selects the amount of ridge development allowable in the valley: development quotas are
selected at the socially efficient level, in which marginal benefits equal marginal cost. The
regulator allocates permits to the Centennial community based on some predetermined
rule—perhaps the number of acres or years spent in the valley. The permits are then free to
be traded. If Riley wants to develop open space by building on the ridgeline, he can buy
more development permits on the open market. In theory, the equilibrium permit price
equates the marginal costs to the marginal benefits of development. Again, the efficient
outcome is achieved.
To see how this works, consider our example of a pollution tax. The EPA faced a situation
in which two firms emitted a total of 20,000 tonnes of BOD per week, whereas the target
level of emissions was only 10,000 tonnes. Instead of imposing a tax, the EPA could have
created 10,000 emission permits and allowed firms to trade them between themselves.
Because it would be illegal to emit beyond one’s permit holding, the target emissions reduc-
tion would be reached: with 10,000 permits available, 10,000 tonnes of BOD could be legally
discharged in total.
What advantages do tradable permit systems possess? In Figure 2.6, the MAC curves for
Jones and Bloggs are shown again. Suppose that each firm is given 5,000 permits. Both must
cut emissions because their unregulated level is 10,000 tonnes, but by how much? Imagine
that, at first, neither firm is willing to trade permits. At an emission level of 5,000 tonnes,
Jones faces marginal costs of 100, and so could save this amount if it could increase its emis-
sions by one tonne. That would involve buying a permit from Bloggs, who would have to be
willing to sell. Bloggs might sell provided that the permit price was greater than its cost of
reducing emissions. The cost to Bloggs of this sale is 50 (its marginal abatement cost at this
level of emission): the minimum Bloggs would take is less than the most Jones will offer—a
deal can be done. Both Jones and Bloggs gain from this trade. If the permit was to change
hands for a price of £80, both would be better off. If all such gains from trade can be cap-
tured, we expect trading to continue until the MACs are equalized across sources in a com-
petitive market for permits. This idea of trading until we ‘equalize MACs’ is a necessary
condition for the cost-minimizing outcome: tradable permits, like taxes, can offer a least-
cost way of controlling pollution.
MARKETS AND THE ENVIRONMENT oe)

MAChones

MAC, Bloggs

MAC,
tax

2,500 5,000 7,500 e?

Emissions, tonnes

Figure 2.6 Tradable permits as the least-cost solution.

Another way of thinking about trading permits is to consider how a firm would react if
offered permits for sale at some fixed price, such as p* in Figure 2.7(a). Jones would choose
to buy e* permits at this price, which necessitates a cut in emissions from e° to e*. Why?
Because if Jones were to buy more permits, they would have spent more on permits, at the
margin, than it would cost to abate. If they were to buy less than e*, they would have spent
too little on permits given the cost of abatement. The optimal amount to purchase is the
level at which marginal costs intersects with marginal benefits, e*: Jones should purchase
permits until the price exceeds the benefits. If the permit price were to increase, the firm
would choose to buy less, and would have to spend more on emissions control (e.g. at p**).
If the price fell, they would buy more permits and spend less on pollution control. The firm’s
MAC curve is its demand curve for permits.
Where does this permit price come from? One way to think about this is as the interac-
tion of supply and demand in the permit market. Supply is determined by the number of
permits available in total, and that is determined by the EPA when it sets the maximum
desired level of emissions, E* in Figure 2.7(b). The supply curve S is vertical at this point,
since no more permits are available from the EPA, irrespective of the price. MAC, is the
aggregate marginal abatement cost curve: its shows the market demand for permits. At E*,
supply and demand are equal at price p*, and this is the market clearing permit price. If all
firms behave as Jones, in a multi-firm market with many dischargers (firms a, b, c, ...), we
end up with the situation that
MAC(a) = MAC(b) = MAC(c) = ...p*,

which is the same efficient outcome as with a tax.

In practice, permit trading can take place in two ways. First, the EPA may decide to
launch the permit market by auctioning permits. All firms bid for permits from a single
seller. Once firms have acquired their initial holding, they can trade with each other as their
circumstances change or as firms enter and leave the industry/area. In this case, the permit
price depends on the bargains that firms strike with each other.
34 ECONOMIC TOOLS FOR THE ENVIRONMENT

MAC
price
permit

Buy ——_—_—_—_—_—_———_—_—
————
permits _: Reduce emissions

Emissions, e

iV)
aS)
=
[aX
=
=—_
cD)
&

2=
p*

E* f0
Aggregate emissions, E

Figure 2.7 Prices in the permit market (a) individual (b) aggregate.

Second, the EPA could give permits away, a practice known as ‘grandfathering’. All
trades will be inter-firm, unless environmental groups buy permits and then withhold
them from the market (to reduce the maximum legal level of emissions). Firms prefer
grandfathering to auctioning, since the financial burden is less on average. As with taxes,
this financial burden has two parts: the resource costs of pollution abatement, and pay-
ments (net of receipts) for permits.
We have seen that permits can generate a least-cost, efficient means of controlling pollu-
tion. So what is the catch?

e Transaction costs imply that fewer trades take place than needed to realize all potential
abatement cost-savings. The transaction costs are the costs associated with finding
potential buyers/sellers, and with negotiating subsequent trades. Evidence from the
sulphur-trading programme in the USA suggests that these can be a high percentage of
the gains from trade.
MARKETS AND THE ENVIRONMENT 35

e If few firms participate in the permit market, they are less likely to behave
competitively. For example, a large, powerful permit seller may withhold some permits
from the market to keep their price high. This kind of behaviour, by both buyers and/or
sellers, may result in a loss of permit market efficiency, but this depends on the precise
circumstances.
e When pollutants are non-uniformly mixed, allowing permits to trade at a one-for-one
rate may result in local violations in water quality standards. Imagine that two firms are
thinking of trading. In Figure 2.8, firm A is a potential buyer from firm B. Because A is
located upstream of B, each unit of A’s emissions does more harm than each unit from
B. If A buys 100 permits from B, total emissions remain constant, but environmental
damage rises, especially in the zone immediately downstream of A. This situation arises
in water pollution control, and several solutions have been proposed. One is zonal
trading—here, a zone rule would prohibit trades between A and B, and would allow
trades between A and another firm, C. But the more trade is restricted in this way, the
lower is the cost-saving potential. Another idea is to use trading rules, which would
restrict the rate at which A and B can trade. Suppose that A’s emissions are twice as
harmful per unit as B’s in terms of average water quality. The regulator could impose an
exchange rate of 0.5/1 per trade between the two. Under this scheme, exchange rates are
calculated for all firms on the river, which is possible given current water quality models.
In reality, however, one of the largest actual Tradable Pollution Permit (TPP) systems in
use, sulphur trading in the USA, does not address the fact that SO, is a non-uniformly
mixed pollutant, and allows emission-based trades to go ahead at a one-for-one rate.
e Existing firms may use permits as a barrier to entry, to keep out new firms who want to
set up.

None of these criticisms means that all of the cost-saving potential of TPPs is lost. We
would expect some cost savings if a change were to be made from regulation to trading.
Relative to pollution taxes, TPPs possess some advantages too. Most importantly, the EPA
does not need to know the firms’ MAC curves to set the system up. All it does is decide how
many permits to issue, and what restrictions—if any—to place on trade, and then police the
system. Firms may also prefer trading to pollution taxes if permits are allocated free of
charge (grandfathered) rather than auctioned. Financial burdens are lower than with an

eFirmC
(no change)

WIV

The Estuary
SSE SESS reer OS Mal

ofc fs eFirmB
(buys 100) 0 10 (sells 100)
eee

miles

Figure 2.8 Permit trading in an estuary.

36 ECONOMIC TOOLS FOR THE ENVIRONMENT

Up until the introduction of the European Union's carbon cap-and-trade scheme, the USA had the most
extensive experience of using tradable permit markets to control pollution. The initial moves towards use
of TPPs came in the 1970s, due to conflicts between achieving national targets for clean air and allowing
economic growth in industrialized states that were in violation of these national targets. Policy initiatives
such as offsets, netting, and banking were brought together under the Emissions Trading Program in
1986. This allowed for limited permit trading in ‘emission credits’ for seven pollutants in 247 control
regions across the USA.
In 1992, amendments to the Clean Air Act paved the way for a nationwide cap-and-trade system for
sulphur dioxide emissions from power stations. The aim was to reduce total emissions to 50 per cent of
existing levels. The market began in 1995. Some 110 of the largest power stations were allocated permits
based on historical emissions, and then allowed to trade. Even though SO, is a non-uniformly mixed
pollutant, permits traded at a one-for-one rate. In 2000, 800 additional power stations were to be
brought into the scheme.
Evidence suggests that the pollution permit markets have performed well. First, total emissions fell by
more than the target level in phase 1, as firms banked permits for future use. The market in permits grew
steadily, and permits prices fell from an initial high of around $1,000/tonne to around $100/tonne. The
increasing volume of trading, which reduced transaction costs, and falling abatement costs caused this
price fall. This last factor was due to suppliers of abatement equipment cutting their prices, since firms
now had an alternative (buying permits), and by reductions in the price of low-sulphur coal due to
deregulation.
Overall, the cost savings of the sulphur trading programme have been estimated at up to 50 per cent
of what the costs would had been if regulation had been employed instead. This implies a saving to the
US economy of nearly $1 billion per year (Carlson et al., 2000). Finally, the total costs of the scheme
seem to be less than the benefits, which include the economic value of avoided damage to human
health, ecosystems, and recreational activities. Tradable permits have provided good value for the
money.

emission tax system. TPP mechanisms also do not need updating iffirms’ abatement costs
shift, since this changes the demand for permits: actual emissions cannot rise above the
maximum permitted (see also Box 2.5).
Whether tradable permit markets can flourish as a tool to manage environmental and
health dilemmas will be determined over time. Conceived in the 1960s, tested in the 1970s
and 1980s, and implemented in the 1990s, tradable pollution permits are now common-
place in debates on how to manage the environment cost-effectively. The most studied
example is the US acid rain trading programme from the 1990s, which reduced sulphur
dioxide emissions by 50 per cent at half the cost of a command-and-control approach. Such
success stories raise the costs to policy-makers who neglect how effective markets can be at
managing risk to society.
Modern climate change policy has focused around the tradable permits market that is the
market for carbon emissions, as an integral part of the cost-effective risk reduction strategy.
In 2005, the European Union created a regional market to trade carbon emissions—the EU
ETS. A cornerstone in the EU’s attempts to address climate change policy in a cost-effective
manner, the EU ETS sets up a cap-and-trade system for nearly 10,000 factories, oil refiner-
ies, electricity utilities, and more. The ETS market allows buyers the flexibility to find
MARKETS AND THE ENVIRONMENT BT

low-cost carbon emissions from around the world. Estimates suggest that a well-function-
ing market would cut the costs of reaching the Kyoto targets by between 50-80 per cent (see
Chapter 9). The main criticism is that the EU ETS caps were set in too loose a fashion in
Phases I and II to induce a significant reduction in carbon emissions. The counter-argument
is that Phases I and II helped the EU define how the system would work—better emissions
data, better monitoring, and creating a positive price on carbon emissions. Phase III begins
in 2013, with the stated goal of reducing carbon emissions in the EU by 21 per cent from
2005 levels (see Parker, 2010).

Summary
Over millennia, humans increased their life expectancy by a few years. About 200 years ago,
something changed, and since then Western culture has witnessed a thirty-year increase in
our longevity. Economists do not see this as a coincidence. In 1776, Adam Smith published
his classic work The Wealth of Nations, which explained the power of the market to create
wealth through trade. Economists argue that understanding how markets collect, codify,
and disseminate diffuse information has helped to create the social order to improve the
quality and length of life in our modern world for many people.
In this chapter, we have examined how markets can work against and for the environ-
ment and natural resources. Markets are a process of discovery. Markets allow us to create
wealth, which in turn allows us to create more health. And even when one market fails, a
new market can be constructed to manage the environment. Markets do not substitute for
good environmental policy; rather, they can be a good tool to promote more protection for
less wealth. Markets can make good environmental policy better by allowing for the flexibil-
ity to protect valuable resources cost-effectively. Remember: markets work for us, and not
the other way around. Identifying ifand how markets can be created or corrected is a major
task for all of us who are interested in providing more environmental protection at less cost.

Tutorial Questions
2.1 Define Pareto efficiency.
2.2 Explain the four conditions that must hold for the existence of a well-defined property
rights system, and address why all four matter.
2.3 Why do economists promote the idea of creating markets for environmental
protection?
2.4 What is a public good and why are public goods considered a market failure?
2.5 Explain the idea behind the phrase the ‘tragedy of the commons’.
2.6 Explain why moral hazard and adverse selection are market failures?
2.7 How can government help solve the problems of hidden information?
2.8 Explain how Pigovian taxes work to control pollution.
2.9 How do tradable pollution permits work in theory and in practice?
38 ECONOMIC TOOLS FOR THE ENVIRONMENT

References and Additional Readings

Arrow, K. (1969). “The organization of economic Hanley, N., Shogren, J., and White, B. (2007).
activity: issues pertinent to the choice of market Environmental Economics in Theory and Practice,
versus nonmarket allocation, The Analysis and 2nd edn. (London: Palgrave Macmillan).
Evaluation of Public Expenditures: The PPB System Hardin, G. (1968). ‘The tragedy of the commons,
(Washington, DC: Joint Economic Committee, Science 162: 1243-8.
91st Congress): 47-64.
Jurado, J., and Southgate, D. (1999). “Dealing with
Baumol, W., and Oates, W. (1988). The Theory of air-pollution in Latin America: the case of
Environmental Policy, 2nd edn. (Cambridge: Quito, Ecuador, Environment and Development
Cambridge University Press). Economics 4(3): 375-89.
Carlson, C., Burtraw, D., Cropper, M., and Palmer, K. Krugman, P., and Wells, R. (2008). Microeconomics
(2000). ‘Sulfur dioxide control by electric utilities: (New York: Worth).
What are the gains from trade?’ Journal of
Ledyard, J. (2008). ‘Market failure? in S. Durlauf and
Political Economy 108: 1292-326.
L. Blume (eds.), The New Palgrave Dictionary of
Coase, R. (1960). “The problem ofsocial cost, Journal Economics, 2nd edn. (New York: W.W. Norton).
of Law and Economics 3: 1-44. [Link]
Cornes, R., and Sandler, T. (1996). The Theory of dictionary.
Externalities, Public Goods, and Club Goods, Ostrom, E. (1990). Governing the Commons
2nd edn. (Cambridge: Cambridge University (Cambridge: Cambridge University Press).
Press).
Parker, L. (2010). Climate Change and the EU
Crocker, T. (1966). “The structure of atmospheric Emissions Trading Scheme (ETS): Looking to
pollution control systems, in H. Wolozing (ed.), 2020 (Washington, DC: Congressional Research
‘The Economics of Air Pollution (New York: W.W. Service). [Link]
gov.
Norton): 61-86.
Platteau, J-P. (2008). ‘Common property resources,
Dales, J. (1968). Pollution, Property, and Prices in S. Durlauf and L. Blume (eds.), The New
(Toronto: University of Toronto Press). Palgrave Dictionary of Economics, 2nd edn.
Dallimer, M., Acs, S., Hanley, N., Wilson, P., (New York: W.W. Norton). [Link]
Gaston, K., and Armsworth, P. (2009). ‘What [Link]/dictionary.
explains property-level variation in avian Samuelson, P. (1954). “The pure theory of public
diversity? An inter-disciplinary approach, Journal expenditure, Review of Economics and Statistics
of Applied Ecology 46: 647-56, 36: 387-9.
Fankhauser, S., Tol, R.S.J., and Pearce, D.W. Sandmo, A. (2008). ‘Pigouvian taxes, in S. Durlauf
(1998). ‘Extensions and alternatives to climate and L. Blume (eds.), The New Palgrave Dictionary
change impact valuation: on the critique of of Economics, 2nd edn. (New York: W.W. Norton).
IPCC Working Group III’s impact estimates), [Link]
Environment and Development Economics 3: dictionary.
59-81.
Schoeb, R. (2006). “The double dividend hypothesis
Finnoff, D., and Tschirhart, J. (2008). ‘Linking of environmental taxes: a survey, in H. Folmer
dynamic ecological and economic general and T. Tietenberg (eds.), The International
equilibrium models, Resource and Energy Yearbook of Environmental and Resources
Economics 30(2): 91-114. Economics 2005/2006 (Cheltenham: Edward
Gayer, T. (2008). ‘Pollution permits) in S. Durlauf Elgar): 223-79.
and L. Blume (eds.), The New Palgrave Dictionary Stavins, R. (2003). ‘Experience with market-
of Economics, 2nd edn. (New York: W.W. Norton). based environmental policy instruments? in
[Link] K. Maler and J. Vincent (eds.), The Handbook of
dictionary. Environmental Economics (Amsterdam: North-
Gordon, S. (1954). “The economic theory ofa Holland/Elsevier Science).
common property resource: the fishery, Journal Weitzman, M. (1974). ‘Prices vs. quantities’
of Political Economy 62: 124-42. Quarterly Journal of Economics 41: 477-91.
Valuing the Environment:
Concepts

In this chapter, we are going to consider:

e What ‘economic value’ means.

e In what ways the environment has economic value.
e Howeconomic values might be measured.
e How measures of the economic value of the environment can help society make better
decisions, particularly through cost-benefit analysis.

In the next chapter, empirical methods for measuring environmental values will be
explained.

3.1 What Does Economic Value Mean?

Economics is often described as the study of how to allocate limited resources in the face of
unlimited wants. The fact that resources are scarce means that using up resources in one way
prevents us from using them in another way. The cost of these forgone uses is called an
opportunity cost, defined as the best alternative use forgone. This concept is very relevant to
the environment: using a river for waste disposal has an opportunity cost of lost recreation
and wildlife benefits. Designating a mountain area as a national park means that we forgo
mineral extraction opportunities. In other words, deciding to use the environment in one
way entails a sacrifice, namely the benefits we could have gained by using it in another other
way. Environmental policy may also imply non-environmental sacrifices: for example, the
decision to impose an energy tax to reduce carbon dioxide emissions may mean that poor
households forgo significant consumption possibilities (Sterner, 2011). However, if society
decides to go ahead with the tax, then a judgement must have been made, either explicitly or
implicitly, that the benefits of reduced emissions are ‘worth it’, in the sense that they justify
the costs.
The idea that the value of something is dependent on what we are willing to give up to
have it is a key economic principle. But how should we express what is being given up? In
40 ECONOMIC TOOLS FOR THE ENVIRONMENT

the preceding example, poor households could be giving up a significant proportion of their
consumption goods. In the case of preserving a mountain area from mineral extraction, the
profits from extraction could also have funded a wide variety of forms of consumption. One
approach, therefore, is to take the most general measure of what is being sacrificed, namely
income. The simplest way to see this is by considering an example. Suppose that poor house-
holds were asked whether they would support a local tax on petrol and diesel fuel used by
cars and lorries, with the objective of improving air quality in their city. One way of putting
this question would be to ask if their willingness to pay (to give up income) for the air quality
improvement was higher than the cost to them of the tax.
An important point to note here is that a change in environmental quality is being valued,
not the total environment. Economic value really only has any meaning when it is defined
over a change; that is, when it is measured with regard to more or less of a good or service
being provided. The change could be ‘marginal; in the sense of a relatively small change in
the amount or level of environmental service (e.g. a 5% reduction in current ambient par-
ticulate levels in a city), or ‘non-marginal’ such as the total loss of a forest or the draining of
a wetland. Willingness to pay (WTP)—or, more exactly, maximum willingness to pay—
would, in this case, measure the benefits to people of a beneficial change in environmental
quality. For instance, for a prospective increase in air quality, the WTP is the most income
that an individual would give up to have the improvement in air quality. The logic here
revolves around rationality: no one would be willing to give up more for a change than it is
worth to them, in whatever way they were to derive value (utility) from this change. An
objection that could be made is that if WTP is used as a measure of value to an individual,
then this measure depends not just on their preferences (how much they dislike air pollu-
tion, relative to their likes or dislikes of other things) but also on their income. A rich house-
hold, with the same preferences, could clearly afford to pay more than a poor household.
Economic values based on WTP would therefore always be biased in favour of the rich.
Economists usually answer ‘Yes, that’s right. But willingness to pay is a pretty useless concept
unless backed up by ability to pay. At the level of the economy as a whole, we clearly cannot
give up more than we actually have or expect to earn’ Economic value, as measured by WTP,
is thus a function of the existing income distribution, and may change as this distribution
changes. If preferences differ, then people with similar amounts of income will be willing to
pay different amounts for the same change in environmental quality. For example, if Joe is
more concerned about urban air pollution problems because his kids suffer from asthma,
but Josephine has no kids, then if their incomes are equal, Joe’s WTP for a given improve-
ment in air quality may well be higher than Josephine’.
To recap, given that resources are scarce, using them in one way implies an opportunity
cost. The value of a particular resource use can be measured in terms of the sacrifice that
people are willing to make to have it. At the most general level, this sacrifice is in terms of
income, so therefore WTP makes sense as a measure of economic value to the individual.
However, we have to accept that this measure is sensitive to changes in the distribution of
income.
An alternative measure of economic value exists, based on the same principles of scarcity.
This involves asking what compensation an individual would accept to give something up.
This is a very familiar idea in everyday life. For example, the value you place on your favour-
ite guitar is equal to the minimum you would accept to go without it (to sell it), which we can
VALUING THE ENVIRONMENT: CONCEPTS AI
call your minimum ‘willingness to accept compensation’ (WTA). For workers, the value of
their working time is measurable by the minimum hourly wage they would accept to work.
This compensation-for-loss concept of value can also be extended to environmental
resources. The value you place on your garden could be estimated by the minimum com-
pensation you would accept to sell it to your neighbour. Similarly, the value of peace and
quiet could be thought of as the minimum compensation you would demand to agree to a
new airport runway being constructed near to your house.
The question of whether we use WTP or WTA as the basis for valuation matters.
According to a famous paper by Robert Willig (1976), the differences between WTP and
WTA for most goods should be small, and should depend on the relationship between
income and demand, and on how much ofa person’s income is spent on the good. However,
experimental findings in economics have revealed quite large differences between WTP and
WTA that seem to violate this finding. Competing explanations have emerged for this dis-
parity. The first, based on work by Michael Hanemann (1991), shows that the difference is
explicable in terms of how close a substitute exists for a commodity. For example, if we were
to compare WTP with WTA for tickets for two ice hockey games, one of which was also
being televised live, we might expect the difference between WTP and WTA to be smaller
for the game that people could watch live on TV instead of only being able to watch it live at
the stadium. If there are no close substitutes, then we could expect quite large differences
between WTP and WTA, whereas if close substitutes exist, then WTP and WTA should not
be that different. The second explanation is rooted in behavioural psychology, in the con-
cept of loss aversion. This concept suggests that people systematically value what they
already have more highly than that which they could acquire, so that losses are always val-
ued more highly than equivalent gains. Experimental economics has also shown that differ-
ences between WTP and WTA can be much reduced as people gain experience in trading in
a good, although such experience can be hard to gain for many environmental goods. This
relates to a third possible explanation for the disparity, namely that people are unsure about
their preferences (Loomes et al., 2003).
It seems likely, therefore, that important choices will often have to be made by econo-
mists between measuring WTP or WTA. Such choices can be based around the concept of
property rights. If people have a moral, legal, or assumed right to a good, we would ask them
about their WTA for a reduction in that good, not their WTP to prevent this reduction. If,
instead, people have no right to an increase in a good, then it is reasonable to ask for their
WTP for an increase, rather than their WTA to forgo such an increase.
To summarize, if an increase in an environmental good is being valued, we can try to
measure either what people’s maximum WTP is to have this increase, or what their mini-
mum WTA is to forgo this increase. If areduction in the same good is being valued, we can
try to uncover either their maximum WTP to prevent such a reduction, or their minimum
WTA to tolerate it. Either approach allows a monetary value to be placed on an environ-
mental gain or loss, which is an estimate of the underlying utility gain or loss for an
individual.
But how might WTP vary for that person for different amounts of an environmental
good being offered? Imagine an experiment in which an individual is asked about his or her
maximum WTP for a succession ofincreases in an environmental good. For one such per-
son, Gavin, we might get results such as are shown in Figure 3.1(a). As the quantity of the
42 ECONOMIC TOOLS FOR THE ENVIRONMENT

(£)
value
total

Total
WTP

50 100
No. of pairs of osprey protected

Figure 3.1(a) WTP for wildlife protection.

good rises (as an example, we might consider pairs of ospreys protected in Scotland),
Gavin’s total WTP increases: for example, he is willing to pay more to protect 100 pairs than
for fifty pairs, since, as a birdwatcher, his utility is higher for 100 pairs than for fifty. Note
that as the number of pairs ‘offered’ continues to increase, his total WTP (the total value to
him of ospreys) increases at a decreasing rate. Transforming Figure 3.1(a) into a marginal
WTP curve—by measuring the increase in total WTP as the number of pairs, Q, rises—we
get Figure 3.1(b), which shows marginal WTP decreasing but always positive! (no satiation
is setting in, so that utility always goes up as consumption rises). Marginal WTP declines as
Q rises due to diminishing marginal utility. Figure 3.1(c) shows Gavin’s marginal WTP
curve, which we now term a marginal value curve, MV°, since it indeed shows the value at
the margin to him of increasing numbers ofospreys. His friend Kitty is also a birdwatcher,
who likes ospreys even more: her marginal value curve, MV‘, thus lies above Gavin’s at
every point (we assume, for simplicity, that Gavin and Kitty have equal incomes). Drawing
MV curves as smooth and continuously decreasing is a theoretical assumption that makes
the analysis easier, but that may not be borne out in reality. However, the assumption of
declining marginal utility does seem to be well supported by a large amount of evidence
from economic research.
Figure 3.2 shows the derivation of WTP and WTA for an individual who is offered an
increase in environmental quality from Q, to Q;. This diagram shows utility as being a func-
tion of two things: environmental quality, Q, and income, Y. The curves Uy and U, are indif-
ference curves. These have the property that along a given indifference curve, utility is
constant. Indifference curves are shaped the way they are drawn as we assume diminishing

' Marginal WTP for any value of Q, say Q*, is equal to the slope of the total WTP curve at Q*. It is the partial
derivative of total WTP with respect to Q, evaluated at Q*.
VALUING THE ENVIRONMENT: CONCEPTS 43

(£)
WTP/value
Marginal

50 100
No. of pairs of osprey protected

Figure 3.1(b) Marginal WTP for wildlife protection.

marginal utility, as in Figures 3.1: as we move along a particular indifference curve, the
additional amount of one good we are willing to give up to have more and more of another
good is falling, since the extra utility from having more of something declines as we con-
sume more and more of that thing. The further an indifference curve is away from the ori-
gin, the higher is the level of utility, thus U, is greater than Up. We start at point a, with an
income of y and an environmental quality of Q,. Now suppose that the environmental

Mvk
MVS/MV«K
(£)

50 100
No. of pairs of osprey protected

Figure 3.1(c) Marginal WTP for two people.

44 ECONOMIC TOOLS FOR THE ENVIRONMENT

$ (Income,y)

Qo Q, Environmental
E quality (Q)

Figure 3.2 Indifference curves and the value of an increase in environmental quality.
Note: Environmental quality increases from Q, to Q,. This increases consumer utility from U, to U, along their fixed budget
line, moving the consumer from point a to point b. This also implies that consumers have a maximum willingness to pay
of WTP. Alternatively, they would be willing to accept compensation of WTA to forgo an improvement in environmental
benefits. Note that WTA is greater than WTP.

quality increases to Q,. With the same income, the individual moves to point b, on a higher
indifference curve. They are thus better off. What is their maximum WTP for this increase
in environmental quality? This is the most income they could give up from point a and still
have utility equal to Uj. We can see that this amount is the vertical distance labelled WTP in
the figure; that is, the distance (bc). This diagram can also be used to work out the minimum
compensation that this individual would have to be offered to forgo the improvement in
environmental quality. Starting at point b, if income rises by the amount shown as WTA,
this keeps the individual at utility level U,, even when the environmental quality stays at Qo.
Thus the difference (da) is equal to WTA. Notice that, given the way in which the diagram
is drawn, WTP is less than WTA for the same change in Q.

3.2 In What Sense Does the Environment Have

Economic Value?

In Chapter 1, we saw how environmental and economic systems are interlinked. Figure 1.1
shows that the environment provides four services to the economy: (i) as a source of energy
and material resources (inputs) to production; (ii) as a waste sink; (iii) as a direct source of
amenity; and (iv) as the provider of other local, regional, and global support services, such
as nutrient cycling and climate regulation. Services (i), (ii), and (iv) can be grouped together
VALUING THE ENVIRONMENT: CONCEPTS 45

Outputs,
profits

Raw materials

Production

Waste
assimilation
services

Figure 3.3 Indirect environmental values.

as they both provide inputs to the production process, of raw materials, energy, nutrients,
and waste disposal services. Climate change may also impact on production; for example, in
agriculture. Figure 3.3 shows these inputs, combined with other inputs such as labour and
capital, producing goods and services for sale in markets. For example, inputs of bauxite,
energy, and waste assimilation capabilities allow the production of aluminium. One mon-
etary value of each unit of output is its price. The value of a change in the level of environ-
mental and resource inputs to the production of any good might thus be approximated by
the value of the change in profits due to this change in environmental service flows or inputs.
The values of the environment in roles (i) or (ii) or (iv) above can thus be partly determined
as the change in the value of profits for a change in the value of environmental inputs: in
other words, if bauxite inputs are reduced by one tonne, what is the value of the associated
profit decrease? For waste assimilation services, essentially the same question could again be
asked: What is the loss in profits associated with a unit reduction in emissions? This mar-
ginal productivity approach for determining environmental values is essentially no differ-
ent from the approach used to determine the value of other inputs to production, such as
labour and capital. Let us call these types of environmental values indirect benefits, since the
environment is valued indirectly through its role in the production process.
For amenity values, we need to consider more direct impacts of the environment on util-
ity. We can speak of environmental ‘goods’ as those environmental inputs of which the
individual prefers more to less (e.g. landscape quality, good air quality). Environmental
‘bads’ would then be those environmental inputs that decrease utility as they increase: for
example, noise or water pollution. Clearly, some environmental goods and bads are mirror
images of each other; for example, river water quality (a good) and river water pollution
(a bad). The economic value of any environmental good can be thought of as the increase in
utility if that environmental input is increased by a given amount; or the reduction in utility
if the quantity and/or quality of that good is decreased. Similarly, for an environmental bad,
46 ECONOMIC TOOLS FOR THE ENVIRONMENT

we are interested in the amount by which utility increases if the environmental bad is
reduced. Ideally, we would seek to measure the marginal utility of changes in environmental
goods (and bads), this being the change in utility for a one-unit increase in the good
expressed in terms of WTP or WTA. These types of environmental values could be called
direct benefits, since their effect on utility is direct, rather than indirectly through their role
in the production of consumption goods and services.
An important point to note here is that the environment can thus have economic value,
in terms of both indirect and direct values, even if it has no market value or price. For exam-
ple, many of the environmental goods, such as landscape quality, clean air, and wildlife, may
have a zero market price and yet will most certainly have an economic value, provided that
at least one person derives positive utility from them. Therefore, economic values and mar-
ket prices are not generally the same thing.
There is no reason to suppose that a given environmental good is equally valuable to eve-
ryone (economists refer to this as preference heterogeneity). It is to be expected that the
marginal utility of the good (that is, the change in utility when the level of the good itself
changes by a small amount) will vary across individuals, whilst some individuals may derive
no utility at all from some environmental resources. For example, if Joe is completely unin-
terested in birdwatching, then an increase in his local population of curlews may well have
zero value to him. Ifhis neighbour Jane is a keen ornithologist, then her marginal utility for
this same increase may be very high. This means that these two people would have different
WTP amounts for a given change in bird populations.
As shown in Figure 1.1, the economic process benefits from the many support services
provided by the natural environment, such as global climate regulation, the maintenance of
a global atmospheric chemistry suitable for sustaining life, the stratospheric ozone layer, and
local nutrient and hydrological cycles. It is possible to think about the value of preventing
changes in these services. For example, the value of preventing further changes in the global
climate through enhanced warming could be measured by looking at the costs (and benefits)
that would result from a given change in greenhouse gas emissions. Chapter 9 considers this
in detail. As another example, reductions in the stratospheric ozone layer could be valued by
estimating the economic costs ofan increased incidence of skin cancer.
A recent way of thinking about these links between the environment and the economy is
the idea of ecosystem services. This conceptualization of the value of conserving or enhancing
ecosystems is most associated with the Millennium Ecosystem Assessment.’ More recently,
the National Ecosystem Assessment (NEA) exercise in the United Kingdom? has shown the
links between the state ofdifferent ecosystems (in the UK) and the economic values that flow
from the services generated by these ecosystems (see the UK NEA). In this world view, eco-
systems are thought of as capital assets (Barbier, 2009) generating a flow ofvaluable services
(economic goods). The conservation of ecosystems can then be described as investment,
since it enables the flow of ecosystem services and their associated goods to be protected into
the future. Ecosystems depreciate when their ability to supply people with useful services is
reduced; for example, ifawetland is drained, or a coastal mangrove forest converted into a
shrimp farm.

> See [Link]

* See [Link]
VALUING THE ENVIRONMENT: CONCEPTS 47

Following from the Millennium Ecosystem Assessment, ecosystem services are often
categorized into four groups:

e Provisioning services, such as the production of food supplies

e Regulating services, such as the maintenance of a good water quality
Support services, such as the provision of habitat for mammals and birds
Cultural services, such as the values associated with landscapes

Table 3.1 shows an illustrative categorization of some ecosystem service flows for one
ecosystem (moorlands and heaths) in the United Kingdom, from the UK’s National
Ecosystem Assessment. Actual or predicted changes in any of these service flows can then
be expressed in terms of economic costs or benefits by multiplying the actual or predicted
change in ecosystem service by the marginal economic value (e.g. from market prices, or
from some other source). For instance, if degradation of moorland leads to a net loss of
carbon, these tonnes of lost carbon could be valued using the CO, permit price from the
European Emissions Trading Scheme, since this is one measure of the value of each tonne
of carbon sequestered. If moorland loss also means a decline in water quality downstream

Table 3.1 Some of the ecosystem service flows from moorlands and heaths
Provisioning services Food provision—livestock and crops:
- Livestock products from sheep and some beef cattle
Food provision—deer and game birds:
- Wild harvest products, including venison and grouse meat
Fibre from sheep wool
Traditional lifestyle products, including honey and whisky
Peat extraction for fuel and horticultural use
Freshwater provision for domestic and industrial use
Alternative energy provision:
- Opportunity for wind energy schemes
- Generation of water flows for hydro-energy in freshwater habitats
Regulating services Climate regulation:
- Carbon storage; maintenance of plant and soil C stores
- Carbon sequestration potential
Natural hazard regulation:
- Potential for flood risk mitigation
* Opportunities for wildfire risk mitigation
Pollution mitigation:
- Interception and retention of airborne pollutants by plants and soil
- Regulation of particulate matter and pH buffering
* Dilution by water from uplands of pollutants in downstream locations
Disease regulation:
- Disease transmission through ticks
- Disease regulation of waterborne bacteria (e.g. Cryptosporidia)
pe —————————————————————————
—————e
Source: Van der Wal et al. (2011).
48 ECONOMIC TOOLS FOR THE ENVIRONMENT

Use values: birdwatching, duck hunting

Direct benefits

Non-use/existence values of the wetland

Value of the wetlands as nursery for ,

fish/shellfish
Indirect benefits

Pollution control function value

Economic
value

Figure 3.4 The economic value of a wetland.

in the catchment, then this could be valued by looking at the additional costs to water com-
panies of cleaning up the water prior to supplying it to customers. If moorlands are replaced
by forests, then lost sheep production could be valued using the market price of lambs.
This idea of ‘valuing ecosystem services’ can be related to a further categorization system
of environmental benefits that has become popular in the literature (Pearce and Turner,
1989). Take, as an example, the preservation ofawetland that is important to birds, but that
also functions as a nursery for fish/shellfish and as a natural pollution control plant. How
might the total economic value ofthis wetland be described (Figure 3.4)?
Consider first what we have called direct benefits; that is, direct sources ofutility. Some
of those who benefit from the wetland in this way may participate in activities that make
the wetland valuable to them, such as birdwatching or duck hunting. Such benefits are
often known as use values, since they require actual participation to enjoy them. Use
values may be consumptive (hunting) or non-consumptive (birdwatching). However,
people other than those who actually visit the wetland may derive benefits, in terms of
the utility they get from just knowing that the wetland is preserved. These types of benefit
have become known as non-use or existence values. They may be motivated by selfish
reasons, or by altruism, either for other members ofthe current generation, or for future
generations. Existence values may be particularly high for unique, irreplaceable natural
assets, such as the Grand Canyon in the United States or Kakadu National Park in
Australia.
The sum of use and existence values gives the total direct benefits of preserving the wet-
land. ‘The wetland’s role as a nursery for fish and shellfish could be evaluated by estimating
biological models of the contribution that the wetland makes to fish/shellfish populations,
and then by looking at the economic (commercial) value of these species. Changes in these
VALUING THE ENVIRONMENT: CONCEPTS 49

Table 3.2 Values provided by tropical coastal and marine ecosystems

Direct values Indirect values Existence and bequest values

Fishing: Nutrient retention and cycling: Cultural heritage:
- Aquaculture * Flood control * Resources for future generations
* Transport * Storm protection - Existence of charismatic species
* Wild resources * Habitat for species + Existence ofwild places
- Water supply + Shoreline stabilization
* Recreation
* Genetic material

Source: Adapted from Heal et al. (2005).

economic values, in terms of gains/losses in consumers’ surplus‘ and producers’ profits from
some change in the wetland, could be calculated. Finally, the wetland’s pollution control
function could be valued either by using the value of avoided pollution damages (say, from
sedimentation of coral reef fisheries, or from nutrient enrichment), or the pollution control
costs that would have to be incurred to replace the role currently being played by the wetland.
The sum of avoided pollution and/or pollution control costs, plus the value of commercial
fisheries, would give the indirect benefits of preserving the wetland. Adding the wetland’s
direct and indirect benefits gives its total economic value. Table 3.2 shows an illustration of a
similar breakdown of benefits for tropical coastal ecosystems, such as a mangrove wetland
(note that this employs a somewhat different categorization of benefits than we have used).

3.3 Why Place Economic Values on the Environment?

Economic estimates of the value of changes in environmental and resource quality may be
useful in a number of contexts. The main use that we review here is cost-benefit analysis
(CBA; also called benefit-cost analysis in the USA), although only a brief explanation of this
method is provided (for more details, see Hanley and Barbier, 2009). Other uses of environ-
mental values are noted at the end of this section.

3.3.1 Cost-benefit analysis

Economists have long been intrigued by the problem of how to decide whether one outcome
is better than another from society’s point of view (see Box 3.1). Ideally, we would like to find
decision-making rules that give consistent outcomes; outcomes that are the same when
applied in the same circumstances. We would also like to find a method that is democratic,
in some sense, and practical; and that can be shown to be consistent with economic theory.
Welfare economics developed out of the search for such a method. Cost-benefit analysis
developed from welfare economics as a practical application of a decision-making rule that

4 As Chapter 2 explains, the consumers’ surplus is the difference between the most you are willing to pay for
something and the price you actually pay. In a market-traded good, it is the area underneath the demand curve but
above the equilibrium price.
50
an
ECONO! ENVIRONMENT
rr

One of the earliest uses of cost-benefit analysis in public policy appraisal was in the assessment of water
resource projects such as new dams orflood control investments in the USA. The 1936 Flood Control
Act stated that the federal government should undertake public investments in flood alleviation if ‘the
benefits to whomsoever they may accrue are in excess of the estimated costs’. As Banzhaf (2009)
explains, this eventually resulted in fierce debate amongst economists in the USA as to the proper role
of CBA in the policy appraisal process. Essentially, the argument was between those who viewed the
philosophical and theoretical structures of CBA to be robust enough for the outcome of a CBA to be
viewed as actually determining whether a particular project should go ahead (this group was associated
particularly with a think tank called Resources for the Future, which is still in existence today) and those
who were sufficiently uneasy about the principle of making interpersonal utility comparisons in dollar
terms to conclude that CBA should only inform the decision-making process. This latter group, most
identified with Harvard University's Water Program, felt that political or ‘expert’ judgement should
always be decisive, and that the role of the CBA analyst was no more than to make clear the trade-offs
involved in deciding whether or not to proceed with a project. Moreover, the Harvard team viewed CBA
as focusing too closely on a single objective—economic efficiency—in contrast to the multiple objectives
of public policy.
These arguments over the correct role of CBA within the public policy process, and of the advantages
and disadvantages of CBA relative to other project and policy appraisal methods, have continued ever
since. However, CBA emerges as a remarkably robust institution, perhaps because of its apparent
simplicity of approach: add up the benefits to society of aparticular project and compare these with the
costs. This still makes sense to a lot of economists and policy analysts as a way of thinking about decisions.

could be used to decide between different policy options or projects in terms of their net
contribution to social well-being. CBA consists of identifying the impacts of a project or
policy, valuing these impacts in terms of their effects on social well-being, and then compar-
ing the good effects (benefits) with the bad effects (costs). Costs and benefits are expressed in
monetary terms to allow comparison. The links with welfare economics come in terms of
how benefits and costs are measured (e.g. using the principles of WTP and opportunity
costs), and with the basis on which the difference between benefits and costs can be viewed as
a proxy for the underlying change in net social welfare. This basis is often referred to as the
Kaldor—Hicks compensation test. This asks: Could the gainers (those who benefit from a pro-
ject) compensate the losers, and still be better off? Acceptance of this principle as the basis for
evaluating contributions to social well-being in turn requires us to accept that: (i) all relevant
benefits and costs can be expressed in the same units; and (ii) benefits and costs can be com-
pared with each other, so that any cost (loss) can always be compensated by some offsetting
benefit (gain). Clearly, not all would agree with these statements (see, e.g., Aldred, 2006).

[Link] The stages of aCBA

(1) Project/policy definition

This involves setting out exactly what is being analysed; whose welfare is being considered;
and over what time period. In terms of ‘whose welfare’, the usual answer is that it is national
well-being that is considered, although defining the ‘relevant population’ is often a difficult
issue. For instance, if a new dam in Indonesia would threaten an internationally rare habi-
tat, should the costs to foreign conservationists be counted?
VALUING THE ENVIRONMENT: CONCEPTS 51

(2) Identify physical impacts of the policy/project

Any project/policy has implications for resource allocation: for example, labour used to
build access roads to a new hydroelectric dam; additional electricity production; land used
up in the creation of the reservoir; or less pollution being generated from a coal-fired power
station that can now be closed early. The next stage of a CBA is to identify these outcomes in
physical magnitudes. Frequently, these changes in resource allocation will not be known
with certainty—For example, how many tonnes of pollution will be displaced? Once physi-
cal impacts have been identified and quantified, it is then necessary to ask which of them are
relevant to the CBA. Essentially, anything that impacts on the quantity or quality of resources,
or on their price, may be said to be relevant, if these impacts can be traced back to a link to
the well-being of the relevant population. Since we specify relevant impacts in terms of util-
ity impacts, it is not necessary to restrict attention to market-valued impacts, since non-
market value changes (such as an improvement in air quality) are relevant if they affect
people's utility.

(3) Valuing impacts

One important feature of CBA is that all relevant effects are expressed in monetary values,
so that they can then be aggregated. The general principle of monetary valuation in CBA is
to value impacts in terms of their marginal social cost or marginal social benefit. Here,
‘social’ means ‘evaluated with regard to the economy as a whole’. But where are these mar-
ginal social benefits and costs derived from? Under certain conditions, this information is
contained in market prices, as Chapter 2 explains. Market prices contain information on
both the value to consumers ofa particular product (say, electricity) being supplied, and the
costs to producers of supplying it. The market wage rate, similarly, shows both the value of
labour to employers and the value of leisure to workers. Assuming that the impacts of the
project are not large enough to actually change these prices, then market prices are a good
first approximation to the values of benefits and costs. But markets often ‘fail’, as Chapter 2
shows. Moreover, for some environmental goods such as biodiversity and river water qual-
ity, no market at all exists from which a price can be observed. In this case, methods described
in Chapter 4, and based on the principles of valuation outlined earlier in this chapter, will
need to be employed.

(4) Discounting of cost and benefit flows

Once all relevant cost and benefit flows that can be expressed in monetary amounts have
been so expressed, it is necessary to convert them all into present value (PV) terms. This
necessity arises out of the time value of money, or time preference. To take a simple exam-
ple, suppose that an individual is asked to choose between receiving £100 today and receiv-
ing that same £100 in one year’s time. The more immediate sum might be preferred due to
impatience (I want to spend the money right now). Alternatively, I may not want to spend
the money for a year, but if I have it now, I can invest it in a bank at an interest rate of, say,
10 per cent and have £100 x (1 + i) = £110 in one year’s time, where j is the rate of interest.
Benefits are more highly valued the sooner they are received. Similarly, a sum of money to
be paid out, or any kind of cost, seems less onerous the further away in time we have to bear
it. A bill of £1 million to repackage hazardous wastes seems preferable if paid in 100 years
time rather than in 10 years time. This is nothing to do with inflation, but more to do with
the expectation that we might expect to be better off in the future. Box 3.2 gives an example
52 ONOMIC TOOLS FOR THE ENVIRONM

Consider a project to support the construction of anew renewable energy source. A plan is proposed to
build a small-scale wind farm ina scenic area. The initial construction costs are estimated to be £750,000.
Following start-up, annual maintenance costs of £5,000 are expected throughout the 15-year lifespan of
the plant. At the end ofthis 15-year period, the wind farm will need to be dismantled and the site
restored, at an expected cost of £35,000. Every year after the initial construction year, the site will
produce electricity with a market value of £150,000, which for now we take to be a constant flow in real
terms (we will ignore the effects of inflation here). Objectors have protested about the visual impact of
the windmills, and so the government has commissioned a contingent valuation study of local residents.
The results suggest that the mean annual compensation demanded by locals is £25 per household: 2,000
households are thought to be affected (this mean is calculated across both those against the project and
those in favour).
It is easy to set up a basic CBA of this project. The initial (‘year zero’) construction costs are not
discounted, as they occur at the start of the project. Maintenance costs are then discounted each year
using the relevant discount factor, for a discount rate of 6 per cent. Annual environmental costs of
(£25x 2,000) are also discounted over the 15 years of the project; and are assumed to stop when the site
is restored at a cost of £35,000 in year 15. This year-15 cost also needs to be discounted. Annual benefits
of £150,000 get discounted each year; you can see how the present value of this fixed amount falls each
year as we move forward in time. The following table shows all workings:

Year Discount factor at6% Benefits Present value Costs Present value
discount rate, (1.06)* (£) of benefits (£) (£) of costs (£)
0 1 0 750,000 750,000
| 0.9433 50,000 141,495 55,000 51,881
2 0.8899 50,000 133,485 55,000 48,944
3 0.8396 150,000 125,940 55,000 46,178
4 0.7921 150,000 118,815 55,000 43,565
5 0.7472 50,000 112,080 55,000 41,096
6 0.7049 50,000 105,735 55,000 38,769
7 0.6650 150,000 99,750 55,000 36,575
8 0.6274 150,000 94,110 55,000 34,507
9 0.5918 50,000 88,770 55,000 32,549
0 0.5583 150,000 83,745 55,000 30,706
] 0.5267 50,000 79,005 55,000 28,968
12 0.4969 50,000 74,535 55,000 27,329
13 0.4688 50,000 70,320 55,000 25,784
14 0.4423 150,000 66,345 55,000 24 326
16 0.4172 35,000 14,602
Total discounted 1,394,130 1,275,779
benefits/costs
ee

As we can see, the total present value of costs is £1,275 million, whilst the total present value of
benefits is £1,394 million. Thus the net present value of the project is positive, at £118,351, which passes
the CBA test at the 6 per cent discount rate. Note how little the site renovation cost of £35,000 amounts
to in present-value terms: less than half its future value.
53

BOx'3.3, Discounting and the Discount Rate

Discounting means placing a lower value on benefits and costs the further away in time they occur. Why
might this make sense? Two main reasons have been given for discounting. These revolve around:

® the productivity of capital, and

® preferences.

These motivate two possible choices for the discount rate to be used in public-sector policy and project
appraisal:

® The social opportunity cost of capital, r, and

® The rate of social time preference, s.

Economies grow over time for many reasons, but an important one is that by building up the stock of
capital, an economy increases its potential output. The act of investing in a new factory is expected to
generate a flow of returns over time to the owner of that capital, in terms of annual sales of goods
produced. Across the entire economy, invested capital generates a positive rate of return, meaning that
the value of consumption goods in year t+ 1 that could be produced should all of the resources of an
economy be invested in year t will be greater than the maximum value of consumption goods that could
be produced in year t. However, capital is scarce: investing £1 million in a new factory means that we
cannot invest the same £1 million in another scheme. Choosing to invest in a particular scheme thus
involves an opportunity cost, which is the return on capital forgone from some other use (in particular,
from its most profitable alternative). Across the economy as a whole, we could rank investment projects
in terms of their rates of return. These rates of return show the net benefits from investing resources
rather than consuming. At the margin, this is known as the opportunity cost of capital, which can be used
to measure the social opportunity cost of capital, r.
The other motivation for discounting is that ‘pure time preference’—the desire for benefits to come
sooner rather than later—is a fundamental feature of human desires. Various motivations have been
suggested for time preference: impatience; the fact that we might not be around in the future to collect
on benefits; that future benefits are less certain than present-day benefits; and that we might expect to be
richer in the future, and thus will regard each extra pound of income as less valuable than we do today.
An important distinction is between a discount rate that applies to individual well-being and that which
might be applied to collective well-being. We could refer to the former as being a reflection of individual
time preference and the latter as a reflection of the rate of social time preference, s.
For more detailed discussion ofthe discount rate and of alternative approaches to discounting, see
Hanley and Barbier (2009: ch. 7).

of discounting ofbenefits and costs, whilst Box 3.3 discusses the concept ofdiscount rates in
more detail.
How is this time effect taken into account, and how are cost and benefit flows made com-
parable regardless of when they occur? The answer is that all cost and benefit flows are dis-
counted, using a discount rate that is here assumed to be a market rate of interest, i. The
present value (PV) of a cost or benefit (X) received in time ¢is then calculated as follows:

PV(X,) = X,[Q +a" ]. (3.1)

The expression in square brackets in equation (3.1) is known as a discount factor. Discount
factors have the property that they always lie between 0 and +1. The further away in time a
54 ECONOMIC TOOLS FOR THE ENVIRONMENT

cost or benefit occurs (the higher the value of t), the lower is the discount factor. The higher
the discount rate i for a given t, the lower is the discount factor, since a higher discount rate
means a greater preference for things now rather than later.

(5) Applying the net present value test

The main purpose of CBA is to help select projects and policies that are efficient in terms
of their use of resources. The criterion applied is thé net present value (NPV) test, which is
how the Kaldor-Hicks compensation principle is implemented. This test simply asks
whether the sum of discounted gains exceeds the sum of discounted losses. If so, the pro-
ject can be said to represent an efficient shift in resource allocation, given the data used in
the CBA. The NPV ofa project is thus as follows:

NEV (=> BALA) = CA), (3:2)

where the summations (indicated by the ‘Y’ symbols) run from t = 0 (the first year of the
project) to t= T (the last year of the project). Note that no costs or benefits before year 0 are
counted. The criterion for project acceptance is as follows: accept if NPV > 0 (i.e. is positive).
Any project passing the NPV test is deemed to be an improvement in social welfare.

(6) Sensitivity analysis

The NPV test described above tells us about the relative efficiency of a given project, given
the data input to the calculations. If this data changes, then clearly the results of the NPV test
will change too. But why should data change? The main reason concerns uncertainty. In
many cases in which CBA is used, the analyst must make predictions concerning future
physical flows (e.g. the quantity of electricity produced per year) and future relative values
(e.g. the wholesale price of electricity). None of these predictions can be made with perfect
foresight. When environmental impacts are involved, this uncertainty may be even more
widespread; for example, if a policy to reduce global greenhouse gas emissions is planned,
then the impacts ofthis in terms of avoided damage may be subject to a wide range of pre-
dictions. Therefore, an essential final stage of any CBA is to conduct a sensitivity analysis.
This means recalculating the NPV when the values of certain key parameters are changed.

[Link] Why is CBA useful?

In one very important sense, the practice of CBA addresses what might be called the funda-
mental economic problem: how to allocate scarce resources in the face of unlimited wants.
Resources are scarce because the sum total of demands on them exceeds their availability, and
using up scarce resources in one way imposes an opportunity cost on society in that we can-
not use those same resources for some other purpose. For example, a proposal in 2007 to
expand irrigated agriculture on the Canterbury Plains in New Zealand suggested diverting
water from two rivers to a newly constructed reservoir that would then be used to supply
irrigation schemes for dairy farmers. However, if land is used up to create a reservoir, that
same land cannot also be used for sheep farming. If water is taken from a river to supply a
reservoir and then to irrigate dairy farms, that same water is not available in the river to main-
tain in-stream ecological quality, or to support water-based recreation such as kayaking.
VALUING THE ENVIRONMENT: CONCEPTS 55

Society might find it useful, in determining whether to allow such schemes, to know whether
the economic benefits of irrigated dairy farming were bigger or smaller than the costs of res-
ervoir construction, lost sheep farming output, losses in river ecological quality and losses in
kayaking opportunities.
CBA is a decision-aiding tool that conveys this manner of useful information to decision-
makers. Not only does CBA allow a comparison of the benefits and costs of particular
actions, reflecting therein the scarcity of resources, but it also allows for ordinary people’s
preferences to be included in government decision-making. Economic values in a CBA
depend partly on what people like (their preferences), what they are prepared to give up to
have more of what they like (their WTP), and what they can afford to pay (their budget
constraints). In a sense, CBA is an exercise in economic democracy, since every citizen gets
an economic vote in terms of his or her WTP. CBA is also a formal way of setting out the
impacts of a project or policy over time, of organizing debate over an issue, and of identify-
ing who enjoys the gains and who suffers the losses from such undertakings. It is also, as
Arrow et al. (1998) have noted, a good way of ensuring consistency and perhaps transpar-
ency in public-sector decision-making.

[Link] Uses of environmental CBA

One way in which CBA can be useful is as part of the policy appraisal process. Worldwide,
much of the funding for environmental valuation studies has come from government
departments and agencies with responsibilities for environmental policy design and imple-
mentation (e.g. in the United Kingdom, with the Forestry Commission and the Environment
Agency); or with responsibility for policies that impact on the environment (e.g. roads pol-
icy). Within the European Union, CBA is an important aspect of implementing the Water
Framework Directive and the REACH directive on chemicals registration. Within both the
UK and the USA, CBA is also a part of the process by which the costs to the economy of new
government regulations—for example, the costs of setting stricter targets for recycling of
waste (garbage)—are regularly assessed.
Much early work on CBA was carried out in a project appraisal context. A good early
example is its use in assessing the development of hydroelectric power in the USA (see
Box 3.4 and Krutilla and Fisher, 1985). CBA is also used by public forest authorities in the
UK in assessing the net benefits of alternative forest management regimes, and in the
appraisal of major transport projects, such as new rail lines. The World Bank also has a long
history of using CBA in project appraisal. Many governments worldwide have official
guidelines on how CBA should be applied to the appraisal of public-sector projects (see, e.g.,
[Link] which describes the proce-
dures followed in the UK for both policy and project appraisal).

3.3.2 Other uses of environmental valuation

[Link] Setting environmental taxes

Environmental valuation has been used in the UK for setting eco-taxes; for example, with
regard to the landfill tax and the tax on quarrying. Valuation has been used in both justifying
56 ONOMIC TOOLS FOR THE El

Kotchen et al. (2006) carry out a CBA of the re-licensing of two hydroelectric dams in Michigan. The
policy context involves a move to reduce the environmental impacts of hydropower operations, by
changing how rivers are managed. The changes investigated by Kotchen et al. involve managing releases
from dams and reservoirs in a way that more closely parallels natural fluctuations in water levels, rather
than timing releases to coincide with maximum electricity demands. This change imposes costs in terms
of lost electricity output on hydro operators at peak periods, which must be compensated for with more
expensive output from other sources—here, from fossil fuel-powered generation. The gain is an
environmental one—in this case, an increase of about 270,000 salmon per year emigrating from the
Manistee River to Lake Michigan. Due to the mix of fossil fuel power supplied to the grid, there is also an
environmental gain from reduced net air pollution, since the peak-period demands are met from less
polluting natural gas-powered generation rather than the more polluting coal sources.
The costs to producers of the change in operations is given by the differences in marginal costs per
kilowatt hour (kWh) between hydro-derived and fossil fuel-derived electricity. This implies that the
annual costs for the two dams rise by about $310,000. For air pollution, the authors consider five
pollutants, including NO,, CO,, and SO,. Changes in air pollution between the two water management
regimes are then converted into dollars using estimates from the literature of marginal damage costs,
reporting a range of possible values. Finally, changes in migrating salmon numbers are converted into
changes in predicted catches for recreational anglers, and then valued using travel cost-derived estimates
of the value of recreational fishing (see Chapter 4).
The conclusion ofthe study is that the benefits of changing the way in which the river system is
managed for hydropower produces benefits that are larger than costs. Annual losses in electricity output
imply costs in the range of $219,132-$402,094, with a best guess of $310,612. Annual benefits from
emission reductions are in the range $67,756-$246,680, whilst gains in recreational fishing are worth
$301,900-$1,068,600, with a most likely estimate of $738,400. The authors conclude that ‘the benefits
exceed the costs of the switch ... even ignoring the air quality benefits entirely, the best estimate of
recreational fishing benefits exceeds the upper bound of producer costs’. In this case then, changing
how water resources are managed to reduce adverse environmental impacts seems to pass the cost-
benefit test.

a tax and determining its level. However, the application of an estimate of the average exter-
nal cost at the current level of activity does not constitute the Pigovian tax that it is made out
to be, since this would typically measure the marginal external cost at the optimal level of
externality. Here, the crucial issue would appear to be to know how marginal damages vary
with the level of the externality-causing activity.

[Link] Environmental damage claims

Under the CERLA and Oil Pollution Acts in the USA, firms that cause accidental environ-
mental damages can be sued in the courts by states and by the Federal government to recover
the monetary value of such damage. The most famous such incident to date has been the
Exxon Valdez oil spill in 1989; another example was the accident involving the oil tanker
American Trader in February 1990, which spilled up to 400,000 gallons ofcrude oil into the
Pacific Ocean off the coast of California (Dunford, 1999). The State of California sued the
‘responsible parties’ for damages. Interestingly, in this case, the two parties produced
VALUING THE ENVIRONMENT: CONCEPTS ya

alternative sets of estimates of these damages, which resulted in the main from the tempo-
rary closure of a number of beaches. The State of California's experts estimated the value of
a lost day’s recreation on the beach to be around $15 per visit; unsurprisingly, the defendants’
economists came up with a lower value of $4-8 per visit. Arguments also raged about how
many visits were lost.

[Link] National accounting

In Chapter 6, we will discuss the issue of making ‘green’ adjustments to national accounting
figures such as gross national product, as a way of producing a better measure of welfare and
an indicator of sustainability. Governments can use environmental valuation in calculating
such green adjustments to the national accounts; for example, to take account of changes in
the level of pollution between this year and the previous year, or to value changes in water
quality in a nation’s rivers.

Summary
Changes in the quantity or quality of environmental resources have economic value if they
have an impact on utility. We can base our measures of these values upon either the most
that people are willing to give up to acquire some (desirable) change or the least they are
willing to accept to forgo it (for an undesirable change, then we can use either the most
people are willing to pay to prevent it, or the lowest compensation they would accept to put
up with it). Governments can and have made extensive use of environmental valuation ina
number of contexts, particularly cost-benefit analysis in the context of policy and project
appraisal. Other uses also exist, including the settling of environmental damage claims and
the calculation of green tax rates.

Tutorial Questions
3.1 Why do economists use WTP as a measure ofthe value someone places on an
environmental good or ecosystem service?
3.2 What determines how WTP for a particular change in environmental quality (e.g. a
fall in air pollution) may vary across people?
3.3 Several studies have shown that WTP differs substantially from WTA for a given
change in environmental quality. Why might this be? Does it matter, from a policy
analysis viewpoint?
3.4 What is meant by the economic value of an ecosystem service? How could changes in
pressures on an ecosystem (such as an estuary or a forest) produce a change in
ecosystem service values? What would we need to know to be able to estimate the
monetary value of such a change?
3.5 Explain how you would undertake a cost-benefit analysis of aplanned new nuclear
power station. Why might the choice of discount rate be particularly crucial to the net
58 ECONOMIC TOOLS FOR THE ENVIRONMENT

present value calculation in this instance? How could uncertainty over (i) the future
price of electricity and (ii) future waste storage costs be included in this analysis?
3.6 What are the advantages and problems of applying cost-benefit analysis to
environmental policy decisions?

References and Additional Readings '

Aldred, J. (2006). ‘Incommensurability and (Washington, DC: The National Academies
monetary valuation, Land Economics 82(2): Press).
141-61.
Kotchen, M., Moore, M., Lupi, E., and Rutherford, E.
Arrow, K., Cropper, M., Eads, G., Hahn, R., Lave, L., (2006). ‘Environmental constriants on
Noll, R., Portney, P., Russell, M., Schmalensee, hydropower: an ex post benefit cost analysis
R., Smith, V.K., and Stavins, R. (1998). ‘Is there a of dam re-licensing, Land Economics 82(3):
role for benefit-cost analysis in environmental, 384-403.
health and safety regulation?’ Environment and Krutilla, J.V., and Fisher, A.C. (1985). The Economics
Development Economics 2: 196-201. of Natural Environments (Washington, DC:
Banzhaf, H.S. (2009). ‘Objective or multi-objective? Resources for the Future).
Two historically competing visions for benefit Loomes, G., Starmer, C., and Sudgen, R. (2003).
cost analysis, Land Economics 85(1): 3-23. ‘Do anomalies disappear in repeated markets?’
Barbier, E.B. (2009). ‘Ecosystems as natural assets) Economic Journal 113 (March): C153-66.
Foundations and Trends in Microeconomics 4(8): Pearce, D.W., and Turner, R.K. (1989). Economics of
611-81. Natural Resources and the Environment (London:
Dunford, R. (1999). “The American Oil Trader spill’ Harvester Wheatsheaf).
AERE Newsletter (May): 12-20. Sterner, T. (2011). Fuel Taxes and the Poor
Hanemann, M. (1991). “Willingness to pay and (Washington, DC: Resources for the Future).
willingness to accept: how much can they differ?’ Van der Wal, R., Bonn, A., Monteith, D., Reed, M.,
American Economic Review 81: 635-47. Blackstock, K.L., Hanley, N., Thompson, D.,
Hanley, N., and Barbier, E.B. (2009). Pricing Nature: Evans, M., Alonso, I., Allott, T., Armitage, H.,
Cost-Benefit Analysis and Environmental Policy Beharry, N., Glass, J., Johnson, S., McMorrow, J.,
Appraisal (Cheltenham: Edward Elgar). Ross, L., Pakeman, R.J., Perry, S., and Tinch, D.
(2011). ‘Mountains, moorlands and heaths’ in
Heal, G.M., Barbier, E.B., Boyle, K.J., Covich, A.P.,
The UK National Ecosystem Assessment Technical
Gloss, S.P., Hershner, C.H., Hoehn, J.P., Pringle,
Report (Cambridge: UNEP-WCMC);: 106-59.
C.M., Polasky, S., Segerson, K., and Shrader-
Frechette, K. (2005). Valuing Ecosystem Services: Willig, R. (1976). ‘Consumers surplus without
Toward Better Environmental Decision Making apology, American Economic Review 66: 589-97,
Valuing the Environment:
Methods

In this chapter, we are going to consider different approaches to estimating the economic
values for environmental change that were explained in Chapter 3. These methods
include:

e Stated-preference approaches, such as contingent valuation and choice experiments.

e Revealed-preference approaches, such as travel-cost models and hedonic pricing.
e Production-function approaches, which value the environment as an input.

We will also consider the problem of ‘benefits transfer’. For more detail on all of the meth-
ods outlined in this chapter, see Hanley and Barbier (2009); or for a more technical treat-
ment, see Freeman (2003) or Haab and McConnell (2002).

4.1 An Overview of the Methods

How do we empirically estimate economic values for non-market environmental goods?

Whilst most of the methods used are aimed at measuring the public’s willingness to pay
(WTP) for changes in the natural environment, some are directed instead at evaluating
the contribution the environment makes as an input to the production of goods and ser-
vices, which are themselves sources ofutility. Directly or indirectly then, all environmen-
tal values are traced back to effects on utility. There are many ways of categorizing
environmental valuation methods: one useful way is to place methods into the following
groups:

1. Stated-preference methods.
2. Revealed-preference methods.
3. Production-function approaches.

Stated- and revealed-preference methods focus on direct impacts of the environment on

utility. Production-function approaches focus on the environment as an input. Valuing
changes in ecosystem service flows can involve the use of all of these approaches.
60 ECONOMIC TOOLS FOR THE ENVIRONMENT

4.2 Stated-preference Approaches

These include the following:

e Contingent valuation
e Choice experiments .

These methods have the common feature that they are all based on surveys in which the
public (or some specific subset of the public, such as users of some recreational resource
such as a national park) is directly questioned about its WTP or willingness to accept com-
pensation (WTA) for hypothetical changes in environmental quality. The contingent valua-
tion method (CVM) is the older and more developed method, but choice experiments are
now a fast-developing technique.

4.2.1 Contingent valuation

Contingent valuation originated in the early 1960s, but did not become widely used until
the mid-1970s. The publication of Mitchell and Carson’s book Using Surveys to Value Public
Goods in 1989 was an important milestone in the development of the method, and by 1995,
Richard Carson had identified over 2,000 published CVM studies. CVM has been applied in
many settings all over the world, often to help answer important questions of public policy
design. In principle, the method is very simple. Given that the absence of prices for many
environmental goods is due to the absence of a market, CVM asks respondents how they
would behave ifsuch a market existed. For example, a CVM survey could ask:
Suppose the only way of improving water quality in your local river was for all residents to agree
to pay a surcharge on their taxes. If you were asked to pay an additional $25 per year, would you
agree in order to have these water quality improvements go ahead?

Respondents are drawn from a random sample of the relevant population, which might in
different settings comprise the general public, local residents, or visitors to a recreation area.
The important point about CVM is that respondents are asked to reveal what they would be
willing to pay (or willing to accept as compensation) for a clearly specified hypothetical
increase or decrease in environmental quality.
There are several main design features of aCVM questionnaire:

e People must be given a reason why they might be asked to pay for something that they
currently do not see themselves as paying for. For example, if the purpose of the CVM
was to estimate the benefits of an improvement in river quality, respondents might be
told that extra local tax revenues would be needed to finance an investment in better
sewage treatment. They must also think that their responses are in some way
consequential—for instance, that they will be used to help inform a decision.
® A bid vehicle must be used that is both credible and uncontroversial. The bid vehicle is
the means by which respondents pay in the hypothetical market (local taxes, in the above
example). Bid vehicles must be credible in the sense that respondents feel that they could
be applied in practice. For example, if the value of a large wilderness area with many
VALUING THE ENVIRONMENT: METHODS 61

access points was being studied, the specification of an entrance fee as a bid vehicle would
not be credible if people thought it would be impossible to enforce, or if it would be too
politically unpopular to imagine it being introduced.
e Respondents should be given adequate, unbiased information on the environmental
good and its hypothetical market, in order to let them make an informed judgement.
e A decision has to be made on how to ask the WTP/WTA question. This can be done using
an open-ended format (“What is the most you would be willing to pay?’), through payment
cards where respondents are shown a series of amounts and are asked to indicate their
maximum WTP/minimum WTA from amongst these amounts; and dichotomous choice
formats, where respondents are asked to say whether they would be willing to pay—or
willing to accept as compensation—a specific amount, known as the bid price. This bid
price is then varied across individuals, which yields yes/no responses to different amounts.
A refinement of the dichotomous choice approach is to ask those people who say ‘no’ to the
first amount offered if they would be willing to pay a lower bid, and to ask those who said
‘yes’ to the first amount offered if they would be willing to pay a higher amount. This is
known as double-bounded dichotomous choice. All of these alternatives have advantages
and disadvantages, but most researchers now avoid the use of open-ended formats.
e ‘Protest bids’ should be identified. When respondents are asked how much they would
be willing to pay, a proportion will likely give a zero response. For some people, this is
because they do not value the good, in that it does not impact on their utility. If lam
asked my WTP to protect a wildlife species that I don’t care about, then I place zero
value on it and, accordingly, will state a zero WTP. I may also say that I would not be
willing to pay anything because | cannot afford it. Both types of response are referred to
as ‘genuine zeros’. But I might bid zero, even if I care about the species, because I am
protesting about being asked the question in this way, or because I do not find the
hypothetical market to be credible. Similarly, respondents who are asked their minimum
WTA to allow a local woodland to be felled may state a zero amount because they feel
that no amount of monetary compensation would make up for the loss of the wood: this
is another form of protest bid. Protest bidders are usually separated out from genuine
zeros and positive bidders before analysis progresses.
e Many surveys include debriefing questions, which seek to analyse how well respondents
understood the survey questions, what exactly they thought they were paying for/being
compensated for, how credible they found the survey, how sure they are about their
responses, and whether the survey had changed their opinions on the issue at hand.

Given all these issues, CVM studies start by undertaking focus group sessions, in which
different scenarios, bid vehicles, information sets, and question formats can be tested out by
the researcher in small groups before surveying begins. Surveys may be carried out by mail,
telephone, Internet, or face-to-face interviews. Once a sufficient sample has been collected,
mean or median WTP/WTA is calculated from the sample, using the individual responses
(once protest bidders have been excluded). This sample average can then be aggregated into
a population mean/median. A bid curve analysis is usually undertaken, whereby WTP
responses are related, using regression analysis, to variables thought likely to influence
them, such as education, experience with the good, and income. The purpose of doing this
62 ECONOMIC TOOLS FOR THE ENVIRONMENT

would be to: (i) see how much of the variation in WTP across the sample could be explained
(and therefore how much is unexplained); and (ii) whether the signs on the variables of
interest are in accord with such a priori expectations as we have: for example, income would
be expected to be positively related with WTP, in that higher-income levels should, on
average—and other things being held equal—imply higher WTP amounts (Jacobsen and
Hanley, 2009). For some CVM designs, notably dichotomous choice versions, bid curves
must be estimated in order to calculate mean WTP.
The fact that CVM has become so widely used implies that it has some advantages as a
method. Principal amongst these are the following:

1. It isa very generalizable method, in that it can be applied in an extremely wide range of
situations, from the benefits of preserving global biodiversity to the benefits of
improving a city’s air quality or protecting a local wetland.
2. It is capable of measuring both use and non-use values. Non-use (existence) values
have been found to be very important in many cases. CVM questionnaires can also be
designed so that the researcher gains some insight into why people value a given
environmental good, and how this valuation changes when, for example, uncertainty
surrounding the supply of the good changes.

However, CVM has also attracted much criticism. In general, this has focused on three
issues. First, CVM measures what people say they would do, which may be different from
what they would actually do. Stated WTP could be greater or less than actual, true WTP for
a number of reasons. If respondents think that their answer may influence how much they
would actually get charged, they may free-ride by understating their WTP. On the other
hand, if respondents believe that their answer is not linked to what they would actually be
charged, but is linked to how likely the environmental change is to happen, then they may
overstate their WTP for an environmental change, which increases their utility. Testing for
such hypothetical market bias is clearly a little difficult for many environmental goods, since
it is the absence of markets for such goods (and thus of market prices) that encourages us to
use CVM in the first place. Nevertheless, experiments can be set up that compare real and
hypothetical payments. Murphy et al. (2005) found that across a number of studies, hypo-
thetical WTP was on average 2.6 times greater than actual WTP. The issue of ‘calibrating’
CVM responses has thus attracted much attention. Research has found that the extent to
which hypothetical WTP over- or understates actual WTP depends on the type of environ-
mental good being studied and the design of theCVM exercise itself—in particular, on how
the payment question is asked, and on how ‘consequential’ people think that their responses
are. Other researchers have found that filtering CVM responses in terms of how sure
respondents are of their WTP statements can help reduce hypothetical market bias (Poe
et al., 2002).
Another criticism of CVM is that it produces estimates of WTP that are insensitive to the
amount of the environmental good being bid for: the ‘scoping’ problem. One reason for
people saying they would pay roughly the same to protect one lake in Ontario from acidifi-
cation as they would pay to protect all lakes in Ontario is that their stated WTP is actually a
symbolic number motivated by a feel-good factor referred to as ‘warm glow’, Attention has
thus focused on the results of scope tests, which measure WTP for different quantities of the
VALUING THE ENVIRONMENT: METHODS 63

same good (Heberlein et al., 2005). For example, different subsamples of apopulation could
be asked their WTP to protect 100, 500, or 1,000 hectares of forests from felling, and statisti-
cal tests performed to see whether WTP was indeed sensitive to the area of forest
protected.
Third, CVM results have been criticized as being dependent on the information they
provide to respondents, and of asking respondents to undertake a task that they are not up
to. In many cases, the population of interest may be quite uninformed about the environ-
mental resource that is being studied: we would not expect people to be able to give a WTP
figure for a function of an ecosystem that they do not understand. For example, suppose that
a policy of protecting the population of an obscure species is being evaluated. We wish to
know the value people place on this species, yet we expect that many folks will never even
have heard of it before. To successfully implement the survey, researchers will need to
inform respondents about the species, the threats to it, and what can be done to avert these
threats. Yet this changes their knowledge in the process of undertaking the survey: thus by
implementing a CVM survey, we are possibly changing the preferences that we wish to
measure (MacMillan et al., 2006). Different types and amounts of information may signifi-
cantly affect stated WTP. The question that CVM must answer, then, is ‘How much infor-
mation is enough, and what should it cover?’

4.2.2 The choice experiment method

The choice experiment method! adopts a particular view on how the demand for the envi-
ronmental goods is best pictured, known as the characteristics theory of value. This states

In many countries, the landscape that we see now has been formed by a tremendous mixture of human
and natural influences. For example, semi-natural habitats such as hay meadows and heather moorland
are the product of a particular type of farming regime, combined with environmental conditions that
make such habitats viable. The built environment is also part of the landscape, and historical remains are
very important in this context. One of the most famous archaeological sites in Europe is Stonehenge, on
Salisbury Plain in England, a World Heritage Site. Stonehenge, as its name implies, is a henge monument
(circle and surrounding ditch) made of stones, in this case massive Sarcen stones. It dates from 5,000 to
3,500 years ago.
In the late 1990s, the quality of the site for visitors was increasingly impaired by the proximity of a noisy
and busy road, the A303. Accordingly, English Heritage, who manage the site jointly with the National
Trust, proposed the construction of a covered tunnel for the A303 in the area around the site, which
would essentially hide all of the passing traffic and greatly reduce noise levels. The cost of this project was
estimated at £125 million: but was it worth it? To answer this question, a contingent valuation study was
commissioned. Mourato and Maddison (2000) designed a survey whereby visitors and the general public
were asked their WTP in higher taxes for the road tunnel project. The general public were included in
order to capture non-use values. The most conservative estimate of benefits showed them to be worth
£150 million; that is, more than the cost of the tunnel.

! Sometimes referred to as ‘choice modelling’ or, less commonly, as ‘conjoint analysis.

64 ECONOMIC TOOLS FOR THE ENVIRONMENT

that the value of, say, a forest, is best explained in terms of the characteristics or attributes of
that forest. Different forests are actually different ‘bundles’ of attributes, and it is these bun-
dles that people value. Moreover, the value of any particular forest then can be broken down
into the values of each attribute of that forest. Using observations of people’s choices
between different bundles of attributes, the researcher can infer: (i) which attributes signifi-
cantly influence their choices, (ii) assuming that price or cost is included as one attribute,
what they are willing to pay for an increase in any other attribute; and (iii) what they would
be willing to pay for a policy that changed several attributes simultaneously. The choice
experiment (CE) method is becoming increasingly popular as a tool for estimating and
indeed investigating environmental values. Policy-makers have seen a powerful set of
advantages for the CE method, in terms of being able to measure benefits for a wide range
of policy changes. Birol and Koundouri (2008) give several examples of the use of the
method in the policy process. For a very useful guide to the CE method, see Henscher et al.
(2005).
In the choice experiment method, the researcher first ofall identifies the main attributes
that are relevant for describing the environmental good in question. This is done using
focus groups, and by finding out from policy-makers and administrators which aspects of
the environmental good are likely to be affected by a policy action. For a river, the attrib-
utes might be in-stream ecological quality, flow rates, and the condition of the river banks.
For a national park management problem, the attributes could be provision of guided
walks, set-aside of conservation areas, traffic management, and management of agricul-
tural areas. If the researcher wants to use the CE to measure economic values, then a price
or cost attribute must also be included. For river quality, it could be local water and sewer-
age rates; for a national park, it could be a tourist tax. The researcher needs to be sure that
the selected attributes are: (i) likely to be relevant in terms ofthe preferences of the popula-
tion to be surveyed; and (ii) likely to be amenable to change by environmental managers.
Different bundles of these attributes are then combined using experimental design
principles. Bundles are often arranged in pairs, and respondents asked to choose between
them and some status quo alternative, in what is known as a ‘choice set’. Typically, each
individual might answer between four and eight choice sets. For example, a study by
Morrison et al. (2002) looked at the benefits of protecting wetlands in Australia. Each
respondent was asked to choose most preferred alternatives amongst pairs of different
wetland management options, such as the choice set shown in Table 4.1 (this has been
adapted a little from the original). Respondents were asked: Which option would you pre-
fer that the government went ahead with?—A, B, or C?

Table 4.1 Choice experiment for valuing Australian wetlands

ce Ee ea ee ee ee
Management Management Management
option C (status
option A option B quo: no change on present)
Wetland area conserved (ha) 1,000 800 700
Bird species conserved (number) 40 30 25
Farm Jobs protected its 16 20
Cost per household in terms of increase 30 15 0
in local taxes over next 5 years (S)
ee ee ee eae eee ee ek eee Pee ee ee ee ee eee
VALUING THE ENVIRONMENT: METHODS 65

A choice experiment questionnaire would be designed, piloted, and implemented just

like a contingent valuation study, as described in the previous section. Similar requirements
exist for the description of the hypothetical market.
Once questionnaires have been completed, the researcher now has data on which options
individuals chose (option A, option B, the status quo), and she can relate these choices to the
levels that the attributes took in these options. In this way, choices can be statistically related
to attribute levels, including price. The usual statistical model employed is known as the
conditional logit model. This means that we can write down the probability that an indi-
vidual i chose a particular option A like this:

exp(UVia)
P.(choose A) = (4.1)
S’ exp(uVi)
‘s

where V is the ‘observable’ part of utility, / is a ‘scale parameter’ that shows the variance of
the errors in the choice model, and J are all the other options that the individual could have
chosen instead of A. A typical assumption is that V is a linear function of the attributes of
the good (and thus of the choice alternatives):

V=a+P,X,+B, X,+... B,X,+ B.C. (4.2)

For each attribute X,, X, ..., the model estimates a value f that shows the effect on utility
of a change in the level of each attribute. Thus B, shows the effect on utility of a change in
attribute X,. The model also estimates a parameter B., which is the effect of a change (increase
or decrease) in the price or cost of the option on the likelihood of choosing that option.
Software packages such as Stata and Limdep can be used for this kind of estimation. It is
interesting to know the B values, since now we know how much utility goes up or down
when the attributes increase or decrease (albeit moderated by the scale parameter). These
values tell us whether people prefer an increase or a decrease in each attribute; we can also
see, by looking at the prob or t-statistic values from the computer output, whether or not
these attributes are statistically significant.
The final steps in a choice experiment are to calculate WTP estimates, based on the B
values already discussed. The f values show the effect on utility of changes in the attributes,
but for cost-benefit analysis we need measures of WTP. For a marginal change in an attrib-
ute, this WTP value is given by, for attribute X,:

IP. = B,,/B.. (4.3)

This value for any attribute (other than price!) is called the implicit price (IP, in equation
(4.3)). For instance, in Table 4.1 one of the attributes was the number ofbird species con-
served. Dividing the B value for this attribute by the f value for the tax increase would show
the (average) WTP of people in the sample to increase the number of bird species conserved
by one. However, often we wish to value multiple changes in attributes. For instance, a new
policy on wetlands conservation could alter the area conserved (labelled A below), the num-
bers of bird species conserved (labelled B), and the provision of recreational trails, labelled R.
ECONOMIC TOOLS FOR THE ENVIRONMENT

The price for this would be an increase in local taxes, which are attribute c. The average WTP
for this suite of changes in attributes can be calculated using the following equations:

1
CS = Ve Vv.) (4.4)
B.
ve =at+B, Ao + Ba By + Pa Ros \ (4.5)

V, =a+B, A +, Bi + B,R;. (4.6)

This might look at bit complicated but is actually very easy, and can be calculated using a
spreadsheet once you have got your estimates from the choice model in equation (4.1).
Equation (4.4) says that the compensating surplus (CS) from an improvement in wetlands
conservation—that is, the average person’s WTP for this package of changes—is given by the
difference between their (measurable) utility before the improvement goes ahead, given by
V,, and their measurable utility after the change, V,, converted into monetary units using the
coefficient on the tax or price attribute, B.. In turn, utility in the ‘before’ and ‘after’ cases is
given by the levels of the attributes in each case (so Ap, By, and R, in the ‘before’ case, and Aj,
B,, and R, in the ‘after’ case), multiplied by the attribute coefficients, and including the term
a. This was the constant in equation (4.2), and it is usually referred to as the alternative spe-
cific constant. It shows the utility that people get simply from either staying in the status quo
or leaving it (depending on whether it is positive or negative), independently of the values
taken by the attributes. By fixing the status quo utility and varying the levels of the attributes,
compensating surplus figures can be produced for as many combinations of attributes and
levels as the design makes possible; that is, for a wide range of policy outcomes.

[Link] Problems with the choice experiment method

Accommodating variation in preferences across people. The standard approach to choice

experiments, which was described above, has one important feature that is worthy of com-
ment. This is that if we use a conditional logit model to represent the choices that people
make, then we are effectively assuming that each person in the sample places the same value
on each attribute used in the design. In other words, we effectively assume that the marginal
utility for Joe if attribute X, is increased—B,—is the same as the marginal utility for Jane;
and that the marginal utility for Joe of an increase in attribute X,, B,, is the same as that for
Jane. ‘This is because we only estimate one value for B, and one value for B, in equation (4.2).
However, we might expect that people will care about the same attribute to differing degrees.
Choice experiment practitioners have thus looked for alternative ways of modelling such
preference heterogeneity. For an example ofdifferent approaches to solving this problem, see
Birol et al. (2006).
Issues with experimental design. Designing a choice experiment is almost an art form!
Decisions must be taken on a great number ofissues:
1. What attributes to include.
2. How to describe them to respondents.
3. What levels are to be used for each attribute.
VALUING THE ENVIRONMENT: METHODS 67

4. What price or cost term will be used.

5. How the attributes and levels are combined in choice sets.
6. How many choice sets respondents can deal with.
7. How many choice options are included in each choice set.

The overall success of the choice experiment in terms of what it tells us about people’s
choices and values depends on these steps. Many papers exist that investigate these issues,
mostly in non-environmental applications of the method (e.g. in a transport, marketing,
and health context): lessons learnt can be found in choice experiment textbooks such as
Henscher et al. (2005).
Hypothetical market bias. Another parallel between choice experiments and contingent
valuation is the possibility that responses in a hypothetical market setting will tell us little
about how respondents would behave in a real market. This issue has been addressed in a
couple of ways within the CE literature, comparing real with hypothetical responses in
terms of (i) how well hypothetical choices predict real choices and (ii) how close predicted
WTP from hypothetical choices is to real WTP in an actual market. Of course, the same
problem faces the CE practitioner as faces the CVM analyst—that for most environmental
goods, we cannot observe ‘real’ market prices! However, some findings exist that compare
real with hypothetical choices. Evidence is presented by List et al. (2006), who compared
actual with hypothetical scenarios for two choice experiments. They argue that two tests are
of interest—whether a hypothetical choice experiment overstates the extent to which people
would actually pay for, say, wetland conservation and the differences, if any, in the marginal
values of the attributes used in the choice experiment between ‘real’ and ‘hypothetical’
choices. They found statistically significant differences between hypothetical and real WTP,
but not between the marginal values of attributes. Ready et al. (2010) also study this
problem—their findings are discussed in Box 4.2.

4.3 Revealed-preference Approaches

In revealed-preference (RP) approaches, the analyst tries to infer the value that people place
on environmental goods from their behaviour in markets for related goods. A major differ-
ence between RP and stated-preference approaches is thus that with RP methods we make
use of people’s actual behaviour, rather than their intentions. Two principal RP approaches
will be discussed here: the hedonic pricing method and travel-cost models.

4.3.1 The hedonic pricing method

The hedonic pricing method (HPM) has its basis in the same characteristics theory of value
as choice experiments. In other words, people are pictured as valuing goods in terms of the
bundles of attributes that these possess. For a house, these attributes might include the
number of bedrooms, the age of the house, the size of garden, and whether it has a garage.
These could all be a particular property’s site characteristics, S; Also important to buyers
and sellers will be the locational aspects of the house, such as how far it is from major
68 ECONOMIC TOOLS FOR THE ENVIRONMENT

As the main text indicates, one worry in the choice experiment method is the extent to which estimates
of WTP produced by this approach suffer from a hypothetical market bias. Ready et al. (2010) investigate
this issue by carrying out a choice experiment with both hypothetical and real payments. Their
application is concerned with the extent to which people's uncertainty about how they value an
environmental good can help explain the degree of hypothetical market bias.
A choice experiment was conducted using students from Penn State University as the subjects. The
good being valued was a rescue programme for injured wild animals called Centre Wildlife Care (CWC).
The payment vehicle was donations to the CWC, which is funded entirely from donations. The attributes
in the design were:

® Whether care was focused on mammals, birds, or turtles

w Whether
the animal was common or rare

@ Whether the animal could be returned to the wild, after care

If respondents chose the ‘no donation’ option, then they were told that animals would need to be turned
away, since funds did not permit their rehabilitation. After making each of four choices, respondents
were asked how certain they were about each choice. Two types of experiment were run, depending on
whether the donations were purely hypothetical or were for real (in other words, respondents had to
hand over the money that they said they would donate at the end ofthe session).
Some 249 surveys were completed. Mean WTP for hypothetical payments was around $5, and
mean WTP with real payments was around $1.70. However, the results also showed that respondents
who gave hypothetical responses that were significantly different from their real behaviour tended to
state higher levels of uncertainty over their choices. This suggests that asking people how sure they
are about choices, and then filtering out those who are rather unsure, would be one way of reducing
hypothetical market bias in choice experiments. Calibrating responses by reclassifying people who
chose a payment option in the hypothetical scenario with a level of stated uncertainty greater
than a given threshold value produces the result that mean WTP with hypothetical payments is
insignificantly different from mean WTP with real payments (calibrated hypothetical WTP = $1.71;
real WTP = $1.68).

employment areas, how good the local school is, and how good its public transport links are.
Call these the neighbourhood characteristics, N;. Finally, the environmental characteristics
(E;) may also be an important determinant of house prices: for example, these could include
noise levels, air quality, scenic views, and proximity to landfill sites (see, e.g., Eshet et al.,
2007).
The basic assumption of the HPM is that people’s valuation of environmental attributes
can be inferred from the amount they are willing to pay for these attributes through the
housing market. For example, other things being equal, a house in a quieter part of town
may sell for more than a similar house in a noisier part of town. If Ivalue peace and quiet,
then I may be willing to pay this premium, whilst sellers will know this when advertising
their house. For the buyer, the highest premium he or she would be willing to pay for any
environmental attribute would indicate the maximum value that he or she places on it. If
these premia could be identified from market transactions, then this would thus tell us
something about the value of those environmental attributes (such as noise) that can be
linked to house prices.
VALUING THE ENVIRONMENT: METHODS 69

The HPM thus proceeds by collecting data on (most usually) house prices from sales
records,’ along with data on E;, N,, and S,. A regression analysis can then be carried out to
estimate the equation:

ait Ce rae ean INT

Vind Sy casVig
Gs), (4.7)

where there are m environmental attributes, n neighbourhood attributes, and q site attrib-
utes, and where P,; is the price of the ith house. For those attributes that turn out to have a
statistically significant effect on house prices, the ‘implicit price’ (that is, the price premium)
can be calculated. This might tell us, for example, that a 1 per cent improvement in air-
quality levels increases house prices by 0.2 per cent on average: this could then be used to
work out the implied WTP in money for this air-quality improvement. In this manner,
money values can be placed on environmental attributes that are linked to house prices. In
Figure 4.1, we show the kind of relationship that one might find in the data: as the level of
air pollution falls, and so air quality rises, house prices go up, other things being equal. In
other words, to ‘buy’ an improvement in local air quality from E, to E,, the house-buyer has
to spend an additional amount equal to AP.
The HPM has been widely used to study the implicit prices of changes in air quality, noise,
and proximity to waste sites. An excellent survey of much of this work can be found in Taylor
(2003). A number of drawbacks can be identified with the technique, as follows:

e Many environmental goods are not linked to housing markets. For these goods, the
HPM will not work. Even for those goods that are so linked, the method provides only
an indication of partial value. For example, air-quality improvements may benefit
visitors and commuters to a city, as well as those who own a house there, but only
house-owners’ values get picked up by the method.

Pa 221

House
prices

E, E,
Air quality

Figure 4.1 The value in the housing market of an improvement in air quality.

2 Information on property rents can also be used.

70 ECONOMIC TOOLS FOR THE ENVIRONMENT

e The method assumes that the housing market is in equilibrium in a rather special sense:
that, for every attribute, house-buyers are able to locate a house that allows them to
equate their marginal value for each environmental attribute with the marginal cost
(implicit price) of these attributes.
e Essentially, we must assume that all buyers and sellers are well informed about how
environmental attributes vary spatially across the area being studied.
e House purchases are investments, in the sense that people make them for a relatively
long period of time, often in the anticipation of capital gains. Their maximum WTP for
a house depends not just on current levels of environmental attributes, but also on
expectations of changes in the levels of these attributes over the time during which they
expect to own the house.
e A property market may be segmented into a set of sub-markets, each with its own
hedonic price functions, facing the researcher with the difficult task of discovering what
this segmentation looks like.

Box 4.3 shows an example of the HPM applied to valuing noise reductions.

43.2 Travel-cost models

Models of this kind are amongst the oldest environmental valuation techniques. They
originated in the United States, in the context of the planning and management of outdoor
recreation in national parks. Informal outdoor recreation is clearly a source of utility for
many people, and such recreation often takes place in areas managed or regulated by gov-
ernments. Activities such as hill-walking, kayaking, angling, climbing, cross-country skiing,
and rock climbing have all increased in terms of participation in the last 50 years, and seem
certain to be of increasing importance to land management in the future, along with more
informal types of recreation such as picnicking and dog-walking. But how could the eco-
nomic value of such activities be measured? What is a day’s rock climbing ‘worth’ to the
climber?
Stated-preference approaches such as CVM have been intensively used to estimate WTP
for outdoor recreation opportunities, and for changes in such opportunities (e.g. for water-
quality improvements in the context of recreational fishing). An alternative approach,
though, is to use travel-cost models. These are based on the observation that expenditure is
necessary to partake in such recreational activities. These expenditures include the time and
money spent in travelling to recreational sites (it may seem odd to talk about time spent as
an expenditure, but time is scarce for everyone, and using up scarce time has an opportunity
cost). An individual can be pictured as being willing to spend up to the value in utility that
they get from making such a trip. Typically, their actual spending will be less than the most
they would be willing to spend.
For example, consider the Picos de Europa national park in north-west Spain. This is a
mountainous national park close to the Atlantic coast, which is popular with walkers and
nature-lovers. Suppose that we want to estimate the value of informal recreation in this
area. Visitors would be irrational if they were to spend more in visiting the park than the
utility that they derive from their visit. For any individual, the total cost of visiting (time
VALUING THE ENVIRON ‘: METHODS FA

Urban noise is a persistent and growing problem for millions of people. Measures are taken to reduce
noise, but these impose costs on society. It is therefore of interest to enquire whether the benefits of
noise reduction strategies justify their costs. Day et al. (2007) use the hedonic price method to estimate
the benefits of reducing urban noise levels in Birmingham, UK. This research was used to inform the UK
Department for Transport’s assessment methods for transport system investments. Data was collected on
10,848 house sales during 1997 and on-the-site attributes of each house sold (e.g. total floor area). Using
a geographic information system, variables were constructed to measure the walking time to different
amenities and disamenities (e.g. landfill sites), and local school quality. Digital noise maps were then used
to summarize each property's exposure to road, railway, and aircraft noise. Further neighbourhood
variables were assembled to measure household wealth, ethnic minorities, age, and the number of
households with children.
Eight different market segments were identified in the data, based on income, property size, the ethnicity
of district, and the number of children. For each of these segments, a separate hedonic price model was
estimated. Road noise was found to significantly affect house prices in around half of all segments, although
aircraft noise only had a significant effect in two out of eight segments. They are then able to generate the
following results for 1 decibel (dB) increases in road and rail noise above a baseline of 56 dB:

Change in road noise Annual benefits, in £1,997 per annum

mean value and (95% confidence interval)
From 55 to 56 dB 31 (24-52)
From 60 to 61 dB 43 (33-75)
From 65 to 66 dB 55 (41-99)
From 70 to 71 dB 67 (49-123)

Change in rail noise Annual benefits, in £1,997 per annum

mean value and (95% confidence interval)
From 55 to 56 dB 83 (43-461)
From 60 to 61 dB 94 (49-519)
From 65 to 66 dB 106 (55-580)
From 70 to 71 dB 117 (61-654)

Notice that the costs of increases in railway noise are higher than those of equivalent increases in road
noise. Also, the economic costs of a1 dB increase are higher the greater the baseline is from which this
increase occurs. This paper is a very good illustration of how complex hedonic price modelling can be.

plus distance) will almost certainly be less than the maximum they would be willing to
spend, which is equal to the value that they place on a trip to the park. They thus enjoy a
consumers’ surplus from visiting, equal to the difference between the most they would pay
(per trip) and what they actually pay. By observing the relationship between visits and
travel costs, it might thus be possible to infer the value (consumers’ surplus) that recrea-
tionalists enjoy.
A simple travel-cost analysis would proceed as follows. Visitors to the national park would
be surveyed, and asked how far they had travelled to make this visit, and how often in the last
va: ECONOMIC TOOLS FOR THE ENVIRONMENT

12 months they had visited the site. They might also be asked how many other similar sites
they had visited in the area, and about their income, recreational experiences, and family size.
In Figure 4.2(a), we show a possible relationship between the number of visits that individu-
als make to the park (V;) and the costs to them per visit (C;), where the C; are the costs of
driving to the park from their home. For example, visitor V, faces a cost per trip of €25 and
makes only two trips per year, whilst visitor V, faces a cost of €12 per trip and makes six trips.
If responses from, say, 1,000 visitors were collected, then this would probably allow us to
estimate such a curve with reasonable precision. The curve in Figure 4.2(a) is really a demand
curve for the site since it shows, at any price, how many trips are taken. Since the area under
a demand curve shows total value, we can measure the area under travel-cost curves and use
these to measure total value, or, more usually, consumers’ surplus per visit.
We show visually how this could be done in Figure 4.2(b). Imagine that we select one indi-
vidual (Begonia) from the sample. Begonia currently faces a travel cost of €18 per visit, and
makes three trips per year, shown as V*. However, the value she places on each visit is more
than €18, so currently she enjoys a consumers’ surplus. Using the average relationship
between visits and travel costs in Figure 4.2(a), it is possible to see what would happen to the
number of trips that Begona would make if the cost to her was increased; for example, if a
hypothetical admission fee to the park were introduced. At an admission fee of A,, her trips
fall to V,; whilst at a higher admission fee of A,, they fall further, to V,. By repeating this exer-
cise, we could trace out the function shown originating at V* and passing through the points
(A,, V,) and (A), V,). The area under this curve and above the horizontal line originating at
Ay (the current cost she actually faces) is the consumers’ surplus she enjoys from making
V* = 3 trips per year.
In practice, it is not necessary to trace out the kind of curve shown in Figure 4.2(b) to
calculate consumers’ surplus, since it can be calculated mathematically from the equation
relating actual trips to actual visits. This might be estimated as follows:

In(V,) =a—0bC, +8, (4.8)

where In(V;) is the log of visits per year per person, and € is an error term. This simple
version ofthe travel-cost model has a number ofdrawbacks, which we now investigate.

—fe¥)—

£25 f--------- V,

(CG)
Cost
trip
per £12 |---------
Vi=a-—bC,

we
ee
ee

No. of trips/year (V,)

Figure 4.2(a) The relationship between visits and travel costs.

VALUING THE ENVIRONMENT: METHODS 73

No. oftrips/year

Additional
trip
costs
per

3
No. oftrips/year

Figure 4.2(b) A derivation of consumers’ surplus from the visit-cost relationship.

Substitute sites. The simple travel-cost model discussed so far was presented in terms of
predicting visits at one single site. Suppose, however, that there are many similar sites within
reach of at least part of our sample of visitors. Clearly, the more alternatives an individual
has, the less likely he or she is to visit the site that is being modelled. Moreover, we might
want to estimate demand across a set of sites, which all differ in some respects from each
other (e.g. fishing rivers with different expected catches, or different types of fish or acces-
sibility). Two approaches may be noted. Most simply, if we are still interested in modelling
demand for one site, we could allow for the influence of other sites by including the costs of
trips to these other sites in the travel-cost equation. However, this approach suffers from the
fact that the alternative sites may be quite different from the site we are interested in respects
other than just travel costs. A better approach has been to more explicitly model site choice
using random utility models. These adapt a probabilistic approach, and predict the proba-
bility that an individual will visit a given site out of a list of alternative sites, as a function of
the attributes of this site relative to those of the alternatives: for example, which river(s) a
person will choose to visit out of all salmon rivers in Ireland. Random utility models can be
combined with count models, which estimate the number of trips that a person will make to
all sites of this general type—all salmon-fishing streams in Ireland—in a year (Martinez-
Espineira and Amoako-Tuffour, 2008). For an example of random utility and count models
in use, see Johnstone and Markandya (2006). Box 4.4 shows an example ofarandom utility
travel-cost model of whitewater kayaking.
The value of travel time. What is the monetary value of leisure time? Above, we argued
that since time is scarce, then using it up in travelling to a recreational site has an opportu-
nity cost. But what is this equal to? This depends very much on individual circumstances.
Imagine the case ofa self-employed carpenter who likes to go fishing. Every hour he spends
fishing is one less hour spent working. In this case, the value of his leisure time is equal to
his hourly earnings in work or, more generally, to the wage rate. For people who are giving
up working opportunities in order to pursue leisure at the margin, then their wage rate
74

Whitewater kayaking is an invigorating sport, which is typically unpriced in that access to rivers is free.
However, kayakers incur travel costs in getting to access points, and these costs enable a travel-cost
analysis to be undertaken. Moreover, the fact that kayakers can usually choose between several different
rivers, with differing site qualities and differing travel costs from their homes, means that a random utility
site choice travel-cost model is particularly appropriate. \
Stephen Hynes and co-authors (2009) questioned 279 kayakers in Ireland about their recreational
behaviour. For each person, trips in the last 12 months to eleven different rivers were recorded. Each
person then rated each river in terms of anumber of attributes that described their kayaking experiences.
These attributes included average quality of parking at the site, average crowding, the quality of the
kayaking site as measured by the star rating system used in the Irish Whitewater Guidebook, water
quality, scenic quality, and travel time from home. A travel-cost measure was constructed for each
respondent for each river. People were asked to rate their levels of skill and experience as a kayaker, and
these variables were used to estimate separate travel-cost models for different skill levels, on the grounds
that highly skilled kayakers might have different preferences for site attributes than those just beginning
with the sport. The following table shows the results from the random utility travel-cost models,
comparing a generic (all observations) model with two skills-based models—dummy variables were
included for each site apart from the most visited:

Variable All kayakers Skill level 12 Skill level 22 RUM with skill
interaction
dummies
Travel cost —0.069 —0.099 —0.059 —0.092
Ngee (14.15)* (11.74)* (i531)
Quality of parking —0.145 —0.089 —0.22 —0.063
(2.04)* -0.71 (2.16)* —0.5/
Crowding 0.153 0.172 0.129 014
CAs -1.51 —1.39 —1.34
Star quality of the white- 0.351 0.163 0.488 -0.214
water site (232) —0.82 (2.86)** —1.36
Water quality 0.142 0.24] 0.397 0.042
—1.39 —1.45 (2.87)* 0.32
Scenic quality 0.285 0.492 0.107 0.199
(2.99)" (3.22)* —0.84 —1.55
Availability of information —0.08 =(:311 0.178 —0.067
on water levels -0.92 (2.19)* —1.52 —0.57
Advanced skill*Travel cost 0.028
(4.22)*
Advanced skill*Quality of —0.198
parking —1.47
Advanced skill*Crowding 0.046
=0135
Advanced skill*Quality of the 0.984
whitewater site (5.84)
75

re re a a ee

Variable All kayakers Skill level 18 Skill level 22 RUM with skill
interaction
dummies
Advanced skill*Water quality 0.137
—0.95
Advanced skill*Scenic quality O127
—0.88
Advanced skill*Availability of 0.068
information on water levels =05
Clifden Play Hole 0.905 0.304 —1.643 =0.737
(2.47)* —0.54 18)" Cesias
Curragower Wave on the Sail eal)=) =1141 1-586 Sees
Shannon (5.34)* (2.89)* (4.10)** (4.94)
The Boyle =W772 —1.586 —1.864 =1:715
5.93)" S51 (4.41)" (5:63)7%
The Roughty —1.641 = 107 =1.397 —-1.432
(4.10)** 200 )on (2.51) 3.48)**
The Clare Glens —3.387 —4.224 —2./34 —3.243
(8.63)"* oa) (4.99) 0.08)**
The Annamoe =2,076 =1,/87 =2,105 —1.888
(6.25)% 3.58) 4.47)" BS)
The Barrow 2.914 —2.408 Soils —2.806
OEM DEON 70 8.86)*
The Dargle —5.011 6.195 —4 303 —4.935
(1233 (8.96)** E8iy* (SO
The Inny =1./69 —0,892 =2:393 —1.684
6.04)" (2.07)* BIO 5,70)"
The Boluisce (Spiddle) —2.344 —1.437 =2 399 e225
6.96)** (2.81) (6.06)** 6.57)*
Notes: Absolute value of z statistics in parentheses; * significant at 5%; ** significant at 1%. Models CL1, CL2, and CL3
have log likelihood values of -913.95, -358.22, and -447.78, respectively.
@ Skill level 1 refers to kayakers who have basic and intermediate proficiency level kayak handling skills.
Skill level 2 refers to kayakers who have advanced proficiency level kayak handling skills.

These results were then used to calculate how consumers’ surplus per trip would change if the
attributes of rivers were to change, or if access to certain rivers was curtailed. This showed that allowing
for differences in the preferences of kayakers according to their skill levels made for differences in
consumers’ surplus estimates. For instance, loss of access to one river (the Roughty) where a
hydroelectric scheme had been proposed would impose a cost of €5.97 per kayaker per trip in a model
where no account was taken of different skill levels, and €7.72 per trip where skill level variations are
included. Kayakers incur these losses in utility since loss of access to this popular river reduces their
available choice set.
76 ECONOMIC TOOLS FOR THE ENVIRONMENT

measures the value of their leisure time. However, for many people, this is not a good
description of the situation. For example, unemployed workers, parents staying at home,
and retired people do not give up work for leisure at the margin. This is also true for the
majority of workers, who tend to be on fixed-hours contracts. For example, if Fanny is a
schoolteacher, then she is unlikely to be giving up earnings (at the margin) during her holi-
days in order to go fishing. In such cases, where people are not making marginal wage/
leisure choices, it is harder to know how to value leisure time.
Most research in this area has used either stated- or revealed-preference approaches to
estimate leisure time values in contexts where people choose between faster, more expen-
sive routes and slower, cheaper ones. These studies reveal that leisure time values are posi-
tively related, on the whole, to income, so that some fraction of the wage for an individual is
a reasonable guess as to the monetary value of their leisure time. However, the valuation of
leisure time is still an inexact science. This is problematic for models of the travel-cost type,
since consumer surplus estimates for recreation have been shown to depend on which value
for leisure time is used (Hynes et al., 2009).
Use values versus non-use values. Finally, it should be obvious that since the travel-cost
approach infers values from expenditure, those who make no such expenditures have no
inferred valuation for the good. If all values were use values, this would not be such a prob-
lem. However, if there are non-use values associated with an environmental resource such
as a national park, then travel-cost approaches cannot pick up such values, since they can
accrue to people who do not visit the site.

4.4 Production-function Approaches

In production-function approaches, the environment is typically valued as an input to the
production of some market-valued good or service (Barbier, 2007). Changes in the quality
or quantity of an environmental resource are valued by estimating the implications of this
for outputs and prices of market goods/services, usually in terms of changes in consumers’
surplus and producers’ profits. This class of methods includes dose-response models, which
are used to study the impacts of pollution on market-valued outputs (e.g. the effects of
increases in low-level ozone on agricultural crops). More recent terminology discusses ‘eco-
system service valuation’ models, but the basic approach is similar.
To take a simple example, consider the role that coastal wetlands play in fisheries.
Wetlands provide important nursery areas for fish and shellfish, including crabs. Ellis and
Fisher (1987) modelled the case of wetlands off the gulf coast of Florida. Human effort (in
terms of numbers of traps set) and wetland acreage were found to jointly determine the
output of blue crabs. Having identified this production function, the authors then found the
profit-maximizing levels ofeffort for different levels of wetland acreage. It was then possible
to see the effects on welfare, measured as changes in consumer and producer surplus, of
reductions in wetland area. A similar analysis was conducted by Barbier and Strand (1998),
who study the link between conserving mangrove forest in coastal areas in the Gulf of
Campeche, Mexico, and shrimp fishing. Mangroves act as a nursery for shrimp. As man-
groves are lost due to conversion to aquaculture or hotel developments, shrimp population
dynamics are impacted: an economic model ofthe fishery thus needs to be considered to
VALUING THE ENVIRONMENT: METHODS TL

Table 4.2 Costs of mangrove losses in Thailand, 1996-2004

Basis for calculation Economic costs of mangrove loss in terms of

impacts on coastal fishery (US$)
Static model: annual welfare loss 99,004
Static model: net present value over the whole period, 570,167
discount rate = 10%
Dynamic model, net present value, discount rate = 10% 1,980,128

Source: Barbier (2007)

estimate the change in producers’ surplus from lost biomass production. Barbier and Strand
calculate the economic cost of each kilometre squared of mangroves lost, in terms of lost
profits to shrimp fishers, of $86,345-$153,300 per year. Barbier (2007) also calculates the
costs to consumers and fishers from historic losses of mangroves in Thailand over the period
1996-2004. These are shown in Table 4.2.
Table 4.2 shows that, for an annual average loss rate of 18 km?, the annual costs measured
from a static model of the fishery are nearly $100,000. Over the 9 years within which data for
mangrove losses exist, this translates into a net present value of losses of $570,167 at a 10 per
cent discount rate. However, the static model does not take into account the impacts of the
loss in mangroves on the dynamics of the fish populations, and the responses of fishers.
Once these dynamic effects are‘allowed for, losses become much greater. Coastal wetlands
could also have value as an input to storm and flood protection, by acting as a natural bar-
rier. If so, then additional values for protecting coastal wetlands exist, in terms of the
expected avoided property and human life costs from flooding and storms.
The economic values of the environment as an input to production are also studied in the
context of dose-response models. These have been very widely used in the literature, mainly
to study the effects of air pollutants on agricultural crops and forests. Polluting emissions
from a variety of sources are modelled in terms of their fate in the environment. For exam-
ple, for SO,, this might involve being blown from one country to another and then falling as
acid rain. This constitutes the ‘dose’ of pollution, to which a physical response occurs. This
physical response (e.g. crop damage, buildings damage, fisheries impacts) is estimated using
models produced by natural scientists. An economic response to such damage can then
occur; for example, if farmers decide to grow less pollution-sensitive crops. Finally, the net
effects on output, consumers’ surplus, and producers’ profits are evaluated, along with other
elements of damage costs (e.g. the costs of restoring old buildings damaged by acid rain).
More recently, economists have considered the costs of climate change in terms of such
models. Here, climate is an ‘input’ to the production of agricultural crops. For example,
work by Bateman et al. (2010) for the United Kingdom suggests that wheat production is
significantly related to a number of climate variables, including growing season tempera-
tures and rainfall. This statistical model is used to simulate the effects of a 1 degree rise in
mean temperatures on wheat production and the market value of outputs. The model allows
us to see how the response is highly variable across the UK—wheat yields rise in some areas
and fall in others, as farmers switch to more profitable activities. Figure 4.3 shows this pre-
dicted spatial variation in output, and how this would be valued using market prices for
78 ECONOMIC TOOLS FOR THE ENVIRONMENT

[A Cereals (ha/grid) A FGM (£/ha)

| Gili = -9 orless @=s =-10orless
3 =-9to-3 C3 =-10to+10
ES =-3 to+3 C3 =+10to +30
) | Be =+3to+9 a = +30to +60
wes = +9 or more mi = + 60 or more
=

Figure 4.3 Predicted changes in agricultural outputs and market values for a 1 degree rise in mean
temperatures in the UK.
Source: Bateman et al. (2010).

inputs and outputs (although this would not reflect changes in net social benefits, since
there are externalities associated with agricultural production).
The above examples of valuing the environment as an input relate closely to the idea of
valuing ecosystem services, which was discussed in Chapter 3. When these ecosystem ser-
vices result in outputs valued by the market, then production-function methods are appro-
priate for valuing changes in such services. But stated- and revealed-preference methods
also have a role to play here. Table 4.3 summarizes the range of methods that economists
might use to value ecosystem services for the conservation of moorlands.

45 Benefits Transfer

Benefit transfer (BT) is the practice of extrapolating existing information on the non-market
value of goods or services (Brouwer, 2000). Typically, the practice involves predicting WTP
values for an environmental quality or access change at one site (location), based on data
collected using either stated- or revealed-preference methods at another, similar site. For
example, we may want to predict the value of improvements in water quality on the River
Trent in England, based on stated-preference data for the River Tyne (another river in
England). Adjustments are often made for differences between the environmental charac-
teristics of the site to which values are to be transferred (known as the ‘policy site’) and those
of the site at which the original data was collected, known as the ‘study site’. Differences in
VALUING THE ENVIRONMENT: METHODS

Table 4.3 Valuing ecosystem service flows in moorlands

—_—:-:e—rr CO eee

Ecosystem service _— Possible change in this service Appropriate valuation methods

Carbon Net change in carbon flux due to change in Market prices for CO, from carbon
sequestration land management markets
Informal recreation: Fall or rise due to loss of moorland to forests Travel-cost models
birdwatching
Landscape quality Change due to change in frequency of heather — Stated-preference approaches
burning for grouse shooting; loss of landscape contingent valuation or choice
features due to wind farm construction experiments)
Water quality Decrease in water quality due to loss of peat Production-function approach: addi-
ional costs to water supply companies
ower down the catchment
Water storage Decrease in water supply due to forest Production-function approach: costs
planting of measures taken to augment supplies
rom other sources

some of the socio-economic characteristics of the affected population between the study
and policy sites are also allowed for (Morrison et al., 2002).
The aim of BT techniques is to provide decision-makers with a monetary valuation of
environmental goods and services in a cost-effective and timely manner, since original
valuation studies are both expensive and time-consuming. Demands for environmental
valuation estimates are rising in the policy community in both Europe and the USA. In
Europe, this is partly being driven by the introduction of the Water Framework Directive,
which requires benefit-cost analysis of water-quality improvements throughout the
European Union, and by the greater emphasis on the application of cost-benefit principles
to environmental policy design in the EU. Papers investigating the use and accuracy of BT
have become increasingly frequent. Applications of BT include Rozan (2004) on improved
air quality in France and Germany, Muthke and Holm-Mueller (2004) on national and
international transfers of water-quality improvement benefits, Jiang et al. (2005) on coastal
land management, and Colombo and Hanley (2008) on agricultural landscapes.
Two main approaches have been followed in the literature. The first is the transfer of mean
WTP values from the policy site to the study site. The transfer of unadjusted mean values has
been criticized, since it does not take into account any possible differences between either
the populations or the goods at the policy and study site. Because of that, an alternative
adjusted mean value approach has developed, which adjusts the mean WTP of the study site
to account for differences in the environmental characteristics of the policy site and/or for
differences in the socio-economic characteristics of the affected population between the two
sites. In the case of the adjusted value transfer, the WTPs are adjusted using data on the
socio-economic and environmental characteristics of the policy site, before the comparison
takes place. Such adjustments are, to a varying degree, somewhat ad hoc.
The second approach to BT is benefit function transfer, where the entire demand func-
tion (or choice equation, in a CE setting) estimated at the study site is transferred to the
policy site. WTP values at the policy site are predicted using independent variables (such as
household income) collected from secondary data at the policy site and parameter values
80 ECONOMIC TOOLS FOR THE ENVIRONMENT

estimated from the study site. In the benefit function transfer, the regression parameters of
the study site and the environmental and population characteristics of the policy site are
used to test the following equation:

predicted WTP (B°, X?) = WTP?, (4.9)

where the predicted WTP (B%,X?) is the WTP at the policy site estimated using the param-
eters of the benefit function of the study site (B°) and the X values (site attributes, socio-
economic characteristics, etc.) of the policy site, and WTP? is the WTP at the policy site.
When several study site data sets are available, a further approach is to use a meta-regression
analysis. Here, the analyst is concerned with understanding the influence of methodological
and study-specific factors on WTP. Data can be pooled across study sites to produce a BT
model for predicting policy site values. Which of these benefits transfer testing approaches
is preferable is still open to debate. Economists are still refining how benefits transfer is best
carried out, and how to test its accuracy.

Tutorial Questions
4.1 What are the main differences between stated- and revealed-preference methods of
environmental valuation?
4.2 List the three main problem areas of contingent valuation as you see it; and explain
how researchers could try to minimize the impacts of these problems on estimates of
environmental benefits that they have been asked to produce.
4.3 Explain how you would undertake (i) a choice experiment and (ii) a travel-cost study
of the benefits of improved management of a national park. What information could
such studies generate that would be useful for national park managers?
4.4 What is ‘benefits transfer’, and how and when should it be undertaken? How accurate
could we expect such transferred valued to be?
4.5 What degree of reliance can we place on estimates of environmental value derived
from the methods described in this chapter? Are some methods more robust or more
valid than others?
4.6 Explain how you would estimate the change in ecosystem service values from a
decision to conserve an area of coastal mangrove wetlands, rather than allowing a new
harbour development to proceed.
4.7 What role does the discount rate play in environmental valuation?

References and Additional Readings

Barbier, E.B. (2007). ‘Valuing ecosystem services as Ecosystem Service Assessments Working Paper
productive inputs, Economic Policy 22: 177-229. EDM 10-10, CSERGE.
—— and Strand, I. (1998). ‘Valuing mangrove- Birol, E., Karousakis, K., and Koundouri, P. (2006).
fishery linkages: a case study, Environment and ‘Using choice experiment to account for
Resource Economics 12(2): 151-66. preference heterogeneity in wetland attributes:
Bateman, IJ, Mace, 'G.M,, Fezzi; C., Atkinson, G. the case of Cheimaditita wetland in Greece’
and Turner, R.K. (2010). ‘Economic Analysis for Ecological Economics 60: 145-56.
VALUING THE ENVIRONMENT: METHODS 81

Birol, E., and Koundouri, P. (2008). Choice validity for policy analysis. Environmental and
Experiments Informing Environmental Policy Resource Economics 31: 477-99.
(Cheltenham: Edward Elgar).
Johnstone, C., and Markandya, A. (2006). “Valuing
Brouwer, R. (2000). ‘Environmental value transfer: river characteristics using combined site
state of the art and future prospects, Ecological choice and participation models; Journal of
Economics 32(1): 137-52. Environmental Management 80(3): 237-47.
Colombo, S., and Hanley, N. (2008). ‘How can List, J., Sinha, P., and Taylor, M. (2006). ‘Using choice
we reduce the errors from benefits transfer? experiments to value non-market goods and
An investigation using the choice experiment services: evidence from field experiments, Advances
method) Land Economics 84(1): 128-47. in Economic Analysis and Policy 6(2): 1-37.
Day, B., Bateman, I., and Lake, I. (2007). ‘Beyond MacMillan, D., Hanley, N., and Lienhoop, N. (2006).
implicit prices: recovering theoretically consistent ‘Contingent valuation: environmental polling or
and transferable values for noise avoidance preference engine?’ Ecological Economics 60(1):
from a hedonic price model, Environmental and 299-307.
Resource Economics 37: 211-32.
Martinez-Espineira, R., and Amoako-Tuffour,
Ellis, G., and Fisher, A. (1987). ‘Valuing J. (2008). “Recreation demand analysis: an
the environment as an input, Journal of application to Gros Morne National Park, Journal
Environmental Management 25: 149-56. of Environmental Management 88: 1320-32.
Eshet, T., Baron, M.G., Shechter, M., and Ayalon, O. Mitchell, R., and Carson, R. (1989). Using Surveys
(2007). “Measuring externalities of waste transfer to Value Public Goods: The Contingent Valuation
stations in Israel using hedonic pricing, Waste Method (Washington, DC: Resources for the
Management 27: 614-25. Future).
Freeman, A.M. (2003). The Measurement of Morrison, M., Bennett, J., Blamey, R., and Louviere,
Environmental and Resource Values (Washington, R. (2002). ‘Choice modelling and tests of benefits
DC: Resources for the Future). transfer, American Journal of Agricultural
Haab, T., and McConnell, K. (2002). Valuing Economics 84(1): 161-70.
Environmental and Natural Resources Mourato, S., and Maddison, D. (2000). “Valuing
(Cheltenham: Edward Elgar). different road options for Stonehenge; in
Hanley, N., and Barbier, E.B. (2009). Pricing Nature: S. Navrud and R. Ready (eds.), Valuing Cultural
Cost-Benefit Analysis and Environmental Policy Heritage (Cheltenham: Edward Elgar).
(Cheltenham: Edward Elgar). Murphy, J., Allen, P., Stevens, T., and Weatherhead,
Heberlein, T.A., Wilson, M.A., Bishop, R.C., and D. (2005). ‘A meta-analysis of hypothetical bias
Schaeffer, N.C. (2005). ‘Rethinking the scope in stated preference valuation Environmental and
test as a criterion for validity in contingent Resource Economics 30(3): 313-25.
valuation, Journal ofEnvironmental Economics Muthke, T., and Holm-Mueller, K. (2004). “National
and Management 50(1): 1-22. and international benefit transfer testing with
Hensher, D., Rose, J., and Greene, W. (2005). a rigorous test procedure’, Environmental and
Applied Choice Analysis: A Primer (Cambridge: Resource Economics 29: 323-36.
Cambridge University Press). Poe, G.L., Clark, J., Rondeau, D., and Schulze,
Hynes, S., Hanley, N., and O’Donoghue, C. W.D. (2002). ‘Provision point mechanisms
(2009). ‘Alternative treatments ofthe cost and field validity tests of contingent valuation,
Environmental and Resource Economics 23: 105-31.
of time in recreational demand models: an
application to whitewater kayaking in Ireland’, Ready, R., Champ, P., and Lawton, J. (2010).
Journal of Environmental Management 90(2): ‘Hypothetical bias in a stated choice experiment,
1014-21. Land Economics 86(2): 363-82.

Jacobsen, J.B., and Hanley, N. (2009). ‘Are there Rozan, A. (2004). ‘Benefit transfer: a comparison
income effects on global willingness to pay for of WTP for air quality between France and
biodiversity conservation?’ Environmental and Germany, Environmental and Resource Economics
Resource Economics 43: 137-60. 29: 295-306.

Jiang, Y., Swallow, S.K., and McGonagle, M.P. (2005). Taylor, L. (2003). “The hedonic method, in P. Champ,
‘Context-sensitive benefit transfer using stated K. Boyle, and T. Brown (eds.), A Primer on Non-
choice models: specification and convergent Market Valuation (Dordrecht: Kluwer).
Environmental Risk and
Behaviour

Environmental protection is like a risky lottery. We invest in protection, but what we really
are doing is creating a new lottery. This new lottery exists because environmental policy
rarely changes the world so that people are 100 per cent safe from risks to their health, or
risks to the environment. Rather, we make environmental policy to reduce the current risk
in favour of a new lower level of risk—a new lottery with better odds of good health and a
clean environment. When thinking about environmental policy and economics, it makes
sense to take the time to better understand how people and policy-makers make decisions
under risk and uncertainty, both in theory and in practice.
In theory, policy-as-lottery implies that people think simultaneously about the probabili-
ties and consequences that define the risks to human and environmental health. People
know that risks exist in their daily lives—they think about how to invest resources to reduce
risks to their heath (e.g. water filters, seat belts, sunscreen, and so on). We think about cost-
effective environmental policy that reduces risks to health by suggesting ways to ‘fix’ market
failures. Advice about policies that create new markets or market-like incentives rests on a
model that presumes that people facing new incentives will act with purpose and make
consistent choices that take into account the consequences of their choices. This model is
called rational choice theory—people are assumed to make consistent decisions that are in
their own best interests within the context of market exchange.
Using rational choice theory to help guide environmental policy makes sense if people
make, or act as if they make, consistent and systematic choices when thinking about risk. In
practice, however, people struggle with decision-making under risk; instead, we use simple
rules and heuristics to make choices. Over the past half century, many empirical studies
have documented how rational choice might, in some circumstances, be a poor guide for
how people react to environmental risk (see Kahneman, 2011). People make choices that
are inconsistent—risk-averse people take a gamble to avoid a loss, we resist change, we
overestimate small risks, we think in discrete bundles, we value today more than the future
but we are inconsistent, we care about others even in an economic context, and our good
intentions can be tainted by money. We also know that the context of choice matters: who
gives us the information, social and cultural norms, the default choice and status quo
reference point, what draws our attention—uniqueness, access, simplicity—how we are
ENVIRONMENTAL RISK AND BEHAVIOUR 83

subconsciously primed to make certain choices, emotional responses to goods and informa-
tion, the degree of commitment to overcome bounded willpower, and ego/self-image (see
Metcalfe and Dolan, 2012).
This ‘anomalous behaviour’ and context-dependent choice arises in many private and
public decisions, and serves to undercut the underpinning of rational environmental
economics and policy (see Shogren and Taylor, 2008). Behavioural economics is the field
of research that attempts to explain systematic deviations from rational choice.
Behavioural economists apply psychological insight to help reshape economic principles
more towards observed behaviour. By ‘reshape’, we mean adding more humanity to
rational choice theory. The end goal is to make environmental policy more efficient by
better matching incentives and behaviour (see Thaler and Sunstein, 2008).
In environmental policy, the classic example of energy efficiency illustrates how ‘behav-
ioural anomalies’ can affect behaviour. An ‘energy paradox’ is said to exist when people
buy less energy conservation than predicted by a present-value calculation given, say, a
tax on carbon. Behavioural anomalies that could explain this result include people dis-
counting the future too highly, people who have trouble calculating expected fuel savings,
people who focus too intensely on the status quo, and people who rely on heuristic deci-
sion-making strategies rather than optimizing net benefits. But all these ideas have rarely
been tested within the same experimental design. They are a collection of ideas, in which
the policy-maker does not know which effect, ifany, dominates choices of energy conser-
vation, and why this particular effect is the key (see Gillingham et al., 2009). Policy options
in such cases are limited to more education, information, and standard setting. If people
are not responding rationally to pricing changes, green taxes will not have the intended
consequences, either in efficiency or distribution of burden (Brekke and Johansson-
Stenman, 2008).
In this chapter, we explore environmental risk and behaviour, in theory and in practice.
We examine how people behave in response to current levels of risk to human and environ-
mental health, and how we respond to changes in this risk—privately and publicly. By risk,
we mean the combination of two elements—the chance that a bad event might happen, and
the consequences that are realized if a bad event does occur. Although actions to improve
our lives are not intended to create risk to others and ourselves, we do generate pollution
and accidents do happen. Cars pollute. Oil spills. Technology fails. How and when these
things happen define the lotteries that we create in a modern economy. Our actions gener-
ate rewards—new medicines, new transportation, new communication systems, and new
risks; for example, Chernobyl, or Fukushima Daiichi. Today, many observers fear that our
choices have put humanity at greater risk due to biodiversity loss and climate change.
Since we cannot achieve a zero-risk society, we make trade-offs based on risks and
rewards. For instance, few would pay the necessary price to reduce transportation risks to
zero. We could set the speed limit at 10 miles/16 kilometres per hour for all roads and high-
ways. No one would ever die from an automobile accident. But we will never see such a rule.
The costs per life saved from driving slow are too expensive—people trade off risk for higher
speed limits. We make similar risk-reward trade-offs in our jobs, recreation, and lifestyles.
Understanding the nature of health, safety, and environmental risk, how people behave
towards risk, and how to identify effective strategies to manage risk is essential for better
public policy towards the environment.
84 ECONOMIC TOOLS FOR THE ENVIRONMENT

5.1 Rational Behaviour under Risk

We begin by defining the rational choice model of decision-making under risk. Assessing
risks depends on how people react to the risks that they face. Since the Renaissance, econo-
mists and others have been developing a systematic theory of risk and behaviour. As people
gained an understanding of how to use trade to inctease value, they began to address how to
master risks connected to both finances and health. More trade meant more wealth and
more risk. Trading partners separated by oceans had an incentive to understand how to
manage and control risk. As trade routes turned into world wars and global stock and bond
swaps, the gains from risk assessment and management as practical arts increased. Those
who had an appreciation of the behavioural underpinnings of risk had a better chance of
winning real and metaphorical battles. This holds for environmental risk too.
To follow the intellectual history of understanding how people behave under risk, con-
sider three gambles:

e Gamble X is a certain payment to you of $30—a sure bet.

e Gamble Y isa coin flip in which you win $100 with a heads, and lose $100 with a tails.
e Gamble Z isa roll of a dice in which you win $2,000 if a 1 is rolled; win $1,000 with a 2;
win $500 with a 3; lose $0 with a 4; lose $1,000 with a 5; and lose $2,000 with a 6.

Which would you choose? Early theorists speculating on how people make choices under
risk first argued that people would prefer the gamble with the highest expected value—the
probability weighted average of all possible outcomes of the gamble:

Expected value = (probability of outcome 1) x ($ of outcome 1)

+ (probability of outcome 2) x ($ of outcome 2)
+ (probability of outcome n) x ($ of outcome n)

In our case:

e Gamble X: 100 per cent chance of $30 = $30 expected value.

e Gamble Y: 50 per cent chance of $100 + 50 per cent chance of -$100 = $0 expected
value.
e Gamble Z: [1/6] chance of $2,000 + [1/6] chance of $1,000 + [1/6] chance of $500 +
[1/6] chance of $0 + [1/6] chance of —$1,000 + [1/6] chance of —$2,000 = $83.33
expected value.

Gamble Z has the highest expected value. But we observe that many people take gamble X
instead. ‘The old adage that a bird in the hand is worth two in the bush reflects the prudent
strategy to go for the sure thing.
But why does the gamble with the lower expected value attract so many people? In the eight-
eenth century, Nicholas Bernoulli created the St Petersburg paradox to show why. Suppose
that you are offered the following proposition. You can buy into a gamble on a fair coin toss.
Ifa head comes up on the first flip, you earn $2; ifit takes two flips to uncover a head, you earn
ENVIRONMENTAL RISK AND BEHAVIOUR 85

$4; if it takes three flips, you earn $8; four flips, $16; five, $32; six, $64; seven, $128; and so on.
What is the maximum you would be willing to pay to buy into this gamble?
Everyone would answer with an amount lower than infinity. But infinity is the expected
value of this gamble:

Expected value = [1/2] chance of $2 + [1/4] chance of $4 + [1/8] chance of $8 + ---

l| $1+$1+$1+$1+ $1+.---
= infinity.

Should you have been willing to pay more? No. Why? One reason is because the variance of
the gamble is also infinite. Variance is considered synonymous with risk because it reflects
the potential volatility of the outcome. The variance reflects the distribution around the
expected value. More variance implies more chance that bad states—low pay-off outcomes—
will be realized.
Nicholas’ cousin Daniel Bernoulli offered a reason why people pay less than infinity for a
gamble with infinite variance: a gain of $2,000 is not worth twice as much as one of $1,000.
People have diminishing marginal returns to wealth. This means that even though you prefer
more money to less, the last dollar you earn gives you less satisfaction than the first dollar
earned. His key insight is that a person’s ‘utility’ (the degree of satisfaction in possessing wealth)
resulting from a small increase in wealth will be inversely proportionate to his current wealth.
Figure 5.1 illustrates the idea of diminishing marginal returns. Increased wealth increases
total utility at a decreasing rate, which is why the utility function is curved. Gambles with
high variance are less attractive—your gain from an extra dollar added to your wealth is
smaller than your loss from an extra dollar taken away. One example of a useful utility func-
tion with this property is

u(w) = square root(wealth) or u(w) = Jw.

Utility

Wealth

Figure 5.1 Diminishing marginal returns.

86 ECONOMIC TOOLS FOR THE ENVIRONMENT

For instance, wealth of $10,000 creates a utility level of 100, while wealth of $1,000,000 cre-
ates a utility level of 1,000, so that a hundredfold increase in wealth increases a person’s
utility by tenfold. When a person acts in this way, we say he or she is risk-averse. A risk-
averse person is more likely to take a certain pay-off over a fair bet—a gamble in which the
expected value is zero; for example, 50:50 odds to win or lose $1,000. Our person is risk-
loving if he or she prefers a fair bet to a certain pay-off equalling the expected value of the
gamble. He is risk-neutral if he is indifferent betweén a gamble and certain pay-off equalling
the expected value of the gamble.
Bernoulli’s insight was formalized into a rational choice framework called expected utility
theory (EU). Since its introduction in the 1940s by the mathematician John von Neumann
and the economist Oscar Morgenstern, EU theory has been the most successful model of
how people make decisions under risk. The formal theory of expected utility reflects the idea
that people make choices about risk based on their beliefs about the probability that good
and bad events will be realized, the consequences of good and bad events, and the utility or
satisfaction a person gets from the consequence that is realized.
Consider our three gambles from an expected utility perspective. Assume that a person
has an initial wealth of, say $2,001, and a square root-style utility function:

e Gamble X: 100 per cent chance of $2,031 = 45.1 expected utility.

e® Gamble Y: 50 per cent chance of ./$2,101 +50 per cent chance of $1,901 = 44.7
expected utility.
e Gamble Z: [1/6] chance of ./$4,001 + [1/6] chance of $3,001 + [1/6] chance of
¥ $2,501 + [1/6] chance of $2,001 + [1/6] chance of ./$1,001 + [1/6] chance of
$1 = 40.9 expected utility.
On the basis of expected utility, the person should prefer gamble X to either Y or Z: 45.1 >
44.7 > 40.9. This is the insight of expected utility theory—expected values and expected util-
ity do not rank alternative gambles in the same way.
Now consider an example of environmental risk. Travis is a carpenter who lives and works
in Las Vegas, which has elevated levels of air pollution. He earns $50,000 per year. But he is
uncertain as to whether he will stay healthy given his ongoing exposure to ozone and particulate
matter. Suppose he believes that one of two states of nature will be realized—a good and bad
state. In the good state, he stays healthy and earns $50,000. In the bad state, he gets sick and earns
$35,000 (= $50,000 — $15,000 in medical expenses and lost work). Let and (1 — 2) represent
his beliefs that either the good or bad state will be realized. Suppose he thinks that the odds that
he will stay healthy are 70:30; that is, 7=0.7 and (1 — 2) = 0.3. Travis is thinking about moving
to Baggs, Wyoming, which has both lower air pollution and lower wages. If he moves to Baggs,
he can earn $40,000 per year in the good state, and $38,000 in the bad state. He believes his odds
of staying healthy are 90:10 that he will not get sick due to air pollution (ie. 7 and (1 — 7”)),
We can now write Travis’ expected utility for both living in Las Vegas and living in Baggs
as the probability-weighted sum ofthe two potential states of nature:

EU(Las Vegas) = 2(w) + (1 — 27)u(w — D)

= (0.7)u($50,000) + (0.3)u($35,000)

= 0.7.7$50,000 + 0.3./$35,000
= 213;
ENVIRONMENTAL RISK AND BEHAVIOUR 87

EU(Baggs) = 2’(w) + (1 — 2’)u(w — D)

(0.9)u($40,000) + (0.1)u($38,000)

= 0.7./$40,000 + 0.3./$40,000
= 200.

With these odds and earnings, Travis’ expected utility is greater in Las Vegas even though
the odds of getting ill are greater, as 213 > 200. He is willing to trade off the greater health
risk for the higher earnings. This is common in many jobs in which workers take greater risk
for more pay.
Now suppose that his odds of staying healthy in Las Vegas drop to, say 50:50, holding
everything else constant. Now his expected utility from staying in Vegas falls to 205. But he
still prefers Las Vegas to Baggs, since 205 > 200. Travis would think about leaving Las Vegas
if the odds changed to 30:70, at which point his expected utility would drop to 198. Now
Baggs would be the more attractive option, as 200 > 198. For sure, numerous factors influ-
ence our decision on where to live, in addition to simply wages and the changes of getting
sick, such as school quality, crime, and recreation opportunities. We would have to adapt
the expected utility framework to account for these factors as well. This is more compli-
cated, but doable—as we saw in Section 4.3.1.
The next step in understanding rational choices within the expected utility model is to
account for a person’s ability to influence the risk that he or she confronts, either privately
or collectively, through alternative risk reduction investments. Travis is not as helpless
against the risk as the model suggests. He has more options than moving to Baggs. He can
purchase market insurance against illness. He can invest in different risk-reduction strate-
gies to change the odds of suffering from some illness due to the air pollution. He can buy
an air filter for his home, or he can eat better and exercise more. His actions can reduce the
likelihood that the bad state will occur or reduce the severity of the bad state if realized, or
do both. We refer to actions to reduce the likelihood of illness as self-protection or mitiga-
tion; and to actions to reduce the severity of a realized outcome as self-insurance or
adaptation.
Travis’ problem is now more complicated. He selects the level of self-protection and self-
insurance that balances the extra gains he gets from lower odds of illness and less severity
with the costs of protecting and insuring himself:

EU(Las Vegas | risk reduction) = a(z)u(w — z — x) + (l— a(z))u(w — D(x) — z — x),

where 7(z) is the probability of the good state, which depends on his level of self-protection,
z;and D(x) is the severity of illness, which depends on his level of self-insurance, x. Including
the private ability to reduce risk is helpful in order to understand choice under environmen-
tal risk, because these actions link risk assessment with risk management. One must account
for these actions to measure risk accurately and to manage risk effectively (Shogren and
Crocker, 1999).
Although risk assessment has amassed a useful record of estimating potential threats to
humans and nature, one problem permeates the risk assessment literature—the under-
emphasis on how people adapt to the risk that they face or have created. Over the past
88 ECONOMIC TOOLS FOR THE ENVIRONMENT

decade, scientists have acknowledged that environmental risk is endogenous. People can
influence many of the risks they confront. Examples abound. People move or reduce phys-
ical activities when air pollution becomes intolerable. They buy bottled water if they fear
that their drinking water is polluted, and they apply sunscreen to protect their skin from
UV radiation. A person can invest in a water filter, move, buy a membership to a health
club, jog, or eat food low in fat and high in fibre—each choice alters his or her risk to health
and welfare. How a person invests resources to increase the likelihood that good things
happen and bad things does not depend on both his or her attitudes towards risk and his
or her ability to reduce risk.
Cases exist in which people have little time to react to protect themselves, such as the
Chernobyl nuclear accident in the former USSR. While it can be argued that we can redefine
the problem so that risk is independent of human action, this approach is self-defeating.
Consider a situation in which bacterial groundwater contamination threatens a household’s
drinking water. The probability of illness among household members can be altered if they
boil the water. An analyst could define the situation as independent of the household’s
actions by focusing on groundwater contamination, over which the household probably has
no control. But this definition is economically irrelevant if the question has to do with the
household’s response to the risks from groundwater contamination. The household is con-
cerned about the likelihood of illness and the realized severity, and it is able to exercise some
control over those events. The household’s risk is endogenous because by expending its
valuable resources, it influences probability and severity.
People substitute private actions for collectively supplied safety programmes: the use of
stronger building materials to reduce the damage from tornadoes, storms, and earthquakes;
more thorough weeding and crop storage in response to the prospect of drought; sand-
bagging and evacuation in anticipation offloods; and improved nutrition and exercise regi-
mens to cope with health threats. At the policy level, these private risk-reduction choices
can affect the success of collective regulations that promote safety. The use of car seat belts
reduces both the probability and the severity of injury, but their mandatory installation can-
not guarantee that passengers will choose to wear them. Highway speed limits are also effec-
tive at reducing fatalities, when drivers observe them. At work, rules promoting personal
protective gear (e.g. hard hats) suffer from the same problem: they protect those workers
who wear them. In each case, individual decisions influence both the probability and the
magnitude of harm.
Endogenous risk implies that observed risks are functions of both natural science param-
eters and an individual’s self-protection decisions. Given the relative marginal effectiveness
of alternative self-protection efforts, how people make decisions about risk differs across
individuals and situations, even though the natural phenomena that trigger these efforts
apply equally to everyone. Assessing risk levels according to natural science parameters
only can be misleading. Relative prices, incomes, and other economic and social parameters
that influence any person’s self-protection decisions affect risk. Just as good public policy-
based economics requires an understanding of the physical and natural phenomena that
underpin choices, good public policy-based natural science requires an understanding of
the economic phenomena that affect risk. Accounting for private decisions can increase the
precision ofrisk assessment. Failure to acknowledge the depths ofprivate choice in environ-
mental risk will result in less environmental protection at a greater cost.
ENVIRONMENTAL RISK AND BEHAVIOUR 89

5.2 Behavioural Economics and Risk

We now consider how the rational choice model matches up with actual behaviour. To
illustrate, think how hazardous material conjures up images of a sanitized storage facility or
an abandoned toxic waste dump-site. The two images induce different perceptions of risk to
public health. Yet such a range of risk perceptions exists among people. Determining
whether an environmental risk needs to be regulated depends on how people are willing to
trade off risks for the benefits that flow from risk-generating activities. People’s willingness
to surrender benefits for reduced risk represents the value that they place on risk reduction.
Estimating this value for risk reduction is a critical component of risk—benefit analysis used
in policy-making on environmental risk.
This value of reduced risk depends on human behaviour—how people perceive risk and
their preference towards risk. People who are wary of risks are likely to value risk reduction
more than those who live to take risks. This statement seems straightforward enough, and
the logic behind it guides most economists who address environmental risk. Those at most
risk, who are most afraid of risk, and/or who have the most income should value risk reduc-
tion the most.
Economists who work with risk use the expected utility framework, which presumes that
people have well-defined preferences for risk and can form rational perceptions of risk. The
working presumption is that people have a solid foundation that drives their choices, such
that when they confront a risk, new or old, they are able to evaluate the odds and conse-
quences in a systematic and consistent way. A person’s stated value for risk reduction is
based on a logical foundation of choice from which we can judge the overall economic efh-
ciency of some policy decision. Without well-grounded preferences and perceptions, a
crack exists in the foundation of the rational theory of choice on which the economist’s
risk—benefit analysis rests (Hanley and Shogren, 2005).
Such cracks do exist. Psychologists and some economists have documented numerous
exceptions to the idea of a rational theory of choice (see Daniel Kahneman’s 2011 book
Thinking, Fast and Slow). These behavioural researchers have shown how people use rules
of thumb, or heuristics, to simplify their reasoning about risk. Using these rules, people
react to risk in broader patterns than predicted by expected utility theory. This suggests that
the standard model used to guide risk—benefit decisions is ‘too thin’, and does not predict
systematic aspects of behaviour under risk that are observed in many situations. The evi-
dence suggests that risk preference and perceptions seem to be influenced by the context of
choice (see Box 5.1).
People who make judgements about risk use heuristics that the popular expected utility
framework fails to capture. There is a long list of behavioural anomalies and paradoxes that
have been uncovered by cognitive researchers. One bias in judgement is when people over-
estimate low-probability risks and underestimate high-probability risks. Figure 5.2 illustrates
the bias. The 45-degree line represents the case in which the general public’s subjective risk
equals objective risk as defined by expert opinion. The flatter dashed line reflects the evidence
from different experiments and surveys examining how people rank the threats posed by dif-
ferent risks. People seem to inflate low risks over which they have little to no control (e.g.
nuclear power) and deflate high risks that they can control to some degree (e.g. driving to
work). They tend to worry more about how and where a risk arises than its magnitude; for
90

Several techniques have been used to construct counter-examples to expected utility theory. The most
common method involves obtaining an individual's response to a pair of choices designed to give
inconsistent answers. Allais (1953) provided the first counter-example with the following two pairs of
choices:

10% chance of $5 million

A: 100% chance of $1 million vs. B: 89% chance of $1 million
1% chance of $0

and

C: 10% chance of $5 million vs. D: 11% chance of $1 million

90% chance of $0 89% chance of $0

If the person maximizes expected utility, he must either prefer the pair (A, D) or the pair (B, C).
However, Allais and numerous other variations have observed that the modal number, and the majority,
of people prefer (A, C) (Machina, 1987). This suggests that expected utility theory is too thin a theory of
choice under risk—people are making systematic choices that are not captured by the theory.

example, synthetic versus natural carcinogens. This poor calibration between experts’ objec-
tive opinions and the laypersons’ perceptions can lead to constraints on choice, for example,
over whether or not to include nuclear power in a country’s energy portfolio.
Some risks are more acceptable than others. People who accept the risk of smoking or
driving without seat belts may not accept the risk associated with nearby treatment, stor-

risk
Subjective

Objective risk

Figure 5.2 Objective and subjective risk perception.

ENVIRONMENTAL RISK AND BEHAVIOUR 91

age, and disposal of hazardous material. Voluntary risks that people think they can control
are more acceptable than involuntary risks that they believe are outside their control.
Technologies that inhibit the sense that this risk is ‘voluntary—for example, nuclear
power—are less acceptable.
The perception gap raises a potential dilemma for the regulation. Suppose experts argue
that the risks from a certain product are unacceptable, while many people perceive the
opposite. Does the policy-maker ban the product or allow people to use their own discre-
tion? The 1997 beef-on-the-bone ban in the United Kingdom in the wake of‘mad cow dis-
ease’ (bovine spongiform encephalopathy, BSE), was a good example of not leaving the
decision to people. The policy-maker’s dilemma is to balance the trade-off between preserv-
ing individual freedom of choice and maintaining public safety. The policy-maker may be
tempted to step in and regulate the risk in the best interests of society. Such paternalistic
action, however, conflicts with societies committed to consumer sovereignty—the assump-
tion that a person is best able to judge what is in his or her own best self-interest.
Risk perception examines laypersons’ perceptions of risky technologies, and the deter-
minants of their relative acceptability. The majority of risk acceptance research has been in
the area of public perception of low-probability/high-consequence technology such as
nuclear power. Laypersons will not accept risk if the hazard is perceived as uncontrollable,
regardless of expert opinion. For example, in the 1980s, laypersons perceived nuclear
power as the number one risk to public safety, while experts ranked it twentieth—below
the risk of a household accident. Regardless of expert opinion, during the late 1970s and
the 1980s, Swedish citizens perceived the nuclear power risk as unacceptable, and policy-
makers agreed to phase out the entire industry within three decades. Today, the citizens of
Japan are reconsidering the future of nuclear power given the tsunami and the Fukushima
nuclear disaster.
Another risk perception effect is when people judge a risk by its familiarity with other
risks they have confronted. They base their decisions on how well what they know repre-
sents the new event. This tendency to stick with their prejudices can cause people to dis-
count risks with novel characteristics. One person underestimates the risk of radon because
it is an odourless and colourless gas; another person overreacts to the risk for the same rea-
son. Occasions exist when people fear the worst even when they have both good and bad
information, because we recall bad events first. People have alarmist reactions to well-
publicized risks to health or the environment.
People attach more weight to small losses than huge opportunities. We tend to dislike
losses more than we like the equivalent gains. This ‘loss-aversion’ idea suggests that people
treat perceived gains and losses differently. We seek out risk when gambles involve loss; but
we avoid risk for the equivalent gambles that involve gains. This evidence suggests that
rather than thinking about overall wealth, people seem to judge the value of gains and losses
from a status quo—a reference point. They judge risk by what they have experienced and
how it affects their current standing. The context in this case is the status quo. A person’s
value for risk reduction will then depend on the reference point, and on the nature of the
gains and losses of the lottery (Tversky and Kahneman, 1981).
In addition, how the risk is ‘framed’ affects choice in ways that are not predicted by
expected utility. If a person has well-formed preferences and values, a choice between
two options should be independent of how they are represented or described. But again,
92 ECONOMIC TOOLS FOR THE ENVIRONMENT

psychologists show how choice and values can be systematically influenced by different
ways of framing an identical problem. The importance of framing effects can be illus-
trated using with a famous example. Answer the following three questions:

Q.1: Which of the following options do you prefer?

A. A 100 per cent chance to win $30.
B. An 80 per cent chance to win $45.
Q.2: This is a two-stage game. In the first stage, there is a 75 per cent chance that the game
will end with no prize and a 25 per cent chance to move to the second stage. If you
reach the second stage, you have a choice between:
C. A 100 per cent chance to win $30.
D. An 80 per cent chance to win $45.
Your choice must be made before the game starts. Please indicate the option you prefer.
Q.3: Which of the following options do you prefer?
E. A 25 per cent chance to win $30.
F. A 20 per cent chance to win $45.

Questions 2 and 3 are identical in odds and rewards, and should produce identical choices
from a person. But, instead, people treat Q.1 and Q.2 the same, not Q.2 and Q.3. People
prefer option C in Q.2 and option A in Q.1, while they prefer option E in Q.3. This is the
so-called ‘certainty effect’—an option framed as a sure thing appears as attractive as a cer-
tain option. This suggests that framing matters to risk policy. Regulators should pay as
much attention to how they provide the information as to what information they provide.
People also place varying levels of trust in information on environmental risk according to
the source.
Many risks associated with the environment are hard to measure—the exact odds and
consequences are ‘ambiguous’—they cover a range of possibilities. Most people are averse
to such ambiguity in risk. Ambiguity implies that the probabilities of bad events are
uncertain—the odds of a bad event might range between | and 10 per cent. Ambiguous risks
dominate many decisions that people must make; for example, with regard to investments,
health care, exercise, and food. And while the expected utility model assumes that people
handle ambiguous risk, consider the following example. Suppose you had the choice
between choosing a ball from an urn with a known number of coloured balls and an urn
with an unknown number of coloured balls, in a pay-off situation. Most people prefer to
draw from the urn with the known distribution, even if the expected pay-offislower. This is
called the Ellsberg paradox. Even in experiments using professionals who deal with risk
every day—such as actuaries, business executives, and life insurance executives—people are
averse to ambiguous risk.
Finally, people find it a challenge to translate their preferences for risk into dollar values
when thinking about risky situations. Frequently, they ‘reverse their preferences’—that is,
they rank gamble A over B, but state a higher dollar value for gamble B than A. That is, they
prefer gamble A to B, but they assign a higher selling price to B than to A. To understand
better, answer the following questions:
ENVIRONMENTAL RISK AND BEHAVIOUR 93

Q.3: Which gamble do you prefer, A or B?

Gamble A: 35/36 chances to win $4.
1/36 chance to lose $1.
Gamble B: 11/36 chances to win $16.
25/36 chances to lose $1.50.
Q.4: Suppose that you own both gambles. What is your selling price for:
Gamble A: $
Gamble B: $

Decision theory based on expected utility requires that you be consistent—the gamble you
select (either A or B) should also be the gamble you put the highest selling price on. But
many people say they prefer gamble A to B, and then assign a higher dollar value to gamble
B than A. If you did so, you can add your name to the long list of people who ‘reversed their
preferences’.
The logical inconsistency is this: suppose that you preferred A to B, and valued A at $2
and B at $8. Now if your statements are a true indication of your preferences, we can turn
you into a ‘money pump in three easy steps: (1) sell you B for $8, (2) ask you to switch B for
A because, after all, you preferred A to B, and then (3) buy A back from you for $2. Now you
have neither gamble, and a $6 hole in your pocket (-$6 = $2 - $8). If you prefer more
money to less, this is not good. Your choices should be consistent, which is what expected
utility presumes. But the preference reversal phenomenon has been duplicated in numer-
ous settings, including with real gamblers in Las Vegas and economists retesting the work
of psychologists (also see Box 5.2 on inconsistent preferences over time; i.e. hyperbolic
discounting).
In conclusion, behaviour towards risk matters to environmental policy because if peo-
ple’s stated values for risk reduction are inconsistent with their underlying preferences, or if
the set preferences are a fantasy, society has less information to use to judge the relative net
benefits of one environmental policy over another. If values are context-specific so that they
change with the policy, then we cannot compare two policies using risk—benefit analysis,
because it would be like comparing apples and oranges. If the values always depend on the
context, economists can question standard risk—benefit analysis or cost-benefit analysis to
guide policy. They cannot rely on the foundation of welfare economics to define a consistent
ranking of environmental policy based on our statements of how much we prefer reduc-
tions in risk.

5.3 Valuing Risks to Life and Limb

We now examine how people value a reduction in risks to human and environmental
health, for both private and collective strategies. We focus on rational risk reductions, rec-
ognizing the caveats that arise from the behavioural issues that we just discussed. Constrained
budgets and increased fiscal accountability prevent a policy-maker from reducing all risk to
all individuals. Deciding which risks to reduce, and by how much, requires evaluation of
94

People make economic decisions every day about what goods to purchase, when to purchase them, and
how much money to allocate to current consumption or save for future consumption. When making
these intertemporal choices, people implicitly discount possible future events to compare them to
current benefits. The standard economic model suggests that dynamically consistent behaviour based on
this implicit discounting assumes that people have a constant marginal rate of time preference, ora
discount rate, when determining the current value of future streams of benefits. While the correct size of
the discount rate to use in cost-benefit analysis has been discussed at length in the literature—including
issues such as the social rate of discount versus producer and consumer rates, the social discount rate
versus the social rate of time preference, tax-induced distortions, intertemporal investment decisions,
and imperfect markets—under each of these scenarios, the form of the discount rate is assumed constant
for cost-benefit analysis. Cost-benefit analysis uses constant discounting to determine present values.
But empirical evidence now suggests that people do not use a constant discount rate when taking
actions that affect the future. Evidence suggests that people are less patient in the near term, implying
higher discount rates relative to the discount rates they use for actions in the far distant future; here,
people seem more patient with lower discount rates. This behaviour is typically called hyperbolic
discounting. Hyperbolic discounting implies that people make inconstant choices and plans over time; for
example, ‘I will fix the water pipe ... tomorrow.’ We believe that whatever we are doing later will not be as
important as what we are doing right now (see Thaler and Benartzi, 2004). Such time-inconsistent plans
help to explain such behavioural regularities as self-control, addiction, low savings rates, and
procrastination.
The evidence has led to a great deal of new research, as well as disagreement, on the role of hyperbolic
discounting in economic analysis. Most of this work on hyperbolic discounting looks for evidence to
either support or reject the use of the method, whether it is the choice of the correct discount rate or the
correct discount model. Relatively few attempts have been made to assess how hyperbolic discounting
can affect the outcomes of environmental and resource policy. Two exceptions are Settle and Shogren
(2004) and Hepburn et al. (2010). Shogren and Settle examine how hyperbolic discounting affects
behaviour within a bio-economic model of a fishery in Yellowstone National Park, Yellowstone Lake,
Wyoming. They find that hyperbolic discounting increases the size of the net policy efficiency gains
compared with constant discounting using the same initial discount rate. In addition, they estimate that
hyperbolic and constant discounting can yield the same gains, but the results lead to different time
frames for the policy. Hepburn et al. (2010) examine a fisheries model given hyperbolic discounting. They
model a scenario that allows a social planner to either commit to a time-consistent policy or re-evaluate
the policy in the future. The planner’s lack of self-control causes him to change policy in the future, which
leads to the collapse of the resource stock.
In sum, most people discount the near term at higher rates than they do the far distant future. This
hyperbolic discounting can lead to various forms of non-constant dynamic choices, such as
procrastination and a lack of self-control. Time-inconsistent choices can lead to too little investment in
current savings and too little concern about current stocks of natural resources and ecosystem services.

each new or revised regulation. Comparability of value across all sectors of the economy
requires that policy-makers rank regulatory alternatives in terms of a common unit.
Arguably, the most common denominator is money, or monetary equivalence. Rational
risk valuation systematically evaluates each regulation by estimating the monetary value—
both benefits and costs—of a reduction in risk.
Valuing the costs and benefits of reduced risk is challenging. Measuring the cost of con-
trolling risk is relatively straightforward; measuring the benefits means monetizing lower
ENVIRONMENTAL RISK AND BEHAVIOUR 95

risks to life and limb. Measuring the benefits of fewer death and injuries is difficult, because
life and limb are typically not bought and sold on the auction block. These goods enter mar-
kets indirectly.
Valuing risk reductions requires that we place a value on death and illness. These efforts
give rise to the loaded term ‘the value of statistical life’, or VSL—the idea of a monetary value
of life or, more correctly, the value of reduced mortality risk. Ethical and moral beliefs force
a person to baulk at the idea. But our everyday choices and the trade-offs that we make
implicitly generate a value on life; it is another matter whether we explicitly quantify the
VSL. Whenever a policy change is enacted or whenever the status quo remains, life and limb
are implicitly valued. Economists take it to the next step by measuring how people will trade
off goods and services for a risk reduction in the change of sudden death or injury.
Economists value a rational reduction in risk as follows:

The economic value of a small risk reduction =

[WTP for a small change in risk]/[the small change in risk].

A person’s value for a risk reduction equals his or her maximum willingness to pay (WTP)
to increase the chances to stay healthy, conditional of his or her previous private actions to
reduce risk. For example, suppose that a person was willing to pay $6 to reduce the risk of
death to 1 life in 1,000,000 from 4 lives in 1,000,000—a 3 in 1 million risk reduction. The
value of life is then $2,000,000 ( = $6/(3/1,000,000)). If the person was willing to pay $0.60,
the implied value of life would be $200,000.
This WTP is called the option price. The option price is the maximum that a person is
willing to pay that keeps him or her indifferent between the gamble and the next-best alter-
native. Consider Travis again and his decision to move to Baggs. Suppose that regulators are
thinking about a policy that would reduce air pollution so as to increase the odds of being
healthy in Las Vegas to 90:10. The maximum option price, OP, that Travis would pay for this
risk reduction is the amount that would make him indifferent between the status quo and
staying and leaving Las Vegas for Baggs:

(0.9)u($50,000 — OP) + (0.1)u($35,000 — OP) = (0.7)u($50,000) + (0.3)u($35,000);

0.9,/($50,000 — OP) + 0.1,/($35,000 — OP) = 0.74/($50,000 + 0.3./($35,000 = 213;

0.9./($50,000 — $3,000) + 0.1,/($35,000 — $3,000) = 213.

In this case, Travis would be willing to pay at most $3,000 to reduce the risk ofillness in Las
Vegas.
Alternatively, we could ask him to reveal his minimum willingness to accept compensa-
tion (WTA), C, to forgo the potential risk reduction:

(0.7)u($50,000 + C) + (0.3)u($35,000 + C) = (0.9)u($50,000) + (0.1)u($35,000)

=) (0s

0.7,/$50,000 + C + 0.3$35,000 + C = 0.9$50,000 + 0.14 $35,000;

96 ECONOMIC TOOLS FOR THE ENVIRONMENT

0.7-/$50,000 + 3,1000 + 0.3./$35,000 + 3,1000 = 220.

Travis would take at a minimum $3,100 to forgo the proposed risk reduction policy.
How do economists measure the value of risk reduction? The literature on risk valuation
has developed two general approaches to measuring the economic benefits of reduced risk:
the human capital and WTP approaches:

e Thehuman capital approach. This values risk reductions by examining a person’s lifetime
earnings and activities. The value of a risk reduction is the gain in future earnings and
consumption. The value of saving a life is calculated as what the individual contributes to
society through the net present value of future earnings and consumption. The human
capital approach has an advantage in that it is actuarial; it uses full age-specific accounting
to evaluate risk reductions. The major drawback of the approach is that it lacks
justification based on traditional economic theory. We are measuring {price and
quantity} rather than preferences and welfare.
e The WTP approach. Most economists prefer to measure the value of risk reductions
assuming that they are capturing a person’s underlying preferences for trading of risk and
reward. The WTP approach is based on traditional economic theory. Here, a person values
a risk reduction if it leads to higher satisfaction. The welfare change is measured by the
maximum that he would be willing to pay to reduce risk, or the minimum compensation
he would be willing to accept for an increase in risk. Economists use this willingness to pay
or accept to infer the implied value of life and limb. Four empirical approaches are used to
determine the WTP for risk reduction: revealed preferences, stated preferences,
experimental auctions, and averting behaviour. These are comparable with the general
valuation methods described in Chapters 3 and 4.

1. Revealed preferences/wage-risk trade-offs. Wage-risk trade-ofts are based on the theory

of hedonic prices. Hedonic price theory captures the idea that a person’s wage rate
depends on skill, education, occupation, location, environment of work, and job safety
or risk. A worker will accept a higher wage for more risk, holding all other job attributes
constant—the greater the risk, the higher the wages. And a worker selects his job to
equate the incremental WTP for each attribute to the incremental contribution of each
attribute to the wage rate. The value of risk reduction is the incremental WTP for the
attribute ‘job safety’. Workers then compare their risk-wage trade-offs to the rate at
which the market is willing to trade risk for wages. The market equilibrium between
workers and employers then determines the risk premium—the extra compensation for
risky jobs. The wage-risk trade-off is determined, other job attributes held constant.
Recent reviews of wage-risk studies set the average value of a statistical life at around $7
million (2008 dollars). These values have been challenged. Critics question the
presumptions that workers know all the risks in the job, and can change jobs without
cost. Also they point out the weak correlation between job safety and environmental
hazards. ‘They also stress that hedonic models consider a segment of the population—
people with a job; children and seniors are underrepresented.
2. Stated preferences. As we saw in Chapter 4, stated preference methods (e.g. contingent
valuation) ask people directly how much they would be willing to pay to reduce risk
ENVIRONMENTAL RISK AND BEHAVIOUR of

through a survey or interview. The approach constructs a hypothetical market, in which

a person buys or sells safety. The method attempts to reveal a person’s WTP for a risk
reduction. The challenge is to make these hypothetical markets realistic and relevant to
people. The biggest complaint is that a person answers a hypothetical question that does
not enforce his or her actual budget constraint.
3. Experimental auctions. Experimental auction markets are a recent approach to directly
value reductions in risk. Experimental auctions use the laboratory to sell real goods to
real people within a stylized setting. Laboratory experiments can isolate and control how
different auctions and market settings affect values in a setting of replication and
repetition. Experiments with repeated market experience provide a well-defined
incentive structure that allows a person to learn that the honest revelation of his or her
true preferences is his or her best strategy. Using demand revealing auctions (e.g. the
second-price, sealed-bid auction mechanism), subjects will participate in an auction
market that allows for learning as participants realize the actual monetary consequences
of their bidding. The non-hypothetical auctions with repeated market experience can
help improve the precision of risk valuation. For example, work in experimental markets
has elicited the ex ante WTP for safer food. These experiments used real money, real
food, repeated opportunities to participate in the auction market, and full information
on the probability and severity of disease resulting from the food-borne pathogens. The
value of statistical life implied by bidding behaviour in laboratory work exceeds—by an
order of magnitude—the value estimated in other valuation methods, ranging from $2
million to $70 million (see Box 5.3).
4. Self-protection/averting behaviour. The averting behaviour method estimates WTP for
risk based on what people pay to protect their families and themselves. People reveal
their preferences for lower risk through the market for self-protection, such as smoke
detectors, seat belts, medicine, bottled water, and water filters. People use these private
markets to reduce risk themselves. This self-protection raises an important issue in the
value of life and limb. The value of life or limb is defined as the cost of an unidentified
single death or injury weighted by a probability of death or injury that is uniform across
people. The WTP approach captures this cost by revealing the unobserved preferences
for risk reduction. But here is the rub. These estimates contain more than just unobserved
preferences—they capture preferences for risk reduction conditional on each person’s
unobserved ability to reduce risk privately.

Consider an example. Suppose that people have identical preferences for risk reduction
from contaminated drinking water, but that they differ in their ability to access private risk
reduction markets. And now say that each person is asked to reveal his or her value for a
collective programme to reduce risk. Each person’s value for this collective risk reduction is
conditional on his or her private actions. Following the standard procedures to value life,
one might assume that people with a low value for collective risk reduction would be willing
to tolerate greater risk. But it just might be that they have access to effective private risk
reduction and have reduced the risk themselves.
Why does this matter? This matters because the statistical value of life used in benefit-
cost estimates is biased upwards because it has not addressed these private actions. To see
this, consider the value of life used by the US Environmental Protection Agency (EPA).
98 ECONOMIC TOOLS FOR THE ENVIR

The World Health Organization has estimated that one in three people around the world get sick from
food-borne illness every year. What is the economic value of reducing the risks from food-borne
pathogens? Almost two decades ago, Hayes et al. (1995) designed a set of experimental auctions to
explore this question. They constructed an experimental auction to elicit both the option price and
compensation measures of value for five different food-borne pathogens. They used additional
treatments to evaluate how subjects respond to changes in the risk of illness for a given pathogen,
Salmonella, and to explore if pathogen-specific values act as surrogate measures of general food safety
preferences. All experiments used real money, real food, repeated opportunities to participate in the
auction market, and full information on the probability and severity of the food-borne pathogen. The
design used a classic second-price auction. The auction is ‘incentive compatible’—a bidder's weakly
dominant strategy is to bid his or her true preferences for risk reduction.
Four results emerged from their experiments. First, people underestimated the objective risk of
food-borne pathogens. Second, values across food-borne pathogens were not robust to changes in the
relative probabilities and severity, suggesting that people were placing more weight on their own prior
perceptions than on new information on the odds of illness. Third, marginal willingness to pay an option
price decreased as risk increases, again suggesting that the people weighted their prior beliefs more than
new information. Fourth, they found support for the theory that values for specific pathogens might act
as surrogates for general food-safety preferences.
Overall, the results suggest that the average subject in our experimental environment was willing to
pay approximately $0.70 per meal for safer food. The Salmonella treatments under alternative risk levels
indicate that the average person would pay about $0.30 per meal to reduce the risk of food-borne
pathogens by a fraction of ten. At the time of the auctions, if these values were to be transferred to the US
population, the value of food safety would have been at least three times the largest available estimates.

Today, the US EPA uses a value of statistical life (VSL) of about $7 million. This value is
generated from the mid-point estimate ofdifferent valuation exercises. A prominent exam-
ple is the US EPA’s application of the VSL to justify the 2000 diesel sulphur rule. Here, the
VSL accounted for nearly 90 per cent of the estimated annual total benefits from improved
air quality, at $62.6 billion out of$70.4 billion.
The VSL estimate depends in part on the private ability to reduce risk. To apply a fixed VSL
to other risk reduction policies assumes that people in another context have the same private
risk reduction opportunities. Whether this assumption holds for all markets is not obvious.
Why should the market for the private reduction of water risk be identical to the market for
toxic air risk? By focusing on collective risk reduction, the statistical life approach can bias the
value of risk reduction, which can lead to inefficient levels of environmental degradation.
Allowing a person to reveal whether he or she would prefer to reduce risk privately or col-
lectively, or both, will elicit a more exact measure of the value of risk reduction.
The value of reduced risk depends on the WTP and the change in risk. If we do not
account for people starting from different baselines due to their personal ability to access
private markets, the value will be biased. Consider two kinsmen, Riley and Ole, who are
identical in every way except for their unobserved skill or access to private risk reduction
markets. Suppose that they are asked to state their WTP for a collective policy that will
increase the odds of a gamble from 50:50 to a 100 per cent chance that the good state of the
world will be realized.
ENVIRONMENTAL RISK AND BEHAVIOUR 99

Riley says he will pay nothing for the change in risk, and Ole says he'll pay $100. The tra-
ditional estimate of the value of this risk reduction is

Value of risk reduction without private action = 1/2[$0/0.5] + 1/2[$100/0.5] = $100.

But this presumes that Riley and Ole face the same baseline and the same change in risk,
50:50, even though their unobserved skill or market access differs. Let Riley have high
skill or access, such that his real odds are 90:10 of agood outcome prior to the collective
policy, such that he is paying fora real change in risk that equals 0.1 = 1.00 - 0.9; whereas
Ole has low skill or access and his real odds are 10:90. Ole is paying $100 for a 0.9 change
in risk: 1.0 — 0.1. The value of risk reduction conditional on private risk reduction is
then

Value of risk reduction with private action = 1/2[$0/0.1] + 1/2[$100/0.9] = $56.

The value of risk reduction is lower than the traditional measure if one accounts for private
actions changing the baseline risk. If the WTP amounts were reversed such that Riley paid
$100 and Ole paid nothing, the new value would exceed the traditional measure:

Value of risk reduction with private action = 1/2[$100/0.1] + 1/2[$0/0.9] = $500.

This example is less likely to be observed, however, since low skill and low risk aversion are
more likely to be correlated. Private actions affect the baseline risk, and this alters the value
of risk reduction of the average person.

5.4 Regulating Risk

Managing and regulating risk in society requires regulators to integrate assessment, psy-
chology, economics, and political factors. Risk-management policies are complicated by
numerous factors: scientific complexity and uncertainty, political and economic pressure
from special interest groups, financial abilities to clean up disposal sites, jurisdictional dis-
putes, unresolved liability, and variations in local, state, and federal policy goals. Successful
strategies to manage risk should address which risks we want to confront now and in the
future, how we will control these risks in a cost-effective manner, and how we balance who
faces what risk.
Consider several ways in which to select the risks we choose to face. A common first reac-
tion is to want to set a target ofzero risk to society. Regulation in the United States—such as
the Delaney Clause of the Food, Drug, and Cosmetic Act, which prohibited the presence of
any known carcinogen as a food additive in processed food—is a zero-risk regulation. As
science becomes better at measuring small amounts of trace chemicals that are potential
carcinogens, the zero-risk approach is restrictive. If almost everything is feared to cause
cancer of some form, what can we eliminate? The costs of hitting a zero-risk target become
prohibitive. Some actions and activities can be so risky that society should ban their use—
zero risk because they are gone from society.
100 ECONOMIC TOOLS FOR THE ENVIRONMENT

The next wave is for society to set an acceptable risk target that can be reached using cur-
rent or new technology. Technology-based standards are a centralized process of setting
permissible levels of contamination or building codes. People or firms who ignore these
standards would be punished in civil or criminal court. Examples include uniform limits on
total emissions per day or year, the emission per tonne of input used in a production pro-
cess, and the type of equipment used in production, such that it is the best available control
technology. One argument advanced by proponents of technology standards is that tech-
nology-based engineering decisions that construct a uniform threshold of acceptable risk
have to be measured; costs and benefits are left unmeasured. But uniform standards are
likely to be inefficient, as we saw in Chapter 2.
A third wave is to promote cost-effective risk reduction. Dollars now enter into the pic-
ture. Costs matter. Cost-effectiveness allows regulators to set a target and then asks that
people be allowed to find the most cost-effective path to achieve the target. The idea is to
take the public’s preferences and perceptions of risk into consideration. This can be accom-
plished through open meetings in which regulators and the public set health and safety
targets. Cost-effectiveness attempts to find the least costly method to achieve the goals. One
advantage of cost-effectiveness is that it does not have to measure the benefits of the target.
The method maximizes lives saved given a fixed budget in which assumptions on values are
built into the model.
If we want to consider trade-offs involved in risk management but still not measure costs
or benefits, regulators can address risk-risk trade-offs, or comparative risk analysis. Risk—
risk analysis compares how we trade off one risk for another. For example, an energy policy
that would switch to more nuclear power and less coal power would shift the nature of risk
to radiation rather than climate change. A shift to more hydropower would shift the risk
towards more protection of endangered species. The framework requires estimation of the
trade-off between consumer health risks and substances that offer a direct health benefit.
The health benefits of drugs, exercise, and diet, for example, fit into this framework. The
benefit of the risk-risk framework is that regulators can convert health outcomes into fatal-
ity risk equivalents, which might allow more meaningful comparisons than a risk—dollar
trade-off.
Finally, policy-makers can use cost-benefit analyses, as we saw in the previous chapter.
Here, we ask for dollar measures of both costs and benefits, and a direct comparison ofthe
trade-off between risk and dollar benefits. As we discussed, economists have devoted con-
siderable effort to determining the value of reduced risk. Cost-benefit analysis can be used
as a tool to measure the economic efficiency of a regulation, Cost-benefit analysis attempts
to measure the costs associated with the risk regulation and the subsequent welfare benefits
from a risk reduction. The costs ofdiffering policy alternatives are then compared with their
benefits to determine if, and to what extent, the risk will be reduced. The goal of cost-benefit
analysis is to maximize economic efficiency and make the resulting risk reductions as large
as possible.
There are many controversial aspects to cost-benefit analysis. The value of risks to life
must be addressed in a policy context. The appropriate discount rate remains a question.
Exponential discount rates place less weight on the future. We must address equity and
distributional questions: Whose risk will be reduced and who will pay? An equity criterion
spreads out the costs and benefits of risk based on some subjective measures, to weigh who
ENVIRONMENTAL RISK AND BEHAVIOUR 101

gets what for which price. Risk can be distributed equally among the population; or it can be
progressively or regressively distributed based on, say, wealth.
Environmental risk to children is a prime example of questions over whose risk we should
be reducing. Evidence suggests that children face disproportionate health risks from envi-
ronmental hazards. These unbalanced risks stem from several fundamental differences in
the physiologies and activities of children and adults. As children develop, their digestive,
nerve, and immune systems are more susceptible to toxic pollutants and other environmen-
tal hazards. Children eat, drink, and breathe more for their weight, and spend more time
outside, exposing themselves to greater amounts of contamination and pollution for their
weight than adults. They also face potential exposures over their entire lifetime. They are
also less able to recognize and to protect themselves. All of this suggests that children require
special attention when dealing with environmental risk.
On the basis of this argument, many politicians promote policies that explicitly aim to
protect children from environmental risks. In the USA, the federal government has been
tasked with safeguarding children from environmental threats through more policy, better
research coordination, and more federal regulatory analysis. All US federal agencies now
make the protection of children a high priority when implementing their statutory respon-
sibilities and fulfilling their overall missions. Agencies promulgating major regulations that
may have a disproportionate impact on children must now evaluate how regulation could
affect children’s risk, and then explain why the planned regulation is preferable to alterna-
tive actions that might have more cost and less risk.
This forces agencies to ratchet up their regulatory standards, with a corresponding
increase in the costs and burden of regulation. The pressure to raise standards across the
board may generate criticism from industry and other groups who argue that analysis of
impacts on children can lead to costly decisions; in other words, Superfund clean-ups based
on exposure of children to toxins, and analytical flaws in the public health data supporting
recent Clean Air Act proposals on ozone and particulate matter. The additional burden may
further delay the regulatory process, and add resource demands to agencies confronting
tight budgetary constraints.
Regulators have many tools at their disposal to reduce risk, either to adults or children.
They can impose mandates, liability rules, pollution taxes, and subsidies, create new mar-
kets, and use informed consent through risk communication. We have discussed taxes and
markets previously. Consider risk communication strategies.
The major benefit of risk communication and informed consent is that people are allowed
to make informed choices based on preferences towards risk rather than uniform govern-
ment bans or regulation. The risk manager must be sure that the information that consum-
ers have will result in more accurate private decisions regarding risk. But the language of the
hazard warnings seems to maximize political interests rather than advancing the primary
objective of informing consumers and enabling them to make better decisions. By ignoring
fundamental economic and psychological concepts of decision-making under risk, warn-
ings will not convey the information necessary for consumers to make sound choices
regarding risks and precautions.
But the regulators and the public must also be aware that risks can be regulated by being
transformed and transferred elsewhere. Transferable risk implies that people protect
themselves by transferring the risk through space, to another location, or through time, to
102 ECONOMIC TOOLS FOR THE ENVIRONMENT

another generation. Most environmental programmes do not reduce environmental risks

by cutting the mass of materials used or causing them to accumulate in the economy. People
select a technology that transfers a risk that creates conflict and induces strategic behaviour.
Some nations and states reduce their air pollution by building tall stacks such that the winds
carry the emissions to those downwind. Some local governments ban toxin storage within
their jurisdictions, thereby shifting the problems elsewhere. Effective risk management
should address the question of transferability. \
We now examine how behavioural economics might affect how we design policy to
address market failure and environmental policy. As we discussed, if their rationality
and willpower are bounded, people do not always react as predicted. One idea for incen-
tives for people with self-control problems and temptation is for the person to choose
from a smaller rather than a larger menu of actions for protection. This smaller-is-better
finding also appears to exist for US and Canadian smokers, who act as if they are happier
with higher cigarette taxes. Few such examples of applying behavioural economics to
incentives exist in the environmental economics literature. If a person is altruistic
towards others, then the existence of altruism itself is insufficient to generate a lower
Pigovian tax. But if people behave as if they are ‘addicted to’ the good generating the
negative environmental impact, an optimal environmental tax should exceed the stand-
ard Pigovian tax.
In the case of environmental issues, there is also scope for incorporating inputs from
behavioural economics to maximize the policy goals. In recent research, some of the envi-
ronmental and ecological economists have used the alternative models of rationality in their
analyses related to environmental policy. Kallbekken et al. (2011) examine behavioural eco-
nomics and Pigovian taxation in the laboratory. They consider how aversion to paying taxes
affects the framing and functioning of the classic Pigovian tax. Two results emerge. First,
people were not confused about the nature of Pigovian taxation; they understood how these
taxes work and why society might need them—they just did not like the ‘t-word’, tax.
Second, reframing the tax as a ‘fee’ increased support, especially when revenues were ear-
marked for the environmental problem. A targeted rebate that reduced inequalities in the
distribution of wealth was also preferred by the subjects, which supports the behavioural
economics notion of‘inequality aversion’.
Another example is Banerjee and Shogren’s (2012) model of environmental protection
given the existence of social preferences. The authors consider how to design incentives
when a firm has some non-monetary reason to protect the environment. Consider a con-
crete piece of reality to help think about the relevance of the theory. People have social
preferences to help out even though it is privately costly to them; for example, a landowner
who protects endangered species on his land even if it reduces his land rents. Some people
have social preferences to protect the environment without having to pay. Paying them to
protect nature might be counterproductive. Money ‘crowds out’ their willingness to do the
good deed. Behavioural economists have argued that monetary rewards weaken intrinsic
motivation; their terms are dramatic—the hidden cost of reward, the over-justification
effect, or the corruption effect. Here, monetary rewards reduce the ability to indulge altru-
istic feelings; or cause others to doubt our true motive for doing a good deed. If the crowd-
ing out effect holds, monetary reward decreases effort—the exact opposite of what
economics would predict. For example, in case of forest habitat preservation in Finland,
ENVIRONMENTAL RISK AND BEHAVIOUR 103

private property owners with a positive attitude towards environmental protection actu-
ally claim less monetary transfer (Mantymaa et al., 2009).
In contrast, other people do not have strong social preferences for the environment. They
are unwilling to pick up the tab to protect a public good. But it is difficult to identify, by
observing people’s behaviour, why they contribute to a social project—whether this is due
to intrinsic motivation or sociality—as people care about reputation too. These folks might
want to protect the environment to ‘buy’ a good reputation. A good reputation might be
useful to attract new customers, gain better access to capital or credit markets, entice new
property buyers, and so on. Offering up monetary rewards to these folks could be counter-
productive if they wish to avoid being viewed as ‘greedy’ rather than generous.
The regulator’s dilemma is that she does not know which person is which—social prefer-
ences or reputation buyer. How does she design a mechanism given that she knows both
types exist, but that she does not know who is who. She does not want to chase away the
person with social preferences by crowding out that person’s incentives to do the right thing;
she does not want to reward the reputation seeker by paying out extra money that could be
spent elsewhere. The open question is whether she can design a mechanism that specifies a
menu of monetary transfer-to-effort that gets the best out of both types of people. First, the
social firm receives less transfer than is optimal and thereby under-invests in effort (relative
to the full information case). In contrast, the reputation-driven firm receives the optimal
level of monetary reward and expends optimal effort. Second, the social firm earns no
information rents; but the reputation-driven firm earns negative information rents. The
reputation-driven firm pays out money; it ‘buys’ a reputation for environmental protection.

Summary
People affect nature, and nature affects people. Together, people interacting with nature
create risks to human and environmental health. More understanding is needed on how we
create risk, how we behave towards risk—rationally or otherwise, how we trade off risk and
rewards, and how we can design incentives to induce people to reduce risks to human and
environmental health. Understanding economic behaviour under risk can help make our
decisions to control risk more effective—reducing more risk for more people and for nature.
Knowing how to assess risk accurately, whether people make risky choices with reason or at
random, what people are willing to pay to reduce risk, and what options exist to control risk
can help us make better decisions on how to save lives and protect nature at less cost.

Tutorial Questions
5.1 Isa zero-risk society possible?
5.2. Explain the pros and cons of using rational choice theory to help think about how to
manage risks to human and environmental health.
5.3 How might private self-protection and self-insurance affect the demand for
governmental reduction in risks to life and limb?
104 ECONOMIC TOOLS FOR THE ENVIRONMENT

5.4 How does expected utility differ from expected values? Why does this difference
matter for policy?
5.5 If people reverse their preferences for risky events, should the government
incorporate or ignore their behaviour when making public policy?
5.6 Explain how one might measure the value of statistical life.

References and Additional Readings

Allais, M. (1953). “Le comportement de Thomme Kallbekken, S., Kroll, S., and Cherry, T. (2011). ‘Do
rationnel devant le risque: critique des postulats you not like Pigou, or do you not understand
et axiomes de l’ecole Americaine, Econometrica him? Tax aversion and revenue recycling in the
21: 503-46. lab, Journal of Environmental Economics and
Banerjee, P., and Shogren, J. (2012). ‘Material Management 62: 53-64.
interests, moral reputation, and crowding out Machina, M. (1987). ‘Choice under uncertainty:
species protection on private land} Journal of problems solved and unsolved? Journal of
Environmental Economics and Management 63: Economic Perspectives 1: 121-54.
137-49. Mantymaa, E., Juutinen, A., Monkkonen, M.,
Baron, R. (2008). Thinking and Deciding (Cambridge: and Svento, R. (2009). “Participation and
Cambridge University Press). compensation claims in voluntary forest
Brekke, K., and Johansson-Stenman, O. (2008). “The conservation: a case of privately owned forests in
behavioural economics of climate change, Oxford Finland; Forest Policy and Economics 11: 498-507.
Review of Economic Policy 24(2): 280-97. Metcalfe, R., and Dolan, P. (2012). ‘Behavioural
Cameron, T.A. (2010). “Euthanizing the value of a economics and its implications for transport,
statistical life, Review of Environmental Economics Journal of Transport Geography (in press).
and Policy 4: 161-78. Settle, C., and Shogren, J. (2004). “Hyperbolic
Ehrlich, L., and Becker, G.S. (1972). ‘Market discounting and time inconsistency in a native—
insurance, self-insurance and self-protection, exotic species conflict, Resource and Energy
Journal ofPolitical Economy 80: 623-48. Economics 26: 255-74.

Gillingham, K., Newell, R., and Palmer, K. (2009). Shogren, J. F., and Crocker, T. (1999). ‘Risk and
‘Energy efficiency economics and policy, NBER its consequences, Journal of Environmental
Working Paper No. 15031. Economics and Management 37: 44-51.

Hanley, N., and Shogren, J. (2005). ‘Is cost-benefit and Taylor, L. (2008). ‘On behavioral-
analysis anomaly-proof?’ Environmental and environmental economics, Review of
Resource Economics 32(1): 13-34. Environmental Economics and Policy 2: 26-44,

Hayes, D., Shogren, J., Shin, S., and Kliebenstein, J. Thaler, R., and Benartzi, S. (2004). ‘Save More
(1995). ‘Valuing food safety in experimental Tomorrow”: using behavioral economics to
auction markets, American Journal ofAgricultural increase employee savings, Journal of Political
Economics 77; 40-53, Economy 112: $164-87.

Hepburn, C., Duncan, S., and Papachristodoulou, Thaler, R., and Sunstein, C. (2008). Nudge: Improving
A. (2010). ‘Behavioural economics, hyperbolic Decisions about Health, Wealth, and Happiness
discounting, and environmental policy, (New Haven, CT: Yale University Press).
Environmental and Resource Economics 46: ‘Iversky, A., and Kahneman, D. (1981). ‘The framing
189-206. of decisions and the psychology of choice, Science
Kahneman, D. (2011). Thinking, Fast and Slow 211; 453-8.
(New York: Farrar, Straus, and Giroux). Viscusi, W.K. (2009). ‘The devaluation of life,
Kahneman, D., and Tversky, A. (eds.) (2000). Regulation & Governance 3: 103-27.
Choices, Values, and Frames (Cambridge:
Cambridge University Press).
Economic Growth, the
Environment, and
Sustainable Development

This chapter discusses the concepts of economic growth and sustainable development.
Section 6.1 outlines what we mean by growth, a general understanding of how economies
grow, and the differences between growth and development. In Section 6.2 we review the
history of economic thinking about the links between growth and the environment. One
controversial aspect of the debate is whether an economy can grow its way out of environ-
mental problems, and this is covered in Section 6.3. Most debates on the links between
growth and the environment are now undertaken in the context of the idea of ‘sustainable
development’. Section 6.4 explains how economists view this concept. The chapter closes
with a review of different economic indicators of sustainability (Section 6.5).!

6.1 Economic Growth and Development

Governments worldwide are concerned with economic growth as a measure of a country’s
performance, both in absolute terms (how fast are we growing?) and in relative terms (are
we growing faster than our neighbours?). Newspapers and television often report actual
growth figures and predictions of future growth. But growth in what? Why does it occur?
Are ‘growth’ and ‘development’ the same thing? This section addresses these questions,
before turning to the relationship between growth and the environment.
Economic growth is commonly understood to reflect an increase in people’s living stand-
ards over time. What we want is an indicator of how well we are doing as a society. This is a
hard question to answer. Economists have developed the idea of gross national product, or
GNP, a Nobel Prize-winning idea. GNP is a monetary measure of the total value of output of
a country in any time period. GNP is also a measure of national incomes to all factors of pro-
duction involved in economic activity (land, labour, and capital), and of total spending (con-
sumption plus gross investment). GNP is often expressed in per capita terms, to allow for the

| We are grateful to one reviewer for pointing out that, as part of the Convention of Biological Diversity Aichi
targets (2011), countries are now required to develop national green accounting measures, which might well
include the kinds of economic indicator of sustainability described here.
106 ECONOMIC TOOLS FOR THE ENVIRONMENT

Table 6.1 The ranking of global economies by GDP in purchasing power parity (PPP) terms, 2008
Se ene ee eee a
Ranking Country Economy (millions of international
dollars)

(a) Top twenty

1 United States 14,093,310
2 China 7,909,261
3 Japan ; 4,358,472
4 ndia 3,358,871
5 Germany 2,904,557
6 Russian Federation 2,260,202
iy United Kingdom 2,178,205
8 France 2 121,724
9 Brazil 1,978,139
10 taly 871,709
1] exico 1,549,490
12 Spain 1,442,936
13 Republic of Korea 1,344,360
14 Canada W301,737
15 Turkey 991,715
16 Indonesia 907,955
17 Australia 831,220
18 Islamic Republic of Iran 778,779
1g Netherlands 673,634
20 Poland 658,611
(b) Countries ranked from 140th to 160th
140 Rwanda 9,985
141 ongolia 9,396
142 alta 8,649
143 ontenegro 8,330
144 auritania 6,021
145 Swaziland 5,754
46 Togo 5,361]
147 Sierra Leone 4 350
148 Suriname 3,812
149 Fiji 3,678
50 Bhutan 3,268
15] Central African Republic 3,216
152 Lesotho 3,205
Ss) Eritrea 3.161
154 Burundi 3,093
535) Guyana 2339
56 The Gambia 2,263
157 Belize alee
158 Seychelles 1,860
By Antigua and Barbuda 1,817
160 Djibouti 1,816
Note: ‘Purchasing power parity’ means that the data have been adjusted to take into account the different
prices in the countries in the table—each dollar of PPP is supposed to buy the same ‘basket’ of goods.
Source: World Bank Development Indicators database, 2010.
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 107

Table 6.2 Per capita gross national income (GNI) for ten highest-ranked and ten
lowest-ranked countries, 2008
ee ee Senn Se
Rank Country GNI per capita (US$)
Ten richest countries on GNI per capita terms
] Lichtenstein F790
Zz Norway 87,340
3 Channel Islands 68,610
4 Luxembourg 69,390
5 Denmark 58,800
6 Switzerland Sy OIG
A Sweden 50,910
8 Ireland 49,770
9 Netherlands 49 340
10 Kuwait 43,930

Ten poorest countries on GNI per capita terms

200 Guinea 350
201 Niger 330
202 Sierra Leone 520
203 Eritrea 300
204 Ethiopia 280
205 Malawi 280
206 Guinea-Bissau 250
207 Liberia 170
208 Democratic Republic of the Congo 150
209 Burundi 140

effects of changes in population. If real per capita GNP is rising, it is usual to say that a coun-
try is experiencing economic growth. Tables 6.1 and 6.2 show gross domestic product (closely
related to GNP) and gross national income (the income measure of GNP) per capita for the
world in 2008. GNP is a measure of living standards in the sense that it measures the size of
the ‘economic cake’ to be divided amongst the inhabitants of a country; that is, the amount
of total income to be divided up. Looking at Table 6.2 in particular, which focuses on per-
person measures, it is striking how the ten poorest countries in the world on this measure are
all from Africa, whilst the ‘ten richest’ list is dominated by countries from Northern Europe,
along with a Middle Eastern oil economy—incidentally, the United States comes after Kuwait
in terms of the ranking by GNI per capita. In terms of absolute gross domestic product (GDP)
levels, the USA is the world’s biggest economy, followed by China.
Economic growth is measured as the change in GDP or GNP over a time period such as
a year. Table 6.3 shows rates of economic growth for four countries: Vietnam, Turkey,
Brazil, and Malawi. Table 6.3 also illustrates the effects of population change on how much
growth benefits the ‘average citizen’ for Malawi, China and Peru: Malawi looks to have
grown appreciably over the 10 years to 2008 (3%), but because of a high population growth,
its per capita GDP only rose by 0.1 per cent on average over the decade. China, on the other
hand, has had a rate of growth of more than 9 per cent in both absolute GDP and GDP per
capita. Rising GNP per capita would not necessarily mean that absolutely everyone is better
108 ECONOMIC TOOLS FOR THE ENVIRONMENT

Table 6.3 Economic growth rates for a selection of countries

1988-98 (mean) 1998-2008 (mean) 2008 only

Vietnam 8.0 7A 6.2
Turkey 43 47 0.9
Brazil js) se) Sal
Malawi 315 SHO) ey

Malawi China Peru

Rate of growth of GDP 1998-2008 3% 9.9% 5%

Rate of growth of GDP per capita, 1998-2008 0.1% 9.2% 3.6%

off, since some could lose or stand still, whilst others gain more than average. GNP tells us
nothing of changes in the income distribution. However, rising real GNP might be claimed
to show that, on average, people were getting better off over time. Lower growth rates have
a big opportunity cost over time, since high rates of GNP growth can imply very big increases
in the absolute level of real GNP over time. Compound growth tells us that an economy
growing at 10 per cent per annum could double its GNP in 7 years. A country that cannot
maintain growth rates falls further and further behind its competitors.
Many critics complain that GNP is an inadequate measure of well-being, as Box 6.1
shows. However, despite these criticisms, nations still use increases in real GNP per capita
to measure their economic performance over time. But how can GNP increase over time? In
other words, what is the theory behind economic growth?

Since the end of the Second World War, countries around the world have measured and compared their
levels of well-being using gross national product, following guidelines laid out in the System of National
Accounts. This seems sensible, since GNP measures our income as a country, as well as the value of what
we are producing. If population is changing, or we wish to compare GNP across different countries, then
we can divide GNP by population to get GNP per capita, which shows how much income, on average,
each person in a country has. GNP has also been promoted as an indicator of national well-being, or
welfare. Another measure of economic performance from the System of National Accounts is net
national product (NNP): this is defined as GNP minus depreciation (wearing out) of the stock of
manufactured capital during the year.
However, GNP has been widely criticized as a measure of well-being. Some of the main criticisms are
as follows:

® The effects of the economy on the environment are not well measured by GNP. For example, if there was
a major oil spill in a country and a great deal of money had to be spent to clean it up, then GNP could rise
(since the pollution clean-up industry increases its output) even if people feel worse off as a result.
® A closely related point is that changes in our natural resource stocks do not show up in GNP; for
example, if farming generates high levels of soil erosion so that the productive stocks of soil ina
country are significantly depleted.
e Although GNP measures the size of the economic pie, it does not tell us how fairly it is divided up.
GNP per capita could be increasing but income inequalities could worsen.
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 109

6.1.1 Why do economies grow?

Economic growth over the long term is due to increases in potential output, rather than
short-term fluctuations. What causes potential output to rise? Two factors are important:
growth in the resources that can be devoted to production, and changes in the productivity
of these resources.
Increasing resources. If acountry experiences an increase in its resource base, then GNP
can increase. This resource base includes capital, labour, land, energy, and material
resources. Labour supplies can rise with population growth or migration, although this
might not mean that GNP per capita is increasing, even though aggregate GNP is. Land
resources can increase if a country exploits overseas colonies, or converts marginal land to
more profitable uses. Energy and material resources can increase through new discoveries,
such as oil or gas deposits. Technological progress can also drive down the costs of exploit-
ing natural resources, and thus make more of the physical resource stock profitable to
exploit. This increases the economic reserves of that resource. Capital stocks also rise over
time. Indeed, capital accumulation was one of the earliest explanations for economic growth
put forward by the Classical economists and by Marx. Capital accumulates over time
through the process of investment. The productivity of capital means that positive net
investment (i.e. gross investment minus depreciation) can occur even with constant con-
sumption. Capital productivity means that each £1 of resources invested in capital will gen-
erate more than £1 in consumption goods over time.
Productivity growth. As people learn more through education and training, the produc-
tivity of all of the resources used in economic activity increases. For example, the average
productivity of miners in terms of tonnes of coal extracted per worker per annum, and the
average productivity of agricultural land in terms of yields per hectare, have both risen
steadily over time. The more a country invests in its workers, through education and train-
ing, the faster we expect growth to be. Education pays by increasing the stock of human
capital through increasing the skills of the workforce. Human productivity can also increase
through learning-by-doing. A famous example of this refers to steelworkers in Horndal,
Sweden in the 1800s, who despite no changes in the machinery they worked with, or in the
size of the workforce, managed to increase output by 2 per cent per year, just through
becoming more experienced at their jobs (Begg et al., 1987). Produced capital can also
become more productive, with the embodiment of knowledge in new designs of machinery.
This source of growth is endogenous in the sense that the greater the amount of resources
devoted to research and development activities, the greater will be the rate of technological
innovation (other things being equal).

6.1.2 Growth versus development

Growth, as measured by rising GNP per capita, is not usually assumed to be the same as
‘development’. This is because development is interpreted much more broadly (see, e.g., the
UN Human Development Report); and because of the limitations of GNP as a measure of
well-being referred to in Box 6.1. For an economy to be experiencing development, as dis-
tinct from growth, we might expect to see a whole series of indicators improving over time.
These might include:
110 ECONOMIC TOOLS FOR THE ENVIRONMENT

Table 6.4 Human Development Index scores, 2007 data

a nnn EEE

Rank on HDI Country HDI score GDP per capita in PPP

] orway 0.97 53,433

2 Australia 0.97 34,923
3 celand 0.96 35,742
4 Canada 0.96 35,812
5 reland 0.96 44,613
lsv/ Uganda 0.5] 1,059
158 Nigeria 0.51 1,969
iisg Togo 0.49 788
160 alawl 0.49 761
161 Benin 0.49 1,312
Source: Human Development Report (2009).

e GNP per capita—this is still important

e A reduction in income inequality
e Animprovement in adult literacy rates
e Areduction in infant mortality
» Reductions in morbidity (illness) and mortality (death) rates amongst adults
» Animprovement ina range of environmental indicators

These indicators are both monetary and non-monetary. The human development indica-
tor (HDI) was introduced by the United Nations (UN) in 1990 (UNDP, 1990) and is pub-
lished annually. A set of three indicators was chosen to represent aspects of development for
a nation. These comprise GDP, a measure of education levels, and a measure oflife expec-
tancy. In Table 6.4, HDI rankings are compared with ranking in terms of GDP per capita.
As may be seen, there is a strong correlation between the two for these countries over these
years. This is because income is a very major determinant of the other measures of well-
being incorporated in the HDI. Countries that do well in terms of GDP per capita also tend
to do well in terms of the HDI. However, both the HDI and GNP measures omit any direct
account of environmental degradation.

6.2 Predictions from the Past

Above, we reviewed the main reasons why GNP per capita can grow over time, and thus
why economic growth occurs. But for how long can growth continue? Does continual eco-
nomic growth come with a health warning? One of the biggest intellectual debates through-
out the history of economics has been concerned with these questions, and part of this
debate has been the nature of the relationship between the economic system and the envi-
ronment. The earliest economists, known as the Classical School, worried about the interac-
tion of these two systems. Thomas Malthus (1766-1834) set out to formalize the implications
for people’s standard ofliving of exponential population growth coupled with linear growth
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 111

in food output from farming. In Malthus’ view (1798), rising living standards were causing
the birth rate to rise. Eventually, food demand would outstrip food supply, and war, disease,
or famine would occur. The population would then crash, before restarting its exponential
growth once food production per head had recovered. It is not surprising, given this picture,
that economics became known as the dismal science, since the only equilibrium situation
was one of subsistence wages. Malthus’ model was overly simplistic, since it ignored many
of the factors that we now see as important, such as technological progress. His predictions
have not come true in general, but they had a great influence on later thinkers, including
Darwin and Keynes.
David Ricardo’s work (1772-1823) led to a similar long-run gloomy prediction as that of
Malthus, although at a more general level. Ricardo (1817) made use of the concept of dimin-
ishing marginal returns: as wages rose above subsistence levels and population increased,
the rise in food demand caused agriculture to expand on to land of lower and lower produc-
tive quality. Since food prices would have to rise to cover the increasing costs of producing
on less and less fertile land, those farmers growing on the most fertile land would earn a
profit, or rent, as the difference between price and production cost. The distributional impli-
cations were profound: as food prices rose, workers got poorer and poorer, whilst landlords
earned higher and higher rents. Ricardo and Malthus thus put forward two different expla-
nations for increasing scarcity. For Malthus, the problem was that we have a fixed amount
of natural resources (land) but increasing demands on those resources. For Ricardo, the
most important fact was that as demand for food rose, the average productivity of land fell
and the cost of producing food rose. These two views, of absolute and relative scarcity, are
now referred to as Malthusian and Ricardian scarcity, respectively.
Ricardo’s work is still very important; for example, in current work on understanding the
movements of cultivation frontiers in developing countries (see Chapter 10), and the
impacts of climate change on agriculture (Seo et al., 2009). However, his gloomy predictions
did not come true due to productivity improvements and the development of colonialism,
which effectively greatly increased the productive land base of European countries such as
the United Kingdom and Germany, as well as expanding trade opportunities. Increasing
use of fossil fuels meant that transportation became faster and cheaper, which also greatly
expanded world trade (Common, 1988).
Two other early economists have also shaped our thinking about environmental and
natural resources. These are John Stuart Mill (1806-73) and W. Stanley Jevons (1835-82).
Mill’s Principles of Political Economy, published in 1857, is now viewed as the climax of
Classical economics. Mill makes clear that economic growth is a race between diminishing
marginal returns and technological progress: technological progress drives down produc-
tion costs as increasing Ricardian scarcity drives them up. Economic growth, through capi-
tal accumulation, results in higher living standards. Mill saw natural resources as productive
(land for food, mines for coal) and asa direct source ofutility in itself. He also suggested that
economies would eventually evolve into a steady state, where growth ceased. In this steady
state, Mill saw it as important that we do not completely devastate the environment in the
pursuit of growth:

Nor is there much satisfaction in contemplating the world with nothing left to the spontaneous
action of nature: with every rood of land brought into cultivation and . . . every flowery waste
112 ECONOMIC TOOLS FOR THE ENVIRONMENT

ploughed up ... If the earth must lose that great portion of its pleasantness . . . for the mere
purpose of enabling it to support a larger population . . . then I hope (they) will content to be
stationary long before necessity compels them to it.
(Mill, 1857; quoted in Common, 1988)

Jevons is usually credited as being one of the first neoclassical economists, in that he
helped introduce the marginal analysis that enabled economics to become systematic and
rigorous in its approach to formulating how markets worked and the action of Adam
Smith’s ‘invisible hand’. Jevons was also concerned about the implications of limited non-
renewable resource inputs for economic growth. Coal was the most important natural
resource powering the British Industrial Revolution from the early eighteenth century
onwards (Warde, 2007). As more and more coal was dug up, both Jevons and Mill worried
that it would become more expensive to extract, since mines would have to be dug deeper
and deeper, increasing labour costs per ton. This is an application of the law of diminishing
returns to resource deposits, and is an example of the concept of Ricardian scarcity. Jevons
indeed viewed limited resource stocks as a great threat to British development, as his work
The Coal Question (1865) made clear.
The idea that perpetual economic growth was neither inevitable nor desirable was thus
first introduced to economics by Mill in 1857. This idea was reinvigorated within economics
in the 1970s by writers such as Mishan and Daly, particularly in the latter’s book Steady State
Economics. This view, along with those of ecologists such as Holling, Erhlich, and Odum, was
influential in the development of the paradigm of ecological economics in the late 1980s and
1990s. The links between economic growth and the environment, and the concept of sustain-
able development, have been a crucial part of this paradigm. Mainstream economics has
become increasingly interested in links between economic growth and measures of personal
well-being; for instance, in the question as to whether rising real incomes necessarily result
in increasing ‘happiness’, as measured by survey questions. One finding that emerges from
this new literature is that increases in absolute incomes do not seem to matter as much as
changes in relative income (Blanchflower and Oswald, 2004), although this is not true for
households below certain thresholds (in other words, rising absolute incomes always makes
very poor households happier). A good review of the income-happiness relationship can be
found in Clark et al. (2008). In developed countries, there seems to be a relationship between
‘happiness’ and local environmental quality. For example, air pollution has been found to
have a direct negative effect on individuals’ subjective well-being in different countries and
using different data sets (see Levinson, 2009; Ferreira and Moro, 2010).

6.3 Growth and the Environment: The Environmental

Kuznets Curve

Can economies grow their way out of environmental problems? What are the links between
economic growth and environmental quality? One concept that has been much used to
address this question is the environmental Kuznets curve (EKC). This was named after Simon
Kuznets, who in 1955 hypothesized an inverted-U-shaped relationship between the equality
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 113

of income distribution and income levels. A similar relationship has been claimed to exist
between income levels and environmental quality by proponents of the EKC hypothesis. This
relationship was first identified empirically by Grossman and Krueger (1995).
Indeed, the relationship between economic growth and environmental quality has been
the focus of much work historically. For example, the ‘limits to growth’ school (Meadows
et al., 1972) argued strongly that this relationship is negative, in that economic growth is, by
definition, bad for the environment as it leads to more resource use and more pollution. The
EKC hypothesis suggests otherwise. In the EKC literature, growth is usually measured as the
change in income (GNP) per capita. Environmental quality is commonly measured by indi-
vidual pollutant emission levels, ambient air quality or water quality. Most empirical EKCs
are estimated for a single pollutant. The EKC hypothesis states that as per capita incomes
grow, environmental impacts rise, hit a maximum, and then decline. This implies an
‘inverted-U’ shape, as shown in Figure 6.1.
Two parts of the curve can be identified, before and after a turning point at Y*. Up to Y*
pollution is rising, and environmental quality is falling. What gives rise to this pattern?
Economic growth is argued to result in rising pollution since:

e Economic growth results in an increasing use of resources, which also gives rise to an
increase in waste. This is known as a ‘scale effect’.
e Ifacountry starts from an early development stage as an agricultural economy, then
industrialization also leads to an increase in emissions as manufacturing takes over
from agriculture as the dominant economic activity. In other words, growth is
associated with a change in the structure of an economy that results in more pollution.

After Y*, though, emissions fall. Why?

e There may be an increasing demand for environmental quality as incomes go up. This
leads to an increase in government protection of the environment, and increasing green
consumerism.

pollution
(e.g.
capita)
per
Environmental
impacts

Income per capita

Figure 6.1 The environmental Kuznets curve.

114 ECONOMIC TOOLS FOR THE ENVIRONMENT

e Technological improvements over time make production per unit of output cleaner
(a ‘technique effect’), whilst economies of scale in pollution abatement might also kick
in (Andreoni and Levinson, 2001).
e Further changes in the structure of the economy occur, such as moves from
manufacturing to service-sector or high-tech industries.
e Increasing scarcity of ‘environmental quality’ drives up its relative price, and this
means that less is ‘consumed’ and more is preserved—although the non-market nature
of many environmental goods means that this pressure fails to translate into
appropriate market signals. Another way of thinking about this is to say that as
pollution increases, marginal damage costs rise, which increases the incentive for
society to take actions to reduce pollution (McConnell, 1997).

What evidence exists to support the EKC theory? Empirical evidence for the EKC has
been found in a number of studies, mainly for local and regional pollutants. These include
those looking at SO,, urban emissions of particulates, and hazardous waste sites. The level
of income per capita where the ‘turning point’ is reached (Y* in Figure 6.1) varies across
these studies. For example, Grossman and Krueger (1995) found a turning point for SO,
and particulates at $4,000-$6,000/year GNP/capita, when using air-quality data from
around the world. Markandya et al. (2006) looked at data on SO, and growth from 1870 for
twelve European countries: they found a turning point in most (eight out of twelve) cases.
They also found that the EKC has shifted over time, mostly in the direction of a lower turn-
ing point. Others, though, find empirical evidence against the EKC hypothesis. This seems
often to be the case for CO, and other global/long-term pollutants, as well as for energy use
and solid waste. For example, Cole et al. (1997) found no EKC for traffic, nitrates, or meth-
ane. Other researchers have found that the turning point is very high relative to current
mean incomes. Richmond and Kaufmann (2006) found no turning point for CO, in non-
OECD countries, and Dijkgraff and Vollerbergh (2005) cast doubt on the likelihood of
rising GDP leading to falling CO, emissions even in OECD economies. Box 6.2 gives some
further evidence on CQ,,.
Why are there differences in these empirical findings? These could be due to three factors.
First, there is the nature of the pollutants studied. EXCs seem more likely to exist for local
and regional pollutants (such as SO, and particulates) and less likely to exist for global pol-
lutants such as CO,. Second, other factors also drive emission levels. These factors have been
found to include trade (the degree of protection); political freedom (e.g. such as measured
by an index of civil liberties); and the effect of economic growth (scale of economy), inde-
pendently of income per capita. Political corruption and rent-seeking can also co-determine
the level of pollution and the level of income at which the EKC reaches a turning point
(Lopez and Mitra, 2000). Third, for some countries and some pollutants, we are not rich
enough yet to be ‘over the hump’. We can also imagine that much EKC analysis has trouble
separating out the effects on pollution of two factors that work concurrently: that economic
growth is occurring, but also that time is passing —itself leading to pollution-reducing tech-
nological advance. Deacon and Norman (2006) and Carson (2010) give a good discussion
and overview ofthe evidence on this and other points.
To conclude, we might imagine a debate between a critic of the EKC hypothesis and a
supporter going something like the following:
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 115

6.2 Testing the Environmental

Kuznets Curve

The standard view on why an EKC might exist is that income growth initially drives up pollution, but that
factors related to continued income growth (such as rising demand for environmental quality) contribute
to a subsequent reduction of pollution at higher income levels. This relationship is typically tested using
‘panel data’; that is, data where we study both the variation in pollution and income across different
countries, but also across time. However, time itself may be related to pollution. For example, as time
passes, technology improves, which might result in cleaner production techniques. How can we be sure
what is driving the patterns contained in the data?
Vollerbergh et al. (2009) examine this question for both CO, and SO.. As they say: ‘... separating the
correlation between pollution and income from the correlation between pollution and time is difficult
because income is (also) correlated with time’. The authors use data on pollution and income from OECD
countries between 1960 and 2000. Performing a ‘traditional’ analysis of the data, controlling for the
effects of time passing and for unobserved country-specific effects, they find a strong U-shape
relationship between per capita income and per capita SO, emissions, with a turning point at around
about the mean income level for the sample. Similar results are also found for CO).
However, they also find that their results are very sensitive to modelling assumptions, which is
unsatisfactory. By imposing the weakest set of restrictions on their models, they are able to show that
rising income always has the effect of increasing emissions of both SO, and CO,—reductions in either
which appear to be due to income rising further are in fact more likely to be due to time effects; and that
this time effect is stronger for SO, than for CO). They attribute these time effects to technological advance
and changes in the structure of OECD economies—for example, due to the decline in iron and steel
manufacturing.

Critic: ‘Although total incomes can rise with economic growth, incomes per capita might
not rise if population is growing faster. Population changes are a very important part of
environmental pressures.’
Supporter: ‘Ah, but higher levels of income usually result in a slowing down of population
growth rates.’
Critic: ‘Not all technological change reduces environmental degradation; for example, new
“exotic” pollutants may be introduced.’
Supporter: ‘Yes, but for certain “classic” pollutants such as SO,, the trends are good.’
Critic: ‘The empirical evidence is that whilst demand for environmental quality goes up as
people get richer, it does not rise as quickly as the demand for other goods. So
consumer pressures will not moderate emissions.’
Supporter: “That is a potential problem, although faster growth will give us more resources
to counter these problems. We need to keep an eye on things, and intervene where
necessary.’
Critic: ‘At the world level, there are awkward problems over whose environmental quality
and whose income we are measuring. Also at world level, the fact that the UK has got
cleaner is partly because we have imported “dirty” products from elsewhere (e.g.
coal-mining): but someone has to do it!’
Supporter: ‘Yes, but trade is good, so we don’t want to restrict it on environmental
grounds. Trade restrictions only end up hurting people more in the long run.’
116 ECONOMIC TOOLS FOR THE ENVIRONMENT

Critic: ‘The world has a limited carrying capacity and thresholds for total global emissions:
once these thresholds/capacities are breached, environmental impacts must increase,
and perhaps exponentially. Such dynamic “surprises” mean that the EKC is a poor
policy guide.’
Supporter: ‘Well, that means we should act with care and not rely on the EKC to sort our
environmental problems out. We need government to intervene where market failures
persist, or where critical environmental limits are being approached.’
Critic: ‘Where environmental effects are irreversible, what if we reach the turning point
after these have been triggered?’
Supporter: “That is one of the difficult cases I referred to above.’

And so on. To conclude, the EKC hypothesis is an interesting empirical phenomenon that
we can observe for some pollutants, although it is difficult to separate out the effects of
income growth from other factors that might have caused pollution levels to change.
However, we should not rely on it as a guide for how we manage the environmental impacts
of growth: we would also be unwise to rely on economic growth to solve our environmental
problems on its own, since solving them typically requires a revision of property rights or a
correction of prices. In a sense, economic growth is both the cause of and cure for environ-
mental problems: growth increases our demands on the environment, but growth also gives
us the time and money to do something about its undesirable side-effects. In other words,
even with economic growth, market failures still remain.

6.4 Broadening the Issue: The Economics of Sustainable

Development
6.4.1 Definitions of sustainable development
‘Sustainable development’ is now a popular phrase. But what exactly does it mean? That is a
difficult question to answer, since sustainable development (SD from now on) means differ-
ent things to different people. People place varying emphasis on the different aspects of SD;
for example, in terms of poverty alleviation compared to environmental management. The
best-known definition is the one given by the UN’s Bruntland Commission in 1987: ‘devel-
opment that meets the needs of present generations without compromising the ability of
future generations to meet their own needs’. Another definition was offered by the
Norwegian economist Ger Asheim: “A requirement to our generation to manage the
resource base such that the average quality of life we ensure ourselves can potentially be
shared by all future generations.’ As Atkinson et al. (2007) note, interest in the concept of
SD is drawn from a ‘broad church’, which implies that many, possibly conflicting, objectives
are brought together under this heading. However, a common feature of many definitions
is that SD is primarily concerned with fairness over time. SD is thus principally an equity,
rather than efficiency issue.
Economists’ views on what constitutes a sustainable development path for an econ-
omy over time may be divided into two broad groups. The first (the outcome approach)
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 117

is concerned with how the economic process affects well-being directly. Such ends-based
definitions of SD include non-declining utility per capita and non-declining consump-
tion per capita. Within a neoclassical model of growth, it is possible to investigate
whether such a sustainable time-path exists. A typical finding is that: (i) optimal eco-
nomic growth implies that consumption will fall at some point in the future, due to dis-
counting; (ii) that there is therefore a trade-off between what is optimal and what is
sustainable; and (iii) that there are many possible sustainable development paths, one of
which will give higher sustainable consumption over time (Pezzey and Withagen, 1998).
The second approach (the opportunity approach) to an economic definition of SD asks us
to consider the means that are available to society to generate well-being or consumption:
its resources. ‘Resources’ consist of physical stocks and the technology that we use to exploit
them. Economists have thought about SD from this viewpoint in terms of capital (Ruta and
Hamilton, 2007). Four forms ofcapital may be distinguished:
1. Produced capital, Kp. This is the ‘capital’ with which most economics students are
familiar. It is comprised of the results of past production, as the excess of output over
consumption. Kp includes machinery, roads, bridges, phone networks, satellites, and so
on, and may be used up in the production of consumption goods and services. This
depreciation needs to be offset with new investment, or else the stock of Kp will decline.
2. Human capital, Kh. This includes all skills and knowledge embodied within people. The
stock of Kh can also depreciate (e.g. if unemployed people lose their skills), and can be
added to through training and education.
3. Social capital, Ks. This has been defined as social networks that facilitate mutually
beneficial collective action (Sanginga et al., 2007). For example, co-operative groups
that manage common-access resources can agree to implement rules for utilizing such
resources (which might include grazing lands, or coastal fisheries) for mutual, long-
term benefits. Social capital can also be viewed as including some measure of the quality
of a country’s institutions: for example, the degree of corruption, political openness, or
the quality of justice. Box 6.3 discusses the measurement of social capital, and how
changes in social capital can be related to changes in environmental quality.
4. Natural capital, Kn. This comprises all gifts of nature, and so includes renewable and
non-renewable energy and material resources; clean air and water; nutrient and carbon
cycles; and biodiversity. Natural capital can clearly be depreciated when, for example, a
non-renewable resource such as oil is used up, when a species dies out, or when the
global stock of atmospheric carbon increases. Investments in Kn would include forest
replanting, cutting emissions of greenhouse gases, and restocking of fisheries.
Economic work on sustainable development typically proceeds from the assumption
that the natural capital stock can be aggregated (added together) in monetary units. In
this way, we would add the dollar value of forest stocks to the dollar value of
agricultural land to the dollar value of oil reserves. Things become more difficult in
practice when we also want to include monetary values ofthe stock of biodiversity, or
carbon sinks. Conceptually, however, we can think of ‘shadow’ prices existing for all
forms of natural capital, which can be used to add together the different elements of
this stock. Such shadow prices would indicate how much better off society would be if
the stock of any capital asset was increased by one unit.
118

We defined social capital as ‘shared norms, trust and social networks’ that help people respond
co-operatively and collectively. A long-standing argument in economics is that, other things being equal,
high levels of social capital lead to greater well-being, since social capital allows people to respond better
to shocks, and since the quality of institutions (e.g. lack of corruption) is related to how much wasteful
effort is expended in avoiding regulation, committing crimes, or gaining rents (Knack and Keefer, 1997).
Some authors have also argued that communities with higher levels of social capital manage
environmental resources (particularly common-property resources) in a better way than areas with low
social capital.
Paudel and Schafer (2009) construct a measure of social capital for parishes in Louisiana, USA, and
relate this index to measures of water quality. Their hypothesis is that variations in water quality over
space and over time will be related to variations in social capital. The social capital index is constructed
from data on membership of civic associations per 10,000 persons over the period from 1988 to 1997.
A panel data model is used to explore the relationship between three measures of water quality (nitrate
levels, phosphate levels, and dissolved oxygen), social capital, and population density. The authors found
no significant relationship between social capital (as measured) and either phosphate or dissolved
oxygen levels. However, they found a significant quadratic relationship between social capital and nitrate
pollution. Interestingly, this showed that higher pollution is associated with lower levels of social capital,
but also with the highest levels of social capital. Pollution was lowest when social capital was at
intermediate levels (i.e. neither high nor low). So, on the basis of this study, the shape of the EKC for social
capital and pollution is the opposite of that proposed between income and pollution.

Table 6.5 shows World Bank estimates of the relative importance of different forms of
capital to a country’s overall wealth (the value of its total capital stocks), divided into pro-
duced capital, natural capital, and ‘intangible capital’—which includes both human and
social capital as defined above. As can be seen, natural capital is relatively more important
in low-income countries as a proportion of total wealth. The share of produced capital is
roughly the same in low-, middle-, and high-income countries. This suggests that the pro-
cess of economic growth involves the transformation of natural capital into human and
social capital.
Thinking about a nation’s total capital stock, two different opportunity-based defin-
itions of sustainability exist. The first, which has become known as weak sustainability,
requires the total capital stock K, where K = {Kn + Kh + Kp + Ks}, to be non-declining.

Table 6.5 Shares of capital in total wealth, 2000

Natural capital Produced capital Intangible capital

share (%) share (%) share (%)
Low-income countries 26 16 59
Middle-income countries 13 19 68
High-income countries 2 17 80
World 4 18 78
Note: ‘Intangible’ includes human and social capital.
Source: World Bank (2006).
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 119

This permits natural capital to be run down (through extracting oil stocks, say) so long as
human, social, or produced capital are increased sufficiently to offset this loss. This view
clearly presumes that we can aggregate the different capital stocks in the same units and
that they are substitutes for each other (Markandya and Pedroso-Galinato, 2007). The
Hartwick rule and the genuine savings measure, both discussed below, rely on this view of
sustainable development. An alternative view has been to maintain that SD requires us to
keep the stock of Kn as non-declining. This might be either in monetary or in physical
terms (Neumayer, 2009). This view has been called strong sustainability, and derives pri-
marily from the view that reductions in Kn cannot be substituted for by increases in other
forms ofcapita (van den Bergh, 2007). A somewhat different position is taken by those who
focus on critical natural capital only. Critical natural capital is the subset of Kn that is
either (i) essential for human survival and/or (ii) not substitutable for by increases in either
other forms of capital. An example might be certain aspects of biodiversity. Modern
schools of thought on sustainability thus revolve around the ease with which natural cap-
ital can be replaced by other forms of capital, in terms of both its direct contribution to
well-being and its role in supporting the production of goods and services (Neumayer,
2009). It turns out to very difficult to establish empirically what this ‘ease of substitutabil-
ity’ might be.
Sustainable development therefore appears to have two distinct economic meanings. In
the outcome approach, it means that consumption or utility does not decline over time. In
the opportunity approach, it means we pass on to future generations at least as much capital
as we have, so that they have no less a chance than us of achieving certain levels of well-
being. In fact, these two approaches can be thought of as two sides of the same coin. Capital
can be interpreted as wealth, which in turn is defined as the present value of future con-
sumption or utility flows. Maintaining capital is thus key to maintaining the flow of well-
being into the future.

6.4.2 Sustainability rules?

A powerful myth in economics is that of the ‘free market’, since all markets operate under
some kind of rules imposed by governments. These rules may be introduced for a wide vari-
ety of reasons, such as protecting the environment, supporting producer incomes, or keep-
ing down prices to consumers. But what kind of interventions by governments might be
beneficial in the case of sustainability?
We have already noted that the optimal time-path of consumption is unlikely to be sus-
tainable, since consumption eventually falls. This path comes about through the maximiza-
tion of the present value (i.e. discounted sum over time) of future consumption. To generalize,
if the market system results in efficient outcomes through utility and profit-maximization,
then it is unlikely to result in a sustainable outcome. With regard to natural capital stock-
based definitions of SD, we have already shown (Chapter 2) how the market system tends to
lead to the overuse of environmental resources; this will mean Kn falling over time in terms
of its physical stocks, whilst the lack of a market value for many environmental resources and
service flows means that this increasing scarcity is not corrected for, even at the level of effi-
ciency, by the market. Market forces might conceivably lead to enough offsetting investment
in other forms of capital even if Kn was falling over time, so that weak sustainability might be
120 ECONOMIC TOOLS FOR THE ENVIRONMENT

achieved even if strong sustainability was not. But what rules could the government impose
on markets if we wanted to ensure a sustainable pattern of development? Two possibilities
exist, dependent on whether one takes a weak or strong view of sustainability.

[Link] The Hartwick rule

In work closely related to the idea of weak sustainability, although it precedes it historically,
John Hartwick showed that an economy dependent on a non-renewable resource as one
input to production could have constant (and therefore non-declining) consumption level
over time provided that it followed a simple rule: to reinvest all rents (the difference between
price and marginal cost) from exploiting the non-renewable resource in produced capital.
This ‘zero net investment’ rule can be shown to result in non-declining consumption under
a range of circumstances (Hartwick, 1977, 1997; Hamilton and Withagen, 2007). The
Hartwick rule is in fact the basis for the genuine savings measure of SD discussed in the next
section. If we adopt a broader notion of capital, then the rents could equally well be invested
in human, social, or natural capital.
Two problems exist with the rule. First, it assumes that utility depends on consumption
only, and that the environment is only important as a source of inputs to production. If the
state of the environment is a direct determinant of utility, then the rule does hold utility
constant even if consumption is not falling—although it is possible to amend the rule to
allow for this (d’Autume and Schubert, 2008). Second, the rule only works if the various
forms of capital are good enough substitutes for each other. We have already mentioned
that this may not be true. However, the rule does suggest that countries which do not rein-
vest sufficient of the rents from natural resource exploitation may be doing so at the expense
of future well-being: Box 6.4 gives some examples of what these losses might be.

[Link] Rules based on the natural capital stock: the safe minimum standard

The ‘strong sustainability’ view is that SD requires a non-declining stock of natural capital,
somehow defined. It is clear that this could be a very restrictive rule for a country to impose
on itself. If no trade-offs are allowed between different components ofKn (wetlands, forests,
etc.), then no economic action that depleted the stock of any component could be allowed,
no matter how large the economic benefits. A way around the potentially high forgone ben-
efits that this implies is the idea of shadow projects (Pearce et al., 1990). This would require
any action that reduces the stock of, say, wetlands, to be offset by a physical project that
generates an offsetting replacement (by creating a new wetland). Costing in such a replace-
ment would be an essential part of cost-benefit analysis. However, considerable practical
and theoretical difficulties exist here: for example, if a 500-year-old oak forest is threatened
with destruction by a development project, does a newly planted oak forest of equivalent
size provide an acceptable offset? What if the new wood is planted on an area of heathland?
Now the area of heathland has decreased, so an offset must be created for this. Finally, for
many environmental effects, no physical substitute exists (e.g. if aunique national park is
threatened). Does this imply that no development is allowed that threatens unique assets?
One answer to these sorts of question is provided by the concept of the safe minimum
standard (SMS). The SMS is usually thought of as a means of assessing a proposed change
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOP 121

As explained in the text, the Hartwick rule states that a country can maintain non-declining consumption
over time if it reinvests rents from non-renewable resource extraction (calculated as the volume of
annual production, valued using the difference between price and marginal cost) in other forms of
capital. This allows the total value of the capital stock to be maintained in terms of its ability to generate
future well-being. An exercise in the World Bank's Where is the Wealth of Nations? report discusses how
much richer a range of countries would be in 2000 if they had followed the Hartwick rule over the period
from 1970 to 2000. The test of ‘how much richer?’ involves comparing the measured (estimated) capital
stock in 2000 with calculations of what the capital stock would have been had each country followed the
Hartwick rule.
In the Hartwick rule scenario, it is assumed that all resource rents were reinvested in produced capital:
investment is set equal to this amount in each year over the period from 1970 to 2000. The resources
covered are oil, gas, coal, bauxite, copper, gold, iron, lead, nickel, phosphate, silver, and zinc. The table
below gives results for a small selection of countries: ‘resource-dependent’ countries, where resource
rents add up to more than 5 per cent of GDP; and others. For the resource-dependent countries, such as
Bolivia, Algeria, and Nigeria, we see that the actual capital stock in 2000 is much lower than it could have
been had these countries followed the Hartwick rule. No countries with resource rents higher than 15 per
cent of GDP have followed the Hartwick rule—all are now poorer in 2000 than they could have been had
they followed this rule. Countries such as Chile and Mexico seem to have roughly followed the rule, since
the difference in the capital stock between the two scenarios is very small. However, there are some
countries, in which GDP does not depend heavily on resource rents, where the actual capital stock in
2000 was higher than it would have been under the rule—Brazil, for example. Such countries seem to
have invested more in their productive capital than the Hartwick rule would suggest. However, even for
these countries, the report points out that the quality of investments may bejust as important as the
quantity:

(1) Country (2) Capital stock in (3) Percentage increase in capital (4) Resource dependence
2000, $billion(1995 stock in 2000 with Hartwick rule, (resource rents as % of
dollars) compared to column (2) GDP)
igeria 33 359 35
Algeria 195 51 Ze
Bolivia 14 116 13
Jamaica 13 40
Chile 151 -3
exiCO Sys: =]
Brazil 1,750 =59
China 2,899 =62 ‘all

Source: Adapted from World Bank (2006: 58-9).

that threatens wildlife or its habitats. It involves first identifying the minimum viable popu-
lation or habitat size for a population, or the minimum required stock of some other natural
asset, or the minimum flow of some ecosystem service. Management alternatives for safe-
guarding this minimum are then identified, and cost estimates made of these actions. If a
proposed development threatens the SMS, then decision-makers are presumed to rule
against it, unless the social opportunity costs of so doing are judged to be too high. These
122 ECONOMIC TOOLS FOR THE ENVIRONMENT

costs should include the costs of future development benefits forgone. For example, Berrens
et al. (1998) describe the means by which an SMS for endangered fish populations in the
Colorado River is identified, and the costs of the management actions needed to defend
these SMS values estimated. In this case, this involved costing restrictions on water abstrac-
tion and hydroelectric power station operation in order to maintain minimum in-stream
flows.
The SMS is attractive as a mechanism that protects important elements of the natural
capital stock. It has also been put forward as a way of preventing ecosystems or biodiversity
from being pushed past thresholds involving irreversible costs to society—for example, for
ocean acidification (falling pH levels in sea water as CO, levels increase). The most obvious
problem with the SMS principle is the difficulty of identifying how society should decide
whether the cost of defending the SMS is ‘too high’. This might be through a referendum or
a decision-maker judgement, although Berrens et al. (1998) suggest that if costs in terms of
lost output lie within the range of historical fluctuations, then the SMS should be automati-
cally safeguarded. Other problems with the SMS include extending the concept to elements
of Kn other than wildlife populations and their habitats; and the difficulty in identifying safe
minimum population/habitat sizes (which might well change over time). The approach also
involves reversing the normal burden of proof in development cases, placing this on the
developer rather than the conservationist to show that SMS-type constraints are not being
violated by a proposal, such as developing a new mining site in Madagascar; or that the costs
of not proceeding with the development are in some sense ‘too big’ to justify forgoing the
new mine (Randall, 2007). The SMS is very much a quantity constraint, imposed on eco-
nomic activity in the name of sustainability to safeguard a subset of the stock of natural capi-
tal: yet it does not address the dilemma of whether it is the value or physical quantity of
natural capital that we should be sustaining, nor the implications of imposing such a rule on
future well-being.
Recently, Rockstrém et al. (2009) have suggested the concept ofasafe operating space for
the planet, a series of thresholds for important ecosystem processes, which the authors
maintain should be viewed as levels that should not be exceeded (see Figure 6.2). They iden-
tify key levels for ‘control variables’ (such as CO, levels) that can be related to each thresh-
old. What they refer to as ‘planetary boundaries are then levels for each control variable that
are a ‘safe distance’ from each threshold. The argument is that these boundaries should be
thought of as constraints that can be related to economic activity—for example, in terms of
maximum allowable levels of greenhouse gas emissions, or maximum allowable rates of
habitat loss. This seems similar in spirit to the idea of safe minimum standards discussed
above, although it is interesting that the authors do not suggest that any consideration ofthe
costs of staying within these boundaries is appropriate. The idea of a safe operating space for
the global economy also reminds us of the idea of environmental ‘limits to growth’, which
was a popular notion in the 1970s.

6.5 Measuring ‘Sustainability’

At the Earth Summit in Rio in 1992, the world’s nations agreed to set about producing
annual statistics on the sustainability of their economies. This resulted in the production
ECONOMIC GROWTH, THE \BLE DEVELOPMI 123

PLANETARY BOUNDARIES
Earth-system process Parameters Proposed Current Pre-industrial
boundary status value

Climate change (i) Atmospheric carbon dioxide

concentration (parts per million
by volume)

(ii) Change in radiative forcing

(watts per metre squared)

Rate of biodiversity loss Extinction rate (number of species

per million species per year)

Nitrogen cycle (part Amount of N, removed from

|}ofaboundary : withthe =_ the atmosphere for human use
(millionsoftonnes peryear)
‘phosphorus cycle)
Phosphorus cycle (part Quantity of P flowing into the
of a boundary with the oceans (millions of tonnes per year)
nitrogen cycle)

Stratospheric ozone Concentration of ozone (Dobson 276 283 290

depletion unit)
Ocean acidification Global mean saturation state of 2.75 2.90 3.44
aragonite in surface sea water

Global freshwater use Consumption of freshwater 4,000 2,600 415

by humans (km? per year)
Change in land use Percentage of global land cover 15 P17 Low
converted to cropland

Atmospheric aerosol Overall particulate concentration in To be determined

loading the atomosphere, on a regional basis

Chemical pollution To be determined

For example, amount emitted to,
or concentration of persistent
organic pollutants, plastics,
endocrine disrupters, heavy metals
and nuclear waste in, the global
environment, or the effects on
ecosystem and functioning of earth
system thereof

Figure 6.2 Critical thresholds and planetary boundaries.

Source: Rockstrém et al. (2009).

ofavery large set of proposals for indicators of SD; for example, by the UN (1995, 2003).
Given that SD is such a broad concept, it is unlikely that any one measure would tell us
all we want to know about the sustainability of the economic-environmental system.
Also, given the complexity of this system and the uncertainties that pervade its interac-
tions, it is probably better to talk of indicators of system performance, rather than exact
measures. Indicators of sustainability have been developed from a number of different
disciplinary perspectives, including economics, ecology, politics, and sociology. We now
discuss two put forward by economists: green net national product and genuine savings.
For an interesting discussion ofthe available indicators, and how these relate to the con-
cepts of wealth, substitutability between types of capital, and the prices used to value
capital changes, see Heal (2012).
124 ECONOMIC TOOLS FOR THE ENVIRONMENT

6.5.1 Green net national product

An extensive literature has emerged on whether the System of National Accounts can be
transformed to produce both a better welfare measure and a possible indicator of SD. GNP
has traditionally been thought of as a welfare measure, and as a measure of national income.
By relating this to the idea of ‘income’ put forward by Sir John Hicks in 1930, some authors
have sought to produce an indicator of SD. Hicks’ view on income was that it represented that
portion of the value of output that could be consumed in any year without reducing one's
wealth (i.e. the potential for future consumption). This clearly has resonances with some defi-
nitions of SD. In this sense, then, an adjusted national income figure would tell us the maxi-
mum level of consumption that was sustainable in any year, in the sense that it leaves enough
of a residual to be invested, which preserves the national capital stock or national wealth
(here, wealth is interpreted as the discounted value of future consumption). If adjusted GNP
were rising, then an economy would have the potential of higher sustainable consumption
levels; as such, a rising ‘green’ GNP would be an indicator of sustainability.
Why, though, is it necessary to adjust the conventional national accounts? Partly this is
because, as noted in Box 6.1, these accounts omit many of the inputs that the environment
provides to the economy, since they are unpriced by the market. When a country depletes
its natural capital, this is typically ignored in the national accounts, even though depreci-
ation of man-made capital is allowed for (to convert from GNP to net national product,
NNP). Calculating green net national product involves correcting for these omissions, and
for other changes that impact on well-being, such as changes in pollution.
The approach draws theoretically on Weitzman (1976). The main intention is: (i) to
include the depreciation of natural capital just as conventional NNP allows for changes in
produced capital; and (ii) to include the value of changes in environmental quality that
impact directly on people’s utility. Greening the national accounts could thus involve
adjustments such as the following:

e For non-renewable resources, deduct from NNP an amount equal to the value of
annual production (less discoveries) multiplied by the difference between price and
marginal costs.
wo For renewable resources, annual production is first deducted from annual growth. This
amount is then valued using the same (price-marginal cost) term.
iy For pollution, deduct an amount equal to the change in the emissions of each pollutant
multiplied by its marginal damage costs.
mn For changes in biodiversity and landscape quality, adjust NNP by an amount showing
people's willingness to pay (WTP) for these changes, since this has direct impacts on
utility.

Ignoring this last correction as being hard to do, this would then give us a green NNP meas-
ure equal to the following:

green NNP = NNP — (p, — mc,)ANR — (p, — mc,)AR — v(AS), (6.1)

where p; and mc, are the price and marginal cost of non-renewable resources, and ANR is
the change in the stock of non-renewables; p, and mc, are the price and marginal cost of
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 125

renewables, and AR the change in their stock; and v is the marginal damage cost for pollu-
tion emissions S. In practice, one would have to aggregate over many different non-renew-
able and renewable resource types, and over many pollutants. If green NNP is rising over
time, then development is judged to be sustainable.
There is certainly disagreement among economists as to exactly how these and similar
adjustments should be made. It is also true that for the adjustments to be correct, the price
and marginal cost values used should be ‘correct’. For example, some have argued that the
prices should be those that result from a competitive, dynamically optimal use of resources;
others that the prices should be those that would hold along a sustainable path. Well-known
problems of property rights mean that using market prices to undertake green adjustments

Boxes
pap anyeegy a
Ne
HOT
NEN tonad al Product an
rn

Pezzey et al. (2006) construct measures of green NNP and genuine savings for Scotland over the period
from 1992 to 1999. The main environmental adjustments to the national accounts which they make are
as follows:

e Estimates of pollution emissions over time for NO,, PM10, SO, CO, CO, and CH, based on sectoral
emission intensities are costed using estimates of marginal damage costs from a range of studies.
® Natural capital changes are valued using estimates of prices and marginal costs to value changes in
stocks of forestry, ocean fisheries, coal, aggregates, and oil.
® The value of environmental amenities in each year is included in green NNP using used hectares
enrolled in agri-environmental schemes, valued using WTP/ha figures taken from the literature.

In addition, they adjust both GS and green NNP for estimates of technological progress, and resource
price changes for oil—giving rise to the use of the term ‘augmented’ in the paper's title.
Scotland is not a highly resource-dependent economy, so the effects of these adjustments on
‘standard’ measures of economic well-being—such as NNP—are not large. Both GS and green NNP show
Scotland to be sustainable over the time period in question, as the following diagram shows:

60,000 4,000
—s# Augmented Green NNP
-e Augmented Genuine Savings 3,500
50,000 +
Cc
2 3,000 c«
= Ss
E 40,000 + =
a. 2,500 &
w
z
fe re)
> 30,000 + 2,000 @
£ ea)
c
z i 1,500 5
20,000 + 2
<
+ 1,000

10,000
; + 500

0 1 ——+ —+t —— ——+ + ~ '

1992 1993 1994 1995 1996 1997 1998 1999
Year

Source: Pezzey et al. (2006).

Genuine savings is positive in all years, and green NNP is rising over time.
126 ECONOMIC TOOLS FOR THE ENVIRONMENT

will mean that the former is unlikely to be true, especially for fisheries, whereas the latter will
almost certainly not be the case. There is also dispute amongst economists about whether
green NNP can indeed be used as a sustainability indicator. Box 6.5 shows example calcula-
tions from Pezzey et al. (2006) for Scotland.

6.5.2 Genuine savings

An alternative economic indicator of SD is the genuine savings concept, put forward by
Pearce and Atkinson (1993). Genuine savings (GS) compares reinvestment in an economy

Gross national saving

From gross national saving the consumption of fixed captial is subtracted to

give the traditional indicator of saving: net national savings.

Net national saving

Current operating expenditures on education are added to net national saving

to adjust for investments in human capital.

The value of natural resource depletion is subtracted. Energy,

metals, minerals, and net forest depletion are included.

Genuine saving

Figure 6.3 How the World Bank calculates genuine savings.

Source: World Bank (2006).
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 127

with depreciation of all forms of capital. In this way, GS measures the value of the net
change, over a period, of all forms of capital in an economy. If we restrict capital to just natu-
ral and produced capital, then GS is often shown as being equal to the following:

GS = S — Ap — An, (6.2)

where GS is genuine savings, S is total (aggregate) savings, Ap is depreciation of produced

capital, and An is depreciation of natural capital. This natural capital depreciation is calcu-
lated in the same manner as the environmental adjustments to NNP noted in equation (6.1),
in that it includes the same components (e.g. reductions in oil reserves) valued in the same
way (multiplying the change in the stock by the rent). Alternatively, we could simply work
out the annual change in each type of capital, and add these together. To be more complete,
we would wish to add changes in the stock of human capital, Kh, to equation (6.2). The World
Bank (2006) do this by adding national expenditure on education to equation (6.2); they also
include changes in certain pollution damages, as shown in Figure 6.3 (although some econo-
mists would argue that pollution damage costs should show up in green NNP, not in GS,
since direct impacts on utility from pollution would impact on the well-being measure that
NNP is supposed to be, whereas GS shows changes in the aggregate capital stock).
The genuine savings measure tests for weak sustainability; that is, it assumes that natural
capital and man-made capital are perfect substitutes for each other. A negative GS indicates
unsustainable development, since it implies that a country is running down its total capital
(perhaps in order to fund current consumption at the expense of future consumption). It is
also an empirical test for whether a country is, on average, following the Hartwick rule. Table
6.6 shows estimates of GS for a range of countries, whilst Box 6.5 shows GS calculations for
Scotland. Note that there are many empirical conceptual and empirical problems with calcu-
lating GS across countries or over time. These include what value to use for the cost of carbon
emissions, and how to measure the effects of technological progress on a country’s wealth.

Table 6.6 Genuine savings estimates for a range of countries, 2000 (all figures are as percentages of gross
national income)

Gross national CFC Education Energy Mineral Net forest PM co, Genuine
savings investments depl. depl. depl. damage damage savings

New Zealand ile, 10.9 6.9 iS 0.1 0.0 0.0 0.4 +11.8
Nicaragua WS: oN 37 0 0.1 0.9 0 0.6 +10.3
Niger 2.6 6.7 Ze) 0 0 4. 0.4 0.4 =(63.//
Nigeria U5)if 84 0.9 50.8 0 0 0.8 0.6 =2319
Norway 369 16.2 6.1 8.0 0 0 0.1 0.2 +18.5
Pakistan IQs 78 22 3.1 0 0.8 1.0 0.9 +8.6
Poland 18.8 (ee 6.3 0.5 0.1 0.0 0.7 ill +11.7
Saudi Arabia 29.4 10 TZ 51 0 0 1.0 Ve -26.5
Notes: CFC, depreciation of produced capital; ‘dep|., depletion of energy and mineral reserves and of forest stocks; PM,
particulate matter (PM10).
Source: World Bank (2006).
128 ECONOMIC TOOLS FOR THE ENVIRONMENT

Genuine savings and green NNP are closely linked to each other, since they are both
derived from the same underlying theory. GS is an easier measure to calculate, since it omits
direct impacts of environmental changes on utility. Asheim and Weitzman (2001) show a
theoretical link between GS and utility. However, GS seems to have emerged as a more
popular indicator than green NNP, possibly because its direct link to changes in a nation’s
assets is more intuitively appealing as a measure of sustainability. However, the ability of GS
to predict changes in future well-being over the long run is so far untested.

6.5.3 Other sustainability indicators

As mentioned above, indicators of sustainable development have been developed by many
disciplines, including ecology. These include ecological footprints, human appropriation of
net primary productivity, and the index of sustainable economic welfare (for more details,
see Bohringer and Jochem, 2007; Erb et al., 2009).

Summary
This chapter has been concerned with the implications for the environment of economic
growth, and the extent to which this growth causes undesirable feedbacks for people’s
well-being. These questions have been of interest to economists ever since Adam Smith.
The debate over the links between economic growth, environmental quality, and quality
of life are now largely conducted under the heading of sustainable development. We
reviewed how economists have interpreted this often-nebulous concept, and saw how
economic indicators of weak sustainability can be calculated. Whilst the economic inter-
pretation of sustainability is only one amongst many, it is at least fairly rigorous. It also
recognizes that people’s abilities (reflected in human capital) are as important to securing
sustainable development as safeguarding the environment and investing in a nation’s
produced assets.

Tutorial Questions
6.1 What do we mean by economic growth? Why do economies grow? In principle, what
are the main implications of such growth for the environment?
6.2 What did the early, Classical economists such as Ricardo and Mill have to say about
the links between economic growth and natural resources? Why did economics
become known as the ‘dismal science’?
6.3, What are the theoretical explanations for the environmental Kuznets curve, and why
is empirical evidence on the existence of such a relationship so mixed?
6.4 Explain the main differences between weak and strong sustainability, and why these
matter for government policy on sustainable development.
6.5 What economic indicators of weak sustainability exist, and how much do we know
about the robustness of these indicators in practice or in theory?
ECONOMIC GROWTH, THE ENVIRONMENT, AND SUSTAINABLE DEVELOPMENT 129

References and Additional Readings

Andreoni, J., and Levinson, A. (2001). “The simple Erb, K.-H., Krausmann, FE, Gaube, V., Gingrich,
analytics of the environmental Kuznets curve, S., Bondeau, A., Fischer-Kowalski, M., and
Journal ofPublic Economics 80: 269-86. Haberl, H. (2009). ‘Analysing the global human
Asheim, G., and Weitzman, M. (2001). ‘Does appropriation of net primary production,
NNP growth indicate welfare improvement?’ Ecological Economics 69: 250-9.
Economics Letters 73: 233-9. Ferreira, S., and Moro, M. (2010). ‘On the use of
Atkinson, G., Dietz, S., and Neumayer, E. (2007). subjective well-being data for environmental
Handbook ofSustainable Development valuation, Environmental and Resource Economics
(Cheltenham: Edward Elgar). 46(3): 249.

Begg, D., Fischer, S., and Dorbusch, R. (1987). Gowdy, J., and McDaniel, C. (1999). “The physical
Economics, 2nd edn (London: McGraw-Hill). destruction of Nauru: an example of weak
sustainability, Land Economics 75(2): 333-8.
Berrens, R., Brookshire, D., McKee, M., and
Schmidt, C. (1998). ‘Implementing the safe Grossman, G., and Krueger, A. (1995). ‘Economic
minimum standard, Land Economics 74: 147-61. growth and the environment, Quarterly Journal of
Economics 110(2): 353-77.
Blanchflower, D.G., and Oswald, A.J. (2004). “Well-
being over time in Britain and the USA, Journal of Hamilton, K., and Withagen, C. (2007). ‘Savings
Public Economics 88: 1359-86. growth and the path of utility, Canadian Journal
of Economics 40(2): 703-13.
Bohringer, C., and Jochem, P. (2007). ‘Measuring the
immeasurable: a survey of sustainability indices, Hartwick, J.M. (1977). ‘Intergenerational equity
Ecological Economics 63: 1-8. and the investing of rents from exhaustible
resources, American Economic Review 67(5):
Carson, R. (2010). “The environmental Kuznets
972-4.
curve: seeking empirical regularity and
theoretical structure, Review ofEnvironmental — (1997). ‘National wealth, constant consumption
Economics and Policy 4(1): 3-23. and sustainable development; in H. Folmer and
T. Tietenberg (eds.), The International Yearbook
Clark, A.E., Frijters, P., and Shields, M.A. (2008).
of Environmental and Resource Economics
‘Relative income, happiness and utility: an
(Cheltenham: Edward Elgar).
explanation for the Easterlin paradox and other
puzzles, Journal of Economic Literature 46(1): Heal, G. (2012). ‘Reflections: defining and measuring
95-144. sustainability, Review of Environmental Economics
and Policy 6(1): 147-63.
Cole, M., Rayner, A., and Bates, J. (1997). “The
environmental Kuznets curve: an empirical Jevons, W. S. (1865). The Coal Question (reprinted
analysis, Environment and Development 1965 New York: A.M. Kelly).
Economics 2: 401-16. Knack, S., and Keefer, P. (1997). “Does social capital
Common, M.S. (1988). Environmental and Resource have an economic payoff?’ Quarterly Journal of
Economics: An Introduction (London: Longman). Economics 112: 1251-88.

d’Autume, A., and Schubert, K. (2008). ‘Hartwick’s Levinson, A. (2009). “Valuing public goods using
rule and the maximin paths when the exhaustible happiness data: the case of air quality; NBER
resource has an amenity value} Journal of Working Paper No. 15156.
Environmental Economics and Management Lopez, R., and Mitra, S. (2000). ‘Corruption,
56(3): 260-74. pollution and the Environmental Kuznets
Deacon, R., and Norman, C. (2006). “The Curve’ Journal of Environmental Economics and
environmental Kuznets curve and how countries Management 40: 137-50.
behave? Land Economics 82(2): 291-315. McConnell, K.E. (1997). ‘Income and the demand
Dijkgraff, E., and Vollerbergh, H.R. (2005). ‘A test for environmental quality, Environment and
for parameter homogeneity in CO, panel EKC Development Economics 2: 383-99.
estimates, Environmental and Resource Economics Markandya, A., and Pedroso-Galinato, S. (2007).
32(2): 229-39. ‘How substituteable is natural capital?’
Easterlin, R.A. (1973). ‘Does money buy happiness?’ Environmental and Resource Economics 37:
The Public Interest 30: 3-10. 297-312.
130 ECONOMIC TOOLS FOR THE ENVIRONMENT

Markandya, A., Golub, A., and Pedrosa-Gallinato, S. Svedin, U., Falkenmark, M., Karlberg, L., Corell,
(2006). ‘Empirical analysis of national income R.W,, Fabry, V.J., Hansen, J., Walker, B., Liverman,
and SO, emissions in selected European D., Richardson, K., Crutzen, P., and Foley, J.A.
countries, Environmental and Resource Economics (2009). ‘A safe operating space for humanity,
35: 221-57. Nature 461(7263): 472-5.
Malthus, T. (1798). An Essay on the Principles of Ruta, G., and Hamilton, K. (2007). “The capital
Population (reprinted 1970 London: Penguin). approach to sustainability, in G. Atkinson,
Meadows, D., Meadows, D., Randers, J., and ‘ S. Dietz, and E. Neumayer (eds.), Handbook
Behrens, W. (1972). The Limits to Growth (New of Sustainable Development (Cheltenham:
York: Universe Books). Edward Elgar).

Mill, J.S. (1857). Principles of Political Economy Sanginga, P., Kamugisha, R., and Martin, A. (2007).
(London: J.W. Parker). “The dynamics of social capital and conflict
management in multiple resource regimes,
Neumayer, E.. (2009). Weak versus Strong
Ecology and Society 12(1).
Sustainability (Cheltenham: Edward Elgar).
Seo, S.N., Mendelsohn, R., Dinar, A., Hassan, R., and
Paudel, K., and Schafer, M. (2009). “The
Krurkulasuriya, P. (2009). ‘A Ricardian analysis
environmental Kuznets curve under a new
of the distribution of climate change impacts on
framework: the role of social capital in water
agriculture in Africa, Environmental and Resource
pollution, Environmental and Resource Economics
Economics 43: 313-32.
42(2): 265-78.
UN (United Nations) (1995). Indicators of
Pearce, D., and Atkinson, G. (1993). ‘Capital
Sustainable Development (New York: UN
theory and the measurement of sustainable
Commission for Sustainable Development).
development: an indicator of weak sustainability,
Ecological Economics 8(2): 103-8. —— (2003). Integrated Environmental and Economic
Accounting (New York: United Nations).
Pearce, D.W., Makandya, A., and Barbier, E.
(1990). Sustainable Development (Cheltenham: van den Bergh, J. (2007). ‘Sustainable
Edward Elgar). development in ecological economics) in
S. Dietz and E. Neumayer (eds.), Handbook
Pezzey, J.C.V., Hanley, N., Turner, K., and Tinch,
of Sustainable Development (Cheltenham:
D. (2006). ‘Augmented sustainability tests for
Edward Elgar).
Scotland’ Ecological Economics 57(1): 60-74.
Vollerbergh, H., Melenberg, B., and Dijkgraaf, E.
— and Withagen, C.A. (1998). “The rise, fall and
(2009). ‘Identifying reduced-form relations with
sustainability of capital-resource economies,
panel data: the case of pollution and income,
Scandinavian Journal of Economics 100(2):
Journal of Environmental Economics and
513 —27.
Management 58: 27-42.
Randall, A. (2007). ‘Benefit cost analysis and a safe
UNDP (United Nations Development Program)
minimum standard, in G. Atkinson, S. Dietz,
(1990). Human Development Report: Concept and
and E. Neumayer (eds.), Handbook ofSustainable
Measurement of Human Development (New York:
Development (Cheltenham: Edward Elgar).
UNDP).
Ricardo, D. (1817). The Principles ofPolitical
(2009). Human Development Report (New York:
Economy (reprinted 1926 London: Everyman).
UNDP).
Richmond, A., and Kaufmann, R. (2006). ‘Is there
Warde, P. (2007). Energy Consumption in England
a turning point in the relationship between
and Wales, 1560-2000 (Rome: Consiglio
income and energy use and/or carbon emissions?’
Nazionale delle Ricerche).
Ecological Economics 56: 176-89.
Weitzman, M. (1976). ‘On the welfare significance
Rockstrém, J., Steffen, W., Noone, K., Persson, A.,
of national product in a dynamic economy,
Chapin, ES. IH, Lambin, E.R, Lenton, T.M.,
Quarterly Journal ofEconomics 90: 156-62.
Scheffer, M., Folke, C., Schellnhuber, H.J., Nykvist,
B., de Wit, C.A., Hughes, T., van der Leeuw, S., World Bank (2006). Where is the Wealth of Nations?
Rodhe, H., Sorlin, S., Snyder, P.K., Costanza, R., (Washington, DC: World Bank).
Strategic Interactions and
the Environment

Many of the world’s major environmental and natural resource management problems arise
from interactions between economic agents. Examples include negotiations between coun-
tries about reducing greenhouse gases and marine fisheries where fishers compete with each
other over fish stocks for profits. These types of interactions are recognized by economists
and mathematicians as problems in game theory. Game theory can be applied to games such
as chess and poker, but it can also be applied to make sense of the ‘games’ that governments,
individuals, households, and firms ‘play’ that affect the environment.
This chapter:

e Introduces game theory.

e Discusses an important game called the prisoner’s dilemma.
e Applies the basic model to fishery management.
e Considers how co-operation may emerge through institutions over time.
e Gives an analysis of the formation of alliances using co-operative game theory.
e Analyses a transboundary pollution problem.

7.1. Introduction

At local, international, and global levels, interactions between individuals, firms, and goy-
ernments over environment goods and bads and natural resources involve strategic choices.
Like chess players, the ‘players’ in environmental and natural resource ‘games’ develop strat-
egies to counter their opponents’ strategies. To start to understand these interactions requires
a theory of strategic interaction between decision-makers. In 1944, a mathematician, John
von Neumann, and an economist, Oskar Morgenstern, introduced game theory to econom-
ics. Their approach concerned players who take choices in response to or in anticipation of
what others decide to do. Their models have revolutionized the analysis of strategic interac-
tions between decision-makers from both normative (i.e. what decision-makers should do)
and positive (what decision-makers actually do) perspectives (for an introductory review,
see Dixit et al., 2009).
132 ECONOMIC TOOLS FOR THE ENVIRONMENT

In 1972 Iceland unilaterally extended its Exclusive Economic Zone (EEZ) beyond its territorial waters, in an
attempt to exclude British trawlers and reduce overfishing. It policed its newly introduced catch quotas
through the Icelandic Coastguard cutting the trawl lines of UK vessels. The UK responded by sending in naval
vessels to protect the fishing fleet. The dispute ended in 1976 after Iceland threatened to close a major NATO
base in retaliation for Britain's deployment of naval vessels within the disputed zone. The British government
backed down, and agreed that after 1 December 1976, British vessels would not fish within the zone.
The interaction between the players (the UK and Icelandic governments) involves a set of strategies:
Iceland plays ‘send coastguard boats’, to which the UK government responds with ‘send the navy to protect
trawlers’. Finally, Iceland plays its trump card, ‘close NATO base’, to which the UK responds ‘concede’. What
is driving these responses? The pay-offs are the economic gains to the UK of access to the fishery, whilst the
pay-offs to Iceland are increasing the economic gains from a larger share of a better-managed fishery. In the
final play, Iceland switches strategies, and in the threat to close a NATO base finds an action that would
impose costs on the UK that exceed the losses in profitability of the UK fishing industry from access to the
Icelandic EEZ. In this game, the two players interact repeatedly and the strategies evolve through time.

Economists have applied game theory to environmental and natural resource allocation
problems. In one of the first applications, Levhari and Mirman (1980) analyse interactions
between two countries sharing a fish stock, a so-called ‘fish war’ (see Box 7.1). The ‘war’ is
waged through fisheries policies where fishery regulators decide to be more or less conser-
vationist in their setting of fishing quotas depending on how conservationist they expect
other countries to be. Countries impose an externality on each other by reducing the fish
stock and thus making fish more expensive to catch. Another early application of game
theory in environmental economics was by Maler (1989), who considers international
negotiations to reduce the level of acid rain. When countries negotiate over the total levels
of SO, emissions to be permitted, there is a strategic interaction in which countries benefit
from co-operation (since acid rain has impacts in ‘receiving’ countries as well as in ‘emitter’
countries), but a mechanism has to be found to encourage those countries responsible for
the externality to agree to a reduction in SO,. Maler also found that a cost-effective solution
to reducing European emissions would result in some countries incurring a net loss, and the
countries with a net benefit paying them compensation. A third example is Agenda 21,
agreed at the Rio de Janeiro Earth Summit in 1992 and recently renegotiated in Nagoya in
2010, which aims to protect biodiversity (Barrett, 1994a; Normile, 2010). Biodiversity rep-
resents a global public good, but the countries that benefit most from conservation may be
the richer developed nations, while the less-developed tropical countries that host much of
global biodiversity bear the opportunity costs of conservation through lost marine, agricul-
tural, and forestry output. The issue is how an agreement can be reached that provides an
incentive for these biodiversity host countries to reduce the rate of biodiversity loss.
Game theory has been applied to analyse national environmental problems; for instance,
the strategic interaction between producers over a common-property resource such as
common land grazing (Mesterton-Gibbons, 2000) and the interaction between regulators
and the firms regulated in pollution control (Batabayal, 1995). The key element of all of
these problems—both domestic and global—is that the actions of one decision-maker affect
the welfare of others. All these problems are the subject matter of game theory.
STRATEGIC INTERACTIONS AND THE ENVIRONMENT 133

7.2 Game Theory

7.2.1 Basic concepts
The elements of game theory are as follows. A decision-maker (player) has preferences over
a set of outcomes, and these preferences determine the choices made, but the outcomes
depend upon the choices made by the other players in the game. Game theory has developed
two distinct approaches to analysing such problems. Non-cooperative game theory con-
cerns how players choose strategies, whilst co-operative game theory concerns how players
choose to form alliances. Non-cooperative game theory can be further classified into static
and dynamic games, where static games have only one turn (one shot) and dynamic games
have a number of turns through time. Dynamic games can be further subdivided into
repeated games, where the same game is repeated, and more complex dynamic games,
where the actual game itself changes through time.
Information, or the lack of it, determines how the game is played: games of imperfect
information are those where the players are uncertain about the outcome of a combination
of choices. For instance, in the fishing problem, uncertainty about the fish stock and harvest
means that the profit is uncertain. Games of incomplete information are where players are
uncertain about the preferences of other players. For instance, a regulator may be uncertain
about the cost to a firm of complying with pollution regulation, and therefore be uncertain
about an appropriate level of monitoring (Russell, 1990). Modern game theory makes use of
advanced mathematics, and a general analysis is beyond the scope of this book. Fortunately,
some simple models—for example, the prisoner’s dilemma—offer insights into a wide range
of environmental problems.

7.2.2 The prisoner's dilemma

The prisoner’s dilemma is an important concept in environmental and resource economics.
The original game, called the prisoner’s dilemma by A.W. Tucker in 1950, has the following
form. Two prisoners, Fred and George, have been caught with stolen goods and are suspected
of burglary, but there is insufficient evidence to convict them unless one of them confesses.
The police can convict both of them of the lesser offence of possessing stolen goods without
further evidence. They are interrogated in separate rooms. The prisoners expect the following
outcomes: if they both confess and agree to testify they both get 2 years in prison; if neither
confesses they will both get a 6-month sentence; if one confesses, he will go free, while the
other will get the maximum sentence of 5 years. The ‘pay-offs’ (measured as years in jail)
from this situation are represented in the strategic form of the game given in Table 7.1. In
each of the four cells, the pay-off to George is given first and the pay-off to Fred second.

Table 7.1 The prisoner's dilemma, strategic form

Fred
Confess Deny
George Confess 2 years, 2 years Free, 5 years
Deny 5 years, free 6 months, 6 months
eee
134 ECONOMIC TOOLS FOR THE ENVIRONMENT

How might this game be played? George considers his options: if Fred denies burglary,
then his best response is to confess, and go free, whilst if Fred confesses, then George's best
response is still to confess. Therefore, George concludes that his best strategy is to confess,
as it gives the best outcome whether Fred confesses or denies. We say that ‘confess’ is a
dominant strategy and that the strategy ‘deny’ is dominated. Fred, following a similar line of
reasoning, decides to confess as well. '
This game has been of interest to game theorists and economists because in equilibrium
both players are rational in the way in which they select their dominant strategy, but the
resulting equilibrium outcome gives lower pay-offs to both players than they could have
achieved had they both selected their dominated strategies; namely, ‘deny’. Thus, if Fred and
George had made a binding pact not to confess, before they were arrested, both players
would be better off compared with the dominant strategies. Co-operation requires either
trust or some other mechanism that enforces a co-operative outcome. This is the essence of
the prisoner’s dilemmas that arise in environmental economics: an optimal solution is often
rejected because of distrust or a lack of co-operation between players. Examples of prison-
er’s dilemmas from natural resources and environmental economics include the following:
e Countries that impose an acid rain problem on each other would both be better off
collectively if they could agree to curtail sulphur dioxide emissions. Without
agreement, it is individually rational for a country to only account for its national
external costs instead of international external costs—that is, for the damages it
imposes on others as well as itself.
e Countries are reluctant to sign global agreements to cut greenhouse gases, since the
actions of others to reduce emissions deliver benefits to non-signatories and signatories
alike.
e Urban dwellers who suffer from congested roads and air pollution would be collectively
better off if they used their cars less, but it may still be rational for individuals to not
change their pattern of car use.
® Sheep farmers sharing common grazing land degrade the land by overstocking because
they cannot agree to binding reductions in stocking rates.
e
ip Fishers, who share a common-property marine fishery, overfish because they cannot
devise a way of sharing the benefits of conservation.

How does the dilemma arise? In the acid rain example, the lack of property rights con-
cerns air quality: neither country has a right to control international air quality, even though
it may be in their interests to agree on improvement. Similarly, collective action to limit car
use is often beneficial, but there is a public good aspect to this (see Chapter 2): good air qual-
ity is a pure public good, since it is non-rival and non-excludable. Thus, there is little incen-
tive for an individual to voluntarily limit his or her car use unless he or she could be sure that
everyone else would do the same. Fisheries and common land grazing are overexploited
because they are either common property (shared by a group of owners), or open access,
owned by all (again, see Chapter 2). The problem is conventionally seen as one of an absence
of, or shared rights over, a resource where each firm imposes an externality on other firms
who share the resource. We now consider the last of these examples in more detail and give
a brief introduction to fishery economics.
STRATEGIC INTERACTIONS AND THE ENVIRONMENT 135

7.2.3 Common property-the fisher’s dilemma

Currently, many of the world’s major marine fisheries are in a state of either full exploitation
or depletion (FAO, 2008: 7). The terms ‘full exploitation’ and ‘depletion’ are defined in rela-
tion to the maximum catch that is possible without depleting the fish stock, termed the
maximum sustainable yield. ‘Fully exploited’ indicates that the catch is approximately equal
to the maximum sustainable yield; ‘depletion’ indicates that the catch exceeds the maxi-
mum sustainable yield. Many economically important fisheries, such as the Grand Banks
cod fishery, have been closed to commercial fishing (Kurlansky, 1999). In this section, we
offer an explanation, based on game theory, of how the problem has arisen, but before that
we provide an economic model of fishing.
The fishery model has a biomass growth function—the biological growth of the fish
stock—and a relationship giving catch as a function of fishing effort (a measure of all the
inputs used in fishing, such as the number of boats and the hours at sea) and the stock. To
simplify the model, we assume that the fishery is always run in a ‘steady state’ where the
catch equals biomass growth; this means that the stock remains constant. This allows us to
derive a relationship between the catch and the harvest effort that does not include the
stock, but accounts for the effect of stock size on the catch per unit of fishing effort. In fishery
economics, this is called the yield—effort curve.
In any fishery, the key relationship is how the fish stock changes through time; that is, the
growth function. In our model, fishers share a fishery in which the stock of fish changes
through reproduction, mortality, and growth according to a logistic growth curve given by

2(x) = 1 —x/K)x. (7.1)

In biology, this is a standard way of representing the growth of a population. Here, g(x) is
the growth function, which gives the rate of stock growth per unit of time; x is the stock of
fish, in tonnes; Yis a growth parameter; and K is the carrying capacity of the marine ecosys-
tem. The parameters can be interpreted as follows: the growth parameter, y, gives the rate of
change of the stock when the stock is very low (close to zero). In that case, the rate of stock
growth is approximately yx. Thus if y= 0.5, the rate of stock growth is equal to half the stock
per period. At higher stock levels, the term (1 —x/K) acts to reduce the growth rate, until the
stock reaches a carrying capacity K and the growth rate is zero. The carrying capacity is the
maximum biomass ofa given species that the ecosystem can support. Note that if =0, then
(1 — x/K) = 1; and if x = K, then (1 — x/K) =0.
The firm’s catch (qi) is given by the following function:

Gi= Oe, (7.2)

In economics, this is a production function in which there is a relationship between inputs,
fishing effort h; (measured, for example, as a number of standard trawler days at sea) by
fishing firm i and the stock of fish. Later on, we allow for the possibility that there is more
than one fishing firm. The other parameter, 0, represents how easy fish are to catch. Note
that this equation can rearranged so that q,/h; = Ox; that is, the catch per unit of effort is
proportional to the stock size: the larger the value of 6, the easier fish are to catch. This
relationship also encompasses the status of technology in the fishery, as new fishing tech-
nology makes the fishing effort more effective by increasing the value of @. An important
136 ECONOMIC TOOLS FOR THE ENVIRONMENT

aspect of the production function is that the catch of fish for a given harvest effort depends
upon the fish stock; if there are more fish, it is cheaper to catch a fish.
We now bring together fish biomass growth with harvesting. We do this by assuming that
the fishery operates in a ‘steady state equilibrium’, where the quantity of fish caught equals
the growth in biomass. Thus:

g(x) — q,=0.
Substituting in equations (7.1) and (7.2), the above steady state equation can be written as
follows:

yx(1 — x/K) — 0h,x = 0,

and by solving for x and substituting back into equation (7.2), we obtain the yield effort
curve:

(see Box 7.2 for the derivation). This is a modified version of equation (7.2) that accounts for
the steady state of stock.
We have described the technical aspects of a simple fishery, so now we introduce prices
and an expression of profit to define an economic problem. First, if we multiply the catch by
a constant price p, we obtain the equation total revenue = pq(h,). If we multiply the fishing
effort by the cost per unit of fishing effort, we obtain the total cost: total cost = wh,, where w
is the cost per unit of fishing effort measured as the cost of a trawler day. Profit is the differ-
ence between total revenue and total cost:

1; = pqth;) ma wh,.

To maximize profit, the firm equates the marginal revenue per unit of harvest with the
marginal cost of harvest at h’ in Figure 7.1. This represents a social optimum where the
marginal revenue from fishing equals the marginal social cost. Ifaregulator were to choose
how to manage a fishery to maximize, then a total harvest effort of h’ is likely to be optimal.
It is the maximum level of profit that the fishery can generate, and if firms were able to
co-operate, they would choose this level ofharvesting and share the profit.
However, in an open-access fishery any firm can enter the fishery, attracted by the profits
of those already fishing there. This is a common situation in many of the world’s fisheries.
The harvest effort increases by firms continuing to enter the fishery until the profits are
driven down to zero, at which point there is no longer an incentive for additional firms to
enter. This occurs where effort is h., total revenue equals total cost, and profit is zero. Open
access results in too much fishing effort and too few fish left in the sea.

[Link] A Nash equilibrium with two firms

We have now set up a simple economic model of a fishery and analysed two extremes: one
profit-maximizing firm and a number (possibly a very large number) offirms that drive the
profit down to zero. Now we consider an intermediate case where the resource is shared by
two firms and there is no agreement between them to co-operate. Assume that there are two
identical firms who share the fishery, thus their production functions are identical:
137

Read this box if you would like to understand how the Nash equilibrium is derived. Taking the yield effort
curve shown in equations (7.1) and (7.2), if there are two identical firms in the fishery, the steady state
total quantity caught is as follows:
x1 — x/K) — @xh = 0.

Solving forx gives x = f(h) = K(y— @h)/yand the yield effort curve—that is, the catch as a function of the
harvest effort only (with the stock eliminated)—is:

q = OhK(y— @h)/y.
For two firms, h=(h, +h,), that is total effort in the fishery, using this definition, **(Unclear.}** we give
the stock as a function of the total harvest f(h). For more than two firms, the harvest of all other firms is
given by h_; thus the total harvest is h=h,+h_,. lf this is substituted into the firms’ profit functions:

mi = pOh,f(h) — wh,,

and the firms are linked together by the shared stock through the term x(h). The Nash equilibrium for
firm iis defined where

SL= pot f+ hf (h) —w

where the derivative f(h) =—@K/y,
This gives the response offirm jas,

h; _ Kp@ (y—h_j @)-wy

2Kp6?

This Nash response curve is given in Figure 7.2.

If there is a single firm,

pees y(Kp@ — w)
i 2Kp@

The two firms will continue to adjust their fishing effort along their response curves until the derivatives
of both firms are zero:

pn — Kp@ — w)
' — (3Kp8)
Define

_ 7(Kp@ —w)
= (Kp)
The total fishing effort with both firms following a Nash strategy is

hs = (2/3)

With a single firm, the total effort is h° =(1/2)k. This gives the first result: the single-ownership firm
always puts in less total harvest effort than the two competing firms, the stock from equation (a) is
reduced, and the quantity caught may or may not be reduced. Finally, the total cost is c= w(q/6x); thus
reducing the stock increases the average cost of catching fish.
138 ECONOMIC TOOLS FOR THE ENVIRONMENT

Total cost = wh

Total revenue = p g(x)

hs AN h h

Figure 7.1 Fishery revenue and cost curves.

7,(h, |hy) = pq,(h, |hy) — why; (hp |hy) = pqn(hy |hy) — who.
The important point to note about these equations is that the profit of firm 1 is affected by
the harvest of firm 2, and vice versa. For instance, the term q,(h, |h,) indicates the catch of
firm 1 given the fishing effort of firm 2. The form of this problem allows us to introduce the
Nash equilibrium, a fundamental equilibrium concept in game theory. The profit of one
firm acting independently depends upon what the other fishing firm does; therefore, the
best a firm can achieve is to take the strategy of the other firm as given and maximize its
profit on that basis. To reach a final equilibrium, firms may iterate towards a point at which
neither firm wants to change.
If we choose some specific parameters for the fishery model: K = 1000, y= 1, @=0.1, w=
$2,000 per unit of effort per month, and p = $10,000 per tonne of fish, we can produce the
pay-off matrix shown in Table 7.2.

Reaction curve firm 1

eee

Reaction curve firm 2

Figure 7.2 Fishery Nash equilibrium.

STRATEGIC INTERACTIONS AND THE ENVIRONMENT 139.

Table 7.2 The fisher’s dilemma

Firm 2 ($ million)
Nash (compete) Co-operate
Firm 1 ($ million) Nash (compete) 1.07, 1.07 NN;, NN; SOONG
Co-operate 0:9; 1.35 EN;, CN; WA IAGE, ee,

By considering Table 7.2 with Figure 7.2, we see that the Nash equilibrium represents
the best response to the other player’s expected strategy. However, if both firms were
able to co-operate, possibly through negotiated agreement, they would both be better off
by $133,389. The Nash equilibrium is thus not optimal from the economy’s point of
view. Shifting the solution from the Nash equilibrium to a co-operative solution is said
to be Pareto optimal, in that two firms are made better off and no firm or individual is
made worse off. From Figure 7.2, the harvest rates h; and h are the profit-maximizing
efforts that the firm would choose if they had single and exclusive ownership of the
resource. Reaction curves give the Nash response of one firm to the other firm’s harvest
effort; that is, they give the profit-maximizing harvest effort given the other firm’s har-
vest effort. A Nash equilibrium occurs at e with an effort of h’ + hy. At this equilibrium,
there is no incentive for the firms to choose another strategy. The line from h; to h}
shows a range of optimal ‘co-operative’ solutions. Along this line, the harvest effort is
chosen so that firms maximize their joint profits. The total Nash equilibrium fishing
effort, h’ + hy’, is greater than under single ownership, but the profit is less: therefore,
this represents an inefficient outcome, as noted above. The solution is also illustrated in
Figure 7.1, where hY =h’ + hy. Note also that the fish stock is greater under the profit-
maximizing harvest at h®. A fishery regulator should prefer the profit-maximizing solu-
tion, as it gives the greatest total welfare to producers and is therefore efficient. In a more
general model, it would also maximize welfare to the economy as a whole—that is, pro-
ducers and consumers—from the fishery.
There are two key results that emerge from this analysis. First, if the two firms co-operate,
they stand to benefit by increasing their profit. The second point is that the problem of sub-
optimal exploitation becomes worse as the number of firms increases, until the open-access
equilibrium is reached, at which all firms earn zero profits. This gives a game-theoretic
interpretation of Hardin’s (1968) ‘tragedy of the commons’. A prisoner's dilemma charac-
terizes the outcome for two or more firms, but the problem becomes more severe when
there are a large number of firms. With a small number of firms, a co-operative outcome
may emerge as an equilibrium, especially when the game is repeated a large number of
times. We discuss this further below.

7.3 Self-governance—Escaping the Tragedy of the Commons

In his paper “The tragedy of the commons, Hardin (1968) foretells of dire consequences for
common-property resources. However, the situation he is describing is more akin to open-
access than limited-entry common property. Using common grazing as an example, Hardin
predicts that:
140 ECONOMIC TOOLS FOR THE ENVIRONMENT

Each man is locked into a system that compels him to increase his herd without limit—in a world
that is limited. Ruin is the destination toward which all men rush, each pursuing his own best
interest in a society that believes in the freedom of the commons. Freedom in a commons brings
ruin to all.
(Hardin, 1968: 1244)

A different view is offered by Elinor Ostrom (1990), who won the 2009 Nobel Prize for
Economics for her work on common-property resources:
Elinor Ostrom has challenged the conventional wisdom that common property is poorly man-
aged and should be either regulated by central authorities or privatized. Based on numerous
studies of user-managed fish stocks, pastures, woods, lakes, and groundwater basins, Ostrom
concludes that the outcomes are, more often than not, better than predicted by standard theo-
ries. She observes that resource users frequently develop sophisticated mechanisms for decision-
making and rule enforcement to handle conflicts of interest, and she characterizes the rules that
promote successful outcomes.
(Nobel Prize Committee, 2009)

Ostrom (1990) observed that a significant number of common-property resources have

avoided the tragedy of the commons as a result of users developing institutions that increase
the efficiency of resource exploitation. She argues that the predictions of the prisoner’s
dilemma and, more generally, Nash equilibrium are not an inevitable outcome for com-
mon-property resources, because the potential exists for communication between players
before they take their decisions. Government intervention is one way of forcing producers
to co-operate, but this is not necessarily the only way.
In one alternative, firms may be able to co-operate by agreeing to abide by the decisions
of an external regulator or referee, who may even be appointed by the firms and paid a fee.
The referee acts by imposing penalties to ensure that the firms do not play their Nash strate-
gies. This offers an escape route from the inefhicient Nash equilibrium and thus from the
prisoner’s dilemma. Firms now have an incentive to co-operate so long as each firm’s share
of the fee is less than the difference between the co-operative solution and Nash equilib-
rium, which is (1.2 — 1.08) = 0.12 from Table 7.2. The penalty agreed by the firms should be
large enough so that there is no incentive to cheat on the agreement; for instance, an amount
greater than the difference between the profit rates—that is, $133 thousand—should give an
adequate incentive, especially if non-compliance is always detected.
Ostrom (1990) found that voluntary institutions work effectively in managing common-
property resources where a relatively small number of firms share the resource. Common-
property institutions tend to break down when the number of firms involved increases, or when
there is a lack of family and community ties between the appropriators of the resource. Examples
of where voluntary institutions have been successful include Turkish inshore fisheries (see Box
7.3), lobster fisheries in Maine, and irrigation schemes in the US Midwest (Ostrom, 1990).

7.4 Repeated Fishing Games

In Section 7.3, we concluded that the mismanagement of common-property resources is
not inevitable. Some shared resources have been well managed without private ownership.
STRATEGIC INTERACTIONS AND THE ENVIRONMENT 141

‘Box7.3. An Example of Self-organization in aComm

Bodrum is located about 400 km west of Alanya on the Aegean Sea. The inshore fishery (Berkes, 1986) is
relatively small, with about 100 local fishers operating two- to three-person boats. In the early 1970s, the
fishery was in a depressed state. Conflict existed amongst the local fishers due to unrestrained access to
the fishery, and local fishers devoted resources to competing for the best fishing spots, which tended to
increase production costs.
In response to this situation, members of the local fishing co-operative began to experiment with a
system for allotting fishing sites. After almost a decade of refinement, the resulting system is as follows.
Each year, a list of eligible fishers is drawn up. The fishing locations are named and listed. In September,
the fishers draw lots and are assigned a location, but on each day of the fishing season, from September
to May, they shift east to the next location. This gives the fishers equal opportunities to catch the
migratory fish stock.
This system means that no resources are wasted by the fishers fighting over preferred locations; and
the system is self-policing, with the fishers enforcing the system themselves by reporting fishers who are
in the wrong location. The fishery is managed efficiently with the tacit support of the Turkish government,
but no direct policy intervention.

One other explanation is that the players are brought together in long-term competition
rather than a single ‘one-shot’ game such as that analysed in our simple fishing example.
Over time, players may develop a system of sharing the resource by agreement, but agree-
ment would only come about because the firms expect to benefit in the long term from
showing restraint. This describes a repeated game where a sequence of games is played
through time. Repeated games have a much larger number of potential strategies than one-
shot games. There is the potential for observing how the other firms play over time, and for
tacit agreements to co-operate or punish to emerge. If the prisoner’s dilemma is played a
large number of times, then co-operation can emerge as equilibrium. This equilibrium is
reinforced by the threat that if one player stops co-operating, everyone will be punished by
a return to a disadvantageous non-cooperative equilibrium. This is called a ‘tit-for-tat’ strat-
egy, and it was found by Axelrod (1984) to be a frequently selected strategy in experiments
where people play repeated prisoner’s dilemma games.
In many real strategic interactions, such as water resource sharing or negotiations over
fishing rights, games are repeated over and over again. It is an equilibrium to ‘confess’ in the
one-shot prisoner’s dilemma, because there is no possibility of repercussions at a later stage
of the game. The key result in this literature is that when a game is repeated many times, co-
operation may emerge as a competitive equilibrium. However, if the game is repeated just a
few times, then the equilibrium is the same as for the one-shot game. If the game lasts only
a few turns, then the players will reason as follows: in the last stage of the game, the other
player has no incentive to do anything other than not co-operate, because in the last stage
we have a one-shot game. On this basis, moving to the previous stage, there is no scope for
retaliation, so the player chooses a Nash strategy and so on, back to the start of the game.
Thus in this finite game we conclude that the outcome is to play the Nash strategy in all
periods. If the game is repeated indefinitely, discounting ensures that the final period is of
no importance, but the prospect of retaliation is important—in the sense that the optimal
outcome is a policy where each player co-operates until the other deviates and then deviates
for the remainder of the game. The reason for this is that the most the player can gain from
142 -ORICAALC
ONOMIC Moric
TOOLS ce
FOR THE
THE ENVIRONMENT
ENVIRONMENT

deviating is a one-period improvement in his or her pay-off, which is followed by a reduc-

tion in pay-ofts for the remainder ofthe game.
Consider what happens if the fisheries game in Table 7.2 is repeated a large number of
times and the players adopt a tit-for-tat strategy. The pay-offs from the fishery game from
Table 7.4, where CC, indicates that both firms co-operate and gives the pay-off to firm 1,
CN, indicates that firm 1 co-operates while firm 2 plays a Nash strategy and gives the pay-off
to firm 2. The present value offirm 1 co-operating for a long time is discussed in Box 7.4.
For the firms to gain by deviating, this equation would have to be positive. The gain
from deviating lasts for only one period, when they receive (NC, — CC,), but they are then
penalized for all periods after that (NN, — CC,). The present value of punishment forever
is (NN, — CC,)/r, where r is the discount rate (see Box 7.4). From this, it is obvious that

How do we evaluate decisions that give costs and benefits over time? The problem is that $1 today is
worth more than a $1 ina year's time, because we can invest that $1 in the bank or in the stock market
and earn interest. So when evaluating a flow of net benefits (benefits minus costs), we cannot simply sum
the net benefits in each year to give a total net benefit. This total would not take account of the lost
investment opportunities. Instead, we calculate the present value of net benefits, which converts $1 in
future years to its value today. For instance, if we compare $1 today with $1 in a year’s time, the value of
$1 ina year's time needs to be adjusted downwards to account for the fact that $1 today can be invested
and earn annual interest equal to r. Let us suppose that r= 0.1 (10%). Then, after a year, $1 equals (1 +r) x
1 =1.1. The relative value of $1 after a year is, therefore, (1/(1 + r)) =(1/1.1) = 0.9090. In other words, the
present value of $1 after a year is about 91 cents. After 2 years of compound interest, the dollar now has
grown to (1 +r)? x 1 = 1.21, the relative value of a dollar after 2 years is (1/(1 + r)?) = 0.8264. A general
value for the discount factor is 6'= 1/(1 + r)', where dis the discount factor and t is the number of years
into the future.
If a resource or an environmental asset is expected to give a constant flow of net benefits (y,) for the
foreseeable future, then we can actually simplify the discounting formula. The present value of a flow of
income is given by

PV,= Oly, + 64y2 +... + Oyr.

If the income is constant, then the present value can be given as

PV,= dly+ Sy +. . + dy.

We can rewrite this geometric progression as

PV,— 6'PV;-= dly— dy

PV; =y(d! — 6™1)/(] — 81),

Note that if T=o0, 6'+!=0, and thus

PV, =y(d')/(1 — 6!)

=y/r.

Therefore, the present value for a constant income over an infinite period is simply the net benefit
divided by the discount rate.
STRATEGIC INTERACTIONS AND THE ENVIRONMENT 143
deviating from co-operation only pays if the benefits of deviating are very high or the
discount rate is very high. For the example given in Table 7.2, the gain from deviation is
(1.35 — 1.20) = 0.15, but the punishment is (1.20 — 1.07)/r. The discount rate would have
to be 80 per cent (r= 0.8) for deviation to be worthwhile, which implies a very low weight
on future pay-offs. A typical discount rate even for a highly impatient individual would
probably be less than 20 per cent.

7.5 Co-operative Games

Co-operative games are concerned with the formation and stability of coalitions. For instance
in a European Union negotiation over sharing out EU fishing quotas, if France and Spain
form an alliance and agree to work together to get a bigger share of the quotas in negotiations
with other countries and coalitions of countries, this is an example of a co-operative game.
Forming an alliance is a strategy. History also tells us that such alliances are not always stable.
The fact that co-operative game theory focuses on alliances as strategies tends to simplify the
description of the details of the rest of the game; for instance, how much fish is caught by
France and Spain. Instead the focus is on the total pay-ofts of a coalition.
Co-operative games arise where, for instance, instead of competing, players decide to
establish coalitions. In relation to a common property, a group of fishers or graziers may
decide to form a group or coalition that determines how the resource is shared. More for-
mally, co-operative games arise where players can form binding agreements in pre-play
negotiations. Strictly, co-operative games are a special case of non-cooperative games, in
that a non-cooperative game can be extended to include the decision to form a coalition in
which a group of players play as if they are single players. This could occur between fishers
within a producer’s co-operative, where they negotiate collectively for fisheries quotas and
licences, but they then need to decide how they share the gains between co-operative mem-
bers. Co-operative game theory focuses on the pay-ofts that different ‘coalitions’ can achieve,
rather than on the details of how the game might be played. The following is a non-technical
account of the basics of co-operative game theory with reference to the following example
(for more detail, see Hanley and Folmer, 1998).
Suppose that three countries share a groundwater reservoir. They have a choice of acting
individually or collaborating in various coalitions, including a ‘grand coalition’ that includes
all countries. Each country receives the pay-off indicated in Table 7.3.

Table 7.3 The groundwater co-operation game

Coalitions of countries Value of coalitions ($ million)

A 10
B 20
a 30
A,B 50
A,C 60
GiB 70
A,B,C 100
144 ECONOMIC TOOLS FOR THE ENVIRONMENT

Table 7.3 gives the pay-offs to different combinations of countries. Let us define a pay-off
function, v(.), that gives the value of the game to various coalitions; for instance, v({A}) =10
gives the value to country A resulting from ‘going it alone’, v({A, B}) = 50 gives the pay-off
froma coalition between countries A and B, and so on. The next question is which coalitions
of players are likely to form. For instance, v({A, B}) = 50 indicates that a coalition between A
and B has a pay-off of 50 units, while the grand coalition has a pay-off of 100, and thus v({A,
B, C}) = 100. Therefore, in this game the grand coalition gives a bigger pay-off than all sub-
coalitions, but we have to check whether the coalition is stable. In other words, do either A,
B, or C have an incentive to leave the grand coalition?
Now that we have set out the basic structure of co-operative games, it remains to discuss
solution concepts that determine how players divide the benefits of co-operation. The
approach to this is to assume that the grand coalition forms and then assess if a pay-off (S)
(where S is a single player or a group of players) can be set that provides an incentive for the
coalition to continue. The first condition is a ‘budget constraint’:

mA) + m(B) + 2(C) = v({A, B, C}) = 100.

This ensures that the pay-off is shared amongst the players. Next, we need to specify indi-
vidual and group (or coalition) rationality. This assesses whether players are able to achieve
higher pay-offs outside the grand coalition, either individually or in other coalitions.
Individual rationality says that 2(A) = v({A}); in other words, the pay-off received by coun-
try A as part of the coalition must be no less than the amount that country A could achieve
alone. This extends to group rationality as m({S}) = v({S}): thus the pay-off to the subset of
players S under the grand coalition must be greater than the pay-off that could be achieved
by S as a separate coalition, v({S}).
Individual and group rationality defines a set of constraints on pay-ofts that would be
acceptable to all players. The set of all such pay-offs that satisfy individual and group rational-
ity is called the core and sets of pay-offs are called imputations. These concepts can be illus-
trated for our specific game using a diagram. Figure 7.3 shows the shares of the grand coalition
as a triangle. In each corner, one player receives a pay-offof 100 and the others nothing. The
lines across the triangle indicate individual rationality. For instance, v[{A}] = 10; thus in

C
(0, 0, 100)
C vie

A B
(100, 0, 0) (0, 100, 0)

Figure 7.3 The co-operative solution: individual rationality.

STRATEGIC INTERACTIONS AND THE ENVIRONMENT 145

C
(0, 0,100)
v {{8}]
v[{B, C)]
v[fA, J]

(100, 0, 0)

Figure 7.4 The co-operative solution: individual and group rationality.

terms of individual rationality, A must receive a pay-off of at least 10 due to individual ration-
ality, which leaves 90 units to share between C and B. The smaller triangle abc is the set of
pay-offs that satisfies the individual rationality constraints. Turning to Figure 7.4, we now
introduce the group rationality constraints as well as the individual rationality constraints.
This accounts for the pay-ofts that the countries can obtain in sub-coalitions. The core dfg of
the game represents a set of possible pay-offs that satisfy both the individual rationality and
group rationality constraints. The actual solution would be determined by negotiation
between the players and relative negotiating power (see also Box 7.5).
Co-operative game theory is a useful tool in environmental and natural resource economics,
as it helps to explain why groups with similar preferences form alliances and agree to negotiate
together. For instance, alliances have tended to emerge in global climate change negotiations in
Cancun and Copenhagen between countries with similar interests (see Chapter 12). It also
allows us to analyse how stable these alliances are and which alliances may form in the future.

7.6 Game Theory and Transboundary Pollution Control

7.6.1 Introduction

Transboundary pollution concerns emissions that cross international boundaries. We choose

to analyse this problem here because it includes elements of both co-operative and non-
cooperative game theory. Non-cooperative game theory analyses the outcome in the absence
of negotiation, whilst co-operative game theory analyses how countries form coalitions and
how stable these coalitions are. Transboundary pollution problems are of three broad types:

e First, there are unidirectional externalities, where an ‘upstream country’ affects a

‘downstream’ country. This form of externality is characterized by water pollution,
where a country pollutes a river and hence imposes costs on the downstream country.
e Second, regional reciprocal externalities are typical of public goods such as European
air quality, as measured by SO, and NO, levels. The actions of acountry affect not only
its own costs or benefits, but have impacts in other countries as well: emissions of
146 ECONOMIC TOOLS FOR THE ENVIRONMENT

International environmental agreements (IEA), such as the Montreal Protocol, for ozone-depleting
substances and the Kyoto Protocol to limit emissions of greenhouse gases, have been characterized by
protracted negotiations and partial agreements. The lack of a ‘higher authority’ makes international
environmental agreements difficult to negotiate and police. Barrett (1994b, 2005) proposes that
international environmental agreements should be self-enforcing, which means that the group of
countries that sign the agreement have no incentives to leave the agreement and those that are
non-signatories have no incentive to join. The condition for self-enforcement for a group of N identical
countries is similar to the group rationality constraint from co-operative game theory. Countries divide
into signatories (s) or non-signatories (n) to an IEA such that N=n-+s. A coalition of signatories is stable if
the following conditions are satisfied:

incentive to leave the agreement: 2,(n + 1) < z,(s); (1)

incentive to join the agreement: z,,(n) = 1,(s + 1); (2)

where the pay-off to signatories is 7,(s), and the pay-off to non-signatories is 7,(n). A coalition is stable if
there is no incentive for a country to accede to the IEA and no incentive for a country to leave. Condition
(1) above says that there is no incentive to leave the coalition because the pay-off to a signatory is greater
than the pay-off to a non-signatory when the number of non-signatories is increased by 1. Condition
(2) says that there is no incentive to join, as the pay-off to a non-signatory is greater than a pay-off toa
signatory when the number ofsignatories is increased by 1.
The implications of this theoretical model are rather depressing. They imply that self-enforcing IEAs
only include a large proportion of the countries when the benefits of co-operation are relatively small.
Where the benefits of co-operation over non-cooperation are large, then the equilibrium tends to
include only a relatively small proportion of the countries. The implication of this result is that it is going
to be very difficult for countries to agree to IEAs, and we may see partial agreements where one group of
countries joins and another group remains outside the agreement.

sulphur oxides from the United Kingdom acidify UK lakes and streams, but also impact
on Swedish and Norwegian lakes and streams (see Box 7.6).
@ Third, global externalities are subdivided into those that involve physical interactions
between countries and those that do not. For instance, by thinning the ozone layer,
chlorofluorocarbon emissions have the potential to cause detrimental health effects on
most of the human population. Likewise, greenhouse gas emissions will, through global
warming, affect everyone (see Chapter 9). Non-physical effects relate to a range of
goods with non-use values. These bring in issues related to the conservation of global
biodiversity (see Chapter 12).

All of these pollution problems involve a strategic interaction between countries and can be
analysed by game theory. We explore this using an acid rain example.

7.6.2 The acid rain game

This section uses a simple hypothetical two-country ‘acid rain game’ to illustrate the applica-
tion of game theory to transboundary pollution problems. This problem introduces concepts
from both co-operative and non-cooperative game theory. We start off by specifying the prob-
lem. There are two countries, the United Kingdom (subscript 1) and Sweden (subscript 2),
STRATEGIC INTERACTIONS AND THE ENVIRONMENT

BOX 76 the Montreal Protocol

Chlorofluorocarbons (CFCs) have been implicated in depleting the stratospheric ozone shield since the
1970s. The depletion of the ozone layer is a truly global pollution problem in that all countries are likely
to be affected, to some degree, by the health problems that the resulting elevated levels of ultraviolet
light will cause. In September 1988, twenty-four countries signed the Montreal Protocol (Barrett, 2005:
ch. 8) to restrict their production and consumption of CFCs to 50 per cent of 1986 levels by 30 June 1998.
In London during July 1990, fifty-six countries agreed to further tighten restrictions on the use of these
chemicals. This agreement involved the phasing out of halons and CFCs by the end of the twentieth
century. An interesting aspect of this agreement is that a fund of $240 million was established to assist
poorer countries to comply with this agreement. This amounts to a side payment to ensure that a
negotiated settlement is achieved. The restrictions were further tightened at the fourth meeting in 1992
in Copenhagen, with a ban on CFC products brought forward to 1996, from 1999, and a ban on trade in
these substances.
The agreements over the reduction in substances that damage the stratospheric ozone layer represents
a relatively successful international environmental agreement, perhaps because the environmental costs
were potentially large and shared by all countries and the costs, due to the development of new
products, were declining through time. The use of side payments also facilitated the inclusion of poorer
countries in the London agreement. This outcome contrasts with the current state of disagreement over
the right course of action in relation to climate change (see Chapter 9). The stability of the Montreal
Protocol is strengthened by the threat oftrade sanctions if countries are found to be non-compliant or
refuse to sign the protocol. This has been an effective deterrent against free-riding.

both of which generate sulphur dioxide from coal-burning. Emissions from the UK affect
Sweden and vice versa. This is a reciprocal externality. Each country has a benefit-of-emissions
function due to profits derived from burning coal (e.g. for electricity generation) and an exter-
nal cost function due to damages caused by acid rain. These functions can also be given as
abatement cost functions (which represent the emission benefit function) and an abatement
benefit function. To make this example more concrete, we use the specific functional forms
and parameter values in Table 7.4, but, if you prefer, look at the diagrams that come later.
The Nash equilibrium is where each country only takes account of its own external costs.
The equilibrium (national) level of abatement for the UK is where

MAC, (a,) = MBA,(a, |A);

that is, the marginal abatement cost MAC, (a,) is equated with the marginal benefit of abate-
ment in country 1, the UK, given the amount of abatement in (and thus the level of emis-
sions from) country 2, Sweden, MBA, (a, |a,). If the two countries agree to co-operate, then
each country takes account of the other’s benefits of abatement. Thus, for the UK:

MAC,(a,) = MBA,
(a, |a,) + MBA,(a, |a;).

Using the numerical example given in Table 7.5, the results of a Nash strategy and a
co-operative strategy are given in that table and Figure 7.5.
Table 7.5 gives the pay-off and the abatement level for all the combinations of strate-
gies. Starting with the strategy in which both countries co-operate, this gives the highest
overall abatement of 482.8 and the highest aggregate welfare of 38.8. However, without a
binding co-operative agreement, both countries have an incentive to follow an unco-
operative strategy—especially the UK, which is less affected by acid rain than Sweden.
148 ECONOMIC TOOLS FOR THE ENVIRONMENT

Table 7.4 The acid rain game

Private costs and benefits External costs and benefits

Private benefit of emissions: External costs:

Be (e)) a foie) = bye? Cfe (e) == cree

ie?

ej = bp)/2b; Total emissions:

where e,_ is the private benefit-maximizing emission C=€) +
Abatement: 4 =(e *:)) € =e) +e)
Private abatement costs: Aggregate abatement:
C4(a;) = (BE (e; )- B°(e; -a;)) that is, the difference between maximum a=e-(e, +e)
private benefits and benefits with abatement Abatement benefit function:
BA(a) = Ce(e")-C£(e" - a)
Parameter values:
cf =0.02; c§=0.05
ote =? SOF hap, = SIC}
Dj) = 0.15; by = 0115

For the co-operative solution to hold, there would have to be a side payment from Sweden
to the UK to make the agreement stable. This is a concept from co-operative game theory,
where it is necessary to ensure that players receive at least as much through co-operation
as they do from not co-operating. In this example, a side-payment transfer to the UK
means that Sweden gains sufficiently from co-operation to compensate the UK and still
be better off. However, there may be a problem of countries ‘pre-committing’ to side pay-
ments in a believable way, which can harm prospects for co-operative agreements.
However, promises of side payments can be used to entice countries into becoming part

Table 7.5 Transboundary pollution—the acid rain game

Co-operate

UK
Nash Pay-off(£ millions) Pay-off (£ millions)
UK: 27.0 UK: 27.3
Si6ia S$: 6.3
Nanshestoee Gams cee
Abatement (million tonnes SO.) Abatement (million tonnes SO,)
K: 90.9 UK: 89.2 ;
Seeks S$:241.4
UK +S: 318.2 K+S: 320.6
Co-operate Pay-off (£ millions) Pay-off(£ millions)
KizZZ.3 K: 23.4
Sy ses) S:15.4
JK +S: 38.2 UK + S:38.8

Abatement (million tonnes SO.) Abatement (million tonnes SO.):

K: 241.4 UK: 241.4
Si ANSI S: 241.4
K+S: 431.1 UK +S: 482.8
aa a ie ee ee
STRATEGIC INTERACTIONS AND THE ENVIRONMENT 149

MAC,
MBA, + MBA;

0 Nash Co-operate
Pollution abatement

Figure 7.5 Transboundary pollution: the acid rain game

of international environmental agreements (IEAs) on transboundary pollution. Side

payments were used during the negotiation process for the Montreal Protocol on the
phasing out of CFCs. Side payments are also likely to be important in future global agree-
ments on biodiversity conservation, and in greenhouse gas emission agreements.

Summary
In this chapter, we have introduced game theory as an approach to modelling the strategic
interaction of a small number of economic agents. Situations in which small numbers of
agents or coalitions of agents interact arise in environmental economics, where resources
are shared. Environmental resources include global commons such as the atmosphere,
regional air quality, and natural resources such as fisheries and grazing areas. The attributes
of these problems are shared rights of ownership or poorly defined property rights, where
agents compete to appropriate the benefits of a resource. Game theory enables us to analyse
what the outcomes of these interactions might be.
An important game is the prisoner's dilemma, where the equilibrium is one where both
players are worse off than they would be if they were to co-operate. Repeated games offer a
new perspective on the problem, in that co-operation may actually emerge as an equilib-
rium, because—through time—players can punish defections by other players.
Co-operative game theory is about the formation and stability ofcoalitions between play-
ers. The approach can be viewed as complementary to competitive games, in that it consid-
ers only the best that players can achieve in different coalitions and abstracts from the
mechanics of how a solution to a game is derived.
Game theory informs us how environmental conflicts and problems might be resolved,
and it also goes some way towards explaining why problems have arisen in the first place—
in particular, where individuals behave rationally, but counter to the common good.
150 ECONOMIC TOOLS FOR THE ENVIRONMENT

Tutorial Questions
7.1 Why are the predictions of the prisoner’s dilemma so important to natural resource
management?
7.2 Distinguish between an open-access fishery and a common-property fishery. Would
you expect overfishing to be worse when there are fifty firms sharing a fishery or two
firms sharing a fishery?
7.3 Why is Elinor Ostrom more optimistic about the world’s commons than Garrett
Harding?
7.4 Why was the Montreal Protocol successful in reducing CFCs, whilst the Kyoto
Protocol has failed to even get all the countries to agree?
7.5 With reference to co-operative game theory, why might coalitions be unstable in
relation to producer groups (e.g. a fishers’ organization) sharing a natural resource?

References and Additional Readings

Axelrod, R. (1984). The Evolution of Cooperation Kurlansky, M. (1999). Cod (London: Jonathan Cape).
(New York: Basic Books). Levhari, D., and Mirman, L.J. (1980). “The great fish
Barrett, S. (1994a). “The biodiversity supergame, war: an example using a dynamic Cournot-Nash
Environmental and Resource Economics 4: 111-22. solution, Bell Journal of Economics 11: 322-44.
— (1994b). ‘Self-enforcing international Maler, K.-G. (1989). “The acid rain game, in
environmental agreements, Oxford Economic H. Folmer and E. van Ireland (eds.), Valuation
Papers 46: 878-94. Methods and Policy Making in Environmental
—— (2005). Environment and Statecraft: The Strategy Economics (Amsterdam: Elsevier).
of Environmental Treaty-making (Oxford: Oxford Mesterton-Gibbons, M. (2000). An Introduction to
University Press). Game-theoretic Modelling, 2nd edn. (Providence,
Batabayal, A.A. (1995). ‘Leading issues in domestic RI: American Mathematical Society).
environmental regulation—a review essay, Mueller, D.C. (1989). Public Choice II (Cambridge:
Ecological Economics 12(1): 23-39. Cambridge University Press).
Berkes, F. (1986). ‘Local-level management and Nobel Prize Committee (2009). Press Release,
the commons problem: a comparative study [Link]
of Turkish coastal fisheries, Marine Policy 10: laureates/2009/[Link].
215=29,
Normile, D. (2010). ‘U.N. Biodiversity Summit yields
Dixit, A.K., Skeath, S., and Reiley, D.H. (2009). welcome and unexpected progress) Science 330:
Games ofStrategy, 3rd edn. (New York: W.W. 743-4.
Norton).
Ostrom, E. (1990). Governing the Commons
FAO (2008). Review ofthe State ofFishery Resources: (Cambridge: Cambridge University Press).
Marine Fisheries (Rome: FAO),
Russell, C.S. (1990). ‘Game models for structuring
Fudenberg, D., and Tirole, J. (1991). Game Theory monitoring and enforcement systems, Natural
(Cambridge, MA: The MIT Press). Resource Modeling 4(2): 143-73.
Hanley, N., and Folmer, H. (1998). Game Theory and Von Neumann J., and Morgenstern, O. (1944).
the Environment (Cheltenham: Edward Elgar). Theory of Games and Economic Behaviour
Hardin, G. (1968). “The tragedy of the commons’ (Princeton, NJ: Princeton University Press).
Science 162: 1243-8,
Trade and the
Environment

Since 1970, trade liberalization has generated a rapid increase in trade volume and has
driven economic growth and prosperity. At the same time, increased globalization due to
trade is blamed for widespread environmental degradation, especially in developing coun-
tries. This chapter shows how economists have analysed these issues. This chapter:
e Reviews trade theory and shows why economists often promote freer trade.
e Extends a simple trade analysis to include possible impacts of trade on the
environment.
e Considers the empirical evidence on the effects of trade policy on the environment.
e Discusses international trade policy and its role in environmental policy.

8.1 Introduction

International trade, the exchange of goods across national borders, can profoundly change
the production, consumption, and welfare of a country. Inevitably, any change in the spatial
distribution of production and consumption affects the environment locally, nationally,
and globally.
Since Adam Smith in 1776 and David Ricardo in 1817, many economists (Krugman and
Obstfeld, 2009) have believed that countries, as a whole, benefit by an increase in welfare
from freer trade and suffer a reduction in welfare when trade is restricted by policies such as
quotas and tariffs. Cline (2004: 180) estimates that the world would benefit by a 0.93 per cent
increase in gross domestic product (GDP) through a move to freer trade. The modest size of
this increase is probably due to the fact that by 2004, trade was already relatively free
(Krugman and Obstfeld, 2009: ch. 9). This is not to say that freer trade will benefit everyone:
some people engaged in producing tradable goods and services may suffer a reduction in
income as a result of trade liberalization. Trade allows countries to specialize in producing
goods for which they have the lowest opportunity cost. This means that they can achieve a
higher level of consumption than they could achieve without trade. These principles have
To2 ECONOMIC TOOLS FOR THE ENVIRONMENT

underpinned global trade negotiations under the General Agreement on Tariffs and Trade
(GATT) and the World Trade Organization (WTO), which succeeded GATT in 1995 with
the Marrakech Agreement. For the foreseeable future, we can expect to see a continued
expansion in the volume and value of world trade that will tend to increase global produc-
tion and consumption (see Figure 8.1). The relationship between trade and world GDP is
not straightforward, but the fact that trade allows countries to specialize and increase the
value of their output leads to economic growth, as measured by the rate of change in real
GDP per capita (Frankel and Rose, 2005). In turn, as GDP grows then the demand for
imported goods also increases.
Economic growth itself, whether induced by trade or not, can be expected to effect the
environment by changing spatial patterns ofpolluting production and natural resource use,
but it is hard to say whether it will be beneficial or detrimental, as we saw in Chapter 6. On
the negative side, the first law of thermodynamics predicts that more output, by increasing
the quantity of material used, will increase global pollution. The process of trade itself,
which involves the transportation of goods to different countries, is also polluting. Trade
liberalization may lead to production reallocating to countries with lax environmental reg-
ulations, the so-called pollution havens hypothesis (Eskeland and Harrison, 2003; Copeland
and Taylor, 2004; Taylor, 2004). Environmental benefits from increased trade include that
trade may lead to a reallocation of production to countries better able to deal with the envi-
ronmental side-effects of production, whilst increased income may lead to consumers
demanding higher environmental standards. Trade also results in a reallocation of produc-
tion and consumption, moving the location of relatively polluting activities spatially from
importers to exporters (Kander and Lindmark, 2006).
From Chapter 3, there is a clear recommendation that if a country aims to regulate pollu-
tion, it should target policy instruments such as regulation, emission taxes, or tradable emis-
sion permits directly on emissions. Restricting production inputs or outputs is a second-best
policy, because these variables may not be directly related to the level of emissions or to
environmental damage. Trade policy, in terms of tariffs and restrictions on the quantity and
quality of goods sold in a country, deals with production inputs and outputs. Therefore,

300 5 — Trade volume

35941. === World GDP
Sep hin ey Value of trade

o
Zz 150 +
100 +
Oe” Deak ersudl celal era
T = T —

1950 1960 1970 1980 1990 2000 2010

Year

Figure 8.1 World gross domestic product (GDP), trade volume, and value.
Source: World Trade Organization (2010); 2000 = 100.
TRADE AND THE ENVIRONMENT 153

restricting trade in inputs and outputs is a sub-optimal approach to environmental policy.

The fact that trade policy is sometimes justified on the grounds of environmental protection
should be viewed with suspicion: a trade restriction on environmental grounds may be little
more than a disguised measure to protect domestic producers. Examples of trade measures
to protect the environment and consumers include the European Union ban on hormone
treated beefon the basis of the Agreement on the Application of Sanitary and Phytosanitary
Measures under the WTO arrangements in 1996 (Petersmann and Pollack, 2003), which
was deemed unjustified in 2008 by the WTO appeals panel. The 1991 GATT ruling on the
United States—Mexico tuna-dolphin dispute (Esty, 1994), where the United States banned
imports of tuna from Mexico not produced in a way that avoided dolphin deaths, is a similar
case. The 1998 WTO ruling against the USA for banning shrimp exports from countries not
adopting technologies to protect sea turtles is a further example of countries being pre-
vented from using environmentally motivated trade restrictions in a way that might benefit
domestic producers (WTO, 2006).
In a perfect world, a country’s domestic environment should be protected by national
environmental policies, and the global environment protected by international environ-
mental agreements enforced through national policies. Both domestic and global policies
should target emissions, not consumer goods and production inputs. However, in some
situations, restrictions to trade may be warranted as the only available sanction on a
country causing an international environmental problem. For instance, the CITES
(International Convention on Trade in Endangered Species) convention established in
1973 aims to protect species by reducing the demand for live animals and animal prod-
ucts. However, this Convention does not address the source of the problem, which is
often the loss of habitat supporting the endangered species, due to the opportunity costs
of habitat conservation.
The Basel Convention on the Transboundary Movements of Hazardous Waste (ratified
in 1992) aims to protect developing countries from becoming a dumping ground for toxic
waste. This agreement acknowledges a potential failure in some countries to protect the
environment from the dumping of hazardous waste from other countries.
In the remainder of this chapter, we analyse these issues from a number of perspectives. In
Section 8.2, using a simple general equilibrium model we show why gains from trade can
occur. We then introduce domestic environmental policy into this model, to see how the out-
come might be changed. Section 8.3 introduces predictions from a more sophisticated general
equilibrium model and discusses the empirical evidence on links between trade and the envi-
ronment. Section 8.4 discusses multilateral trade treaties and the environment.

8.2 Why do Countries Gain from Trade?

In this section, we introduce a simple general equilibrium (or whole-economy) model to
provide an initial analysis of the effects of environmental policy and trade on the economy.
A general equilibrium approach is warranted as trade liberalization affects the whole econ-
omy. The analysis starts with the idea of absolute advantage and then introduces compara-
tive advantage as the ‘driving force’ of country specialization and trade.
154 ECONOMIC TOOLS FOR THE ENVIRONMENT

Table 8.1 Trade and absolute advantage

Labour input needed per unit of output UK USA

Food 5 (5/2 cloth) 3 (3/6 cloth)
Cloth 2 (2/5 food) 6 (6/3 food)

Note: Opportunity costs are given in parentheses per unit of output.

In 1776, Adam Smith wrote:

What is prudence in the conduct of every private family, can scarce be folly in that of a great
kingdom. If a foreign country can supply us with a commodity cheaper than we ourselves can
make it, better buy it off them with some part of the produce of our own industry, employed in
a way in which we have some advantage. The general industry of the country, being always in
proportion to the capital which employs it, will not thereby be diminished . .. but only left to find
out the way in which it can be employed with the greatest advantage.

(Adam Smith, 1776: book IV, ch. 2)

Here, Smith introduces the idea of absolute advantage; that is, a country specializing in the
good that it can produce at the lowest cost. Consider the example in Table 8.1, in which two
countries produce two goods, food and cloth, using a single factor of production, labour.
Assume that transport costs are zero and that labour can be freely allocated between the two
sectors.
If each country is endowed with 120 units of labour and under autarky (no trade) both
countries, on the basis of what consumers demand, allocate 60 units of labour to each sector,
this means that the United Kingdom will produce 12 units of food and 30 units ofcloth (12,
30) and the USA 20 units of food and 10 units of cloth (20, 10).
If trade now opens up between the countries, which allows the UK to specialize by allocat-
ing all labour to cloth and the USA by allocating all labour to food production, this increases
world production from (32 (food), 40 (cloth)) to (40 (food, USA), 60 (cloth, UK)). As long as
the exchange rate of cloth for food between the two countries is less than the domestic
opportunity cost of the good, there is a gain from trade. The opportunity cost is the amount
of food that must be sacrificed to produce an extra unit of cloth and vice versa. Thus if trade
between the UK and the USA opens at an exchange rate (terms of trade) of between 0.5 units
(the opportunity cost ofa unit of food in the USA) of cloth per unit of food and 2.5 units (the
opportunity cost of food in the UK) of cloth per unit of food, trade will be mutually
beneficial.
A more important concept than absolute advantage is that of comparative advantage,
which was introduced by Ricardo in around 1817 (Sraffa, 1951-73: vol. I). This states that
countries can gain from trade so long as the opportunity cost of producing goods is differ-
ent between potential trading partners, even when there is no absolute cost advantage to
one of the countries. Consider Table 8.2. The UK now has an absolute cost advantage in
both sectors; however, total production can be increased if the UK transfers labour from
food to cloth and the USA transfers labour from cloth to food. The outcome is illustrated in
Figure 8.2.
TRADE AND THE ENVIRONMENT 155

Table 8.2 Trade and comparative advantage

Labour per unit of output UK USA

Food 5 (5/2 cloth) 6 (6/12 cloth)
Cloth 2 (2/5 food) 12 (12/6 food)
Note: Opportunity costs are given in parentheses.

Figure 8.2 shows the production possibility frontiers for the UK (PPF,,) and the USA
(PPFy 4). The consumption possibility frontiers for the UK (CPFy,) and the USA (CPF ys,)
show the possible rates of exchange between the two countries and the global production
possibility frontier. In the absence of trade (autarky), the production possibility frontier also
determines what a country can consume. Consider the diagram for the UK. The production
possibility frontier gives the maximum combinations of cloth and food that can be pro-
duced. The slope of the production possibility frontier also tells us how much cloth the UK
has to give up if it wishes to produce an extra unit of food. This is the opportunity cost of a
unit of food in terms of cloth. How is this calculated? One unit of food requires 5 units of
labour and cloth requires 2 units of labour per unit; therefore an extra unit of food means
that the UK must forgo 2.5 (5/2) units of cloth.
Figure 8.2 also shows the consumption opportunities that a country has with free trade.
The global production possibility frontier (PPF) adds together the UK PPF and the USA PPF.

Cloth

70 Global PPF

28257,

20 24 42 44 Food
Figure 8.2 Comparative advantage.
156 ECONOMIC TOOLS FOR THE ENVIRONMENT

Thus the maximum cloth production is 70 units, which includes 60 from the UK and 10 from
the USA; whilst the maximum food production is 44, which includes 24 from the UK and 20
from the USA. However, which country should specialize in food and which in cloth? It is
now that we compare the opportunity costs in the two countries. The opportunity cost of
food in the USA is 0.5 units of cloth, while in the UK it is 2.5 units of cloth. Thus the USA
specializes in food production and the UK in cloth. At point b in Figure 8.2 on the global
PPF, the USA specializes completely in food and the UK completely in cloth.
Trade opens up between the countries and there is an exchange rate that gives the units
of cloth per unit of food. A country will trade at any exchange rate that is less than the
opportunity cost of production. For instance, the UK will accept an exchange rate for cloth
of less than 2.5 units of cloth per unit of food and the USA will accept an exchange rate of
greater than 0.5 units of cloth per unit of food.
Let us assume that trade opens between the two countries at an exchange rate of 0.7 units
of food per unit of cloth, and that this implies 1.43 units of cloth per unit of food. This
exchange rate is represented by the consumption possibility frontiers (CPF )x., CPFys,4). The
slope of these lines gives the exchange rate as cloth per unit of food. The gains from trade are
measured as the vertical distance between the PPF and the CPF; that is, the extra cloth that
can be consumed for any given level of food consumption.
This section has established that if a country has a comparative advantage in a good, then
it can specialize in that good and gain by trading with another country. The gain is in terms
of increased global output and thus consumption. Next, we consider how this outcome
changes when one or more of the goods generate external costs, such as pollution, in its
production process.

8.2.1 Extending the basic trade model to include the environment

What happens if we now introduce an environmental cost into this model? This could make
the model quite complex, so let us assume that only cloth produced in the UK imposes an
external cost, the impacts of which are confined to the UK, and that everything else is
unchanged. Cloth produces air pollution ey, according to the equation:

Cn = OClOthigg.

Thus without regulation and with trade, the UK specializes in cloth and produces 60 units
of cloth, and thus 300 units of polluting emissions.
Imagine that the UK environmental regulator has specified that 200 units of emissions
from cloth production is an acceptable standard and imposes an emission standard through
a command-and-control policy, an emission tax, or a tradable emission permit scheme.
From Figure 8.3, when the UK pollution regulator restricts the production of cloth to
40 units, this has two effects. First, the production of food in the UK increases as resources are
switched from the ‘dirty’ polluting good (cloth) to the ‘clean’ good, food. Second, gains from
trade for the UK decline as the gap between the UK’s production possibility frontier PPF,,, and
the consumption possibility frontier CPF Ux(reg) Compared to CPF,,x narrows for each level of
food consumption by the UK. The USA is unaffected if we assume fixed terms of trade for
cloth; however, if the reduction in cloth by the UK means that there is an increase in the terms
TRADE AND THE ENVIRONMENT 15/

of trade for cloth (more food is required per unit of cloth), then the CPF, will pivot back
towards the origin, and consumption of cloth at any given level of food production will fall.
It is not possible to conclude from the analysis presented in Figure 8.3 that the pollution
regulation is welfare reducing. This is because the reduction in pollution may be of greater
value to the population of the UK than the reduction in gains to trade. Our conclusions
might also change if pollution is associated with the consumption of a traded good, rather
than its production.

8.2.2 Trade and the environment

We have now established the basic concept of comparative advantage and analysed one pos-
sible effect of domestic environmental policy on trade. Copeland and Taylor (2003) explore
the interactions between trade, growth, and the environment using a general equilibrium
model (whole economy) for a small open (open to trade) economy. From this theoretical
base, they develop a framework for empirical analysis of trade patterns. The full version of
their model is quite complex; therefore our analysis presents a simplified version.
A production possibility frontier, PPF, gives the maximum output of goods that can be pro-
duced in a country given its resources (capital and labour) and technology. Figure 8.3 shows
a more typical PPF than the linear PPFs given in Figure 8.1. As we already know, the slope of
a PPF gives the opportunity cost of good X in terms of good Y; however, in Figure 8.3 the PPF
has a concave shape that indicates that as resources are switched from one good to another,

Cloth

70 Global PPF

60
Global PPF with UK pollution regulation

28.57 |

20 24 42 44 Food

Figure 8.3 The effects of domestic environmental policy on trade.

158 ECONOMIC TOOLS FOR THE ENVIRONMENT

the opportunity cost of producing more is increasing. This is because the resources switched
from producing Yto producing X are progressively less well suited to the production of X.
Where does a country choose to produce on its PPF? Price signals will dictate the mix of
X and Y, and with free trade these prices will be determined on world markets. Here, we
assume that a country aims to maximize revenue or national income given by

G= p,X + p,Y.

National income is represented in Figure 8.4 as the line labelled p, and for any level of
national income this line has the slope —(p,/p,). Thus the highest attainable national income
is where the line p is just tangent to the PPF. In other words, the price ratio that gives the
exchange rate of Y for X just equals the opportunity cost of an extra unit of X in terms of
output forgone of Y. The slope of the PPF is called the marginal rate of transformation. In
Figure 8.4, the equilibrium is a point a. Other solutions, such as any point along the price
line p’, are sub-optimal as national income can be increased by moving out to the frontier,
while output combinations along p” are infeasible.
To introduce the environment into this model, we assume that one good X generates pol-
lution Z, whilst producing the other good Y is clean. Therefore growth—for instance, as a
result of trade liberalization that leads to an increase in the output of X—will increase pollu-
tion, if everything else in the economy is kept the same. Following on from Grossman and
Krueger (1993), Copeland and Taylor (2003) distinguish between different effects of growth
on the economy and environment. The scale effect, or balanced growth effect, involves equal
proportional increases in labour and capital and a rise in overall output. A composition effect,
or capital accumulation, entails an increase in capital and tends to increase output in capital-
intensive dirty industries. Composition effects can also refer more generally to changes in

Figure 8.4 The production possibility frontier and national income.

TRADE AND THE ENVIRONMENT 159

the structure of a domestic economy because of trade, and the effects on pollution that this
leads to: for example, if expanding exporting industries are relatively ‘dirty’ or if trade results
in the import of dirty products and reduction in domestic production (Runge, 1995). The
technique effect involves a policy change that leads to firms switching to less pollution-inten-
sive production techniques. Such a policy change may be induced by increased consumer
income that results in pressure upon politicians to reduce pollution. This is an example of
how trade may induce an environmental Kuznets curve (EKC, see Chapter 6). We consider
each of these effects in turn using diagrams.
The scale of the economy is measured as the total value of X and Y at fixed world prices.
The scale effect is illustrated in Figure 8.5(i). An increase in capital and labour pushes the PPF
outwards, this leads to an increase in output of both goods and an increase in pollution
shown in the bottom panel of the diagram. The equilibrium shifts from a to b and the pollu-
tion increases (downwards in the bottom panel) from Z, to Z,. Note that the price of the
dirty good X is adjusted to account for the costs of paying an emission tax.
The composition effect involves capital accumulation and a shift in share of output from
the clean output Yto the capital-intensive polluting output X. From Figure 8.5(ii), the equi-
librium shifts from a to b and pollution increases from Z, to Z,,.
The technique effect can come from a policy shift that, by increasing an emission tax (say),
induces firms to adopt a technology that reduces emissions per unit of output. This change
in techniques is represented in the lower panel. Changing technique with a constant output
reduces emissions from Z, to Z,, but the resources used for abatement pivot the PPF towards
the origin along the X-axis. This leads to a new equilibrium at b and a lower final level of
pollution with the new technique of Z,.
These three effects measure how the economy changes in response to growth (scale and
composition effects) and policy (technique effects). Technological changes may also result

(i) Scale effect (ii) Composition effect (iii) Technique effects

Figure 8.5 Scale, composition, and technique effects.

160 ECONOMIC TOOLS FOR THE ENVIRONMENT

from other factors (such as exogenous technological change, or changes in consumer prefer-
ences). The model developed by Copeland and Taylor (2003) is also a basis for determining an
optimal level of pollution. In their model, pollution is viewed as an input into the production
of X. It could also be treated equivalently as a joint output, but let us treat it as an input. Like any
other input then, the amount demanded of the pollution input depends upon its price. The
price of emissions is an emission tax (or the price in a tradable emission permit market). The
economy will continue to demand pollution up to the point at which the marginal benefit of
pollution is equal to the emission tax. More specifically, the marginal benefit is the change in
the national income G(p, K, L, z) with respect to pollution where pollution is treated as an input,
shown by G,. The equilibrium demand for pollution is illustrated in Figure 8.6.
Thus we have established a demand curve, but for equilibrium we also need to consider
the marginal costs, or supply, of pollution. The marginal damage MD for a representative
(typical) consumer is given by their willingness to pay (WTP) to reduce emissions by one
unit. As income increases, it is expected that the willingness to pay for MD will increase.
Consumers face a trade-off, as pollution tends to increase their incomes and thus their util-
ity, but by degrading the environment pollution is also utility reducing. Define the total will-
ingness to pay as [Link], where N is the population of the country. In conventional
microeconomics, supply curves for a firm are the marginal cost (above the average variable
cost), whereas here the supply cost is society's marginal damage (cost) from pollution. Thus
society's willingness to allow pollution depends upon the marginal cost of pollution.
As we saw in Chapter 6, the EKC analyses a relationship between income and pollution.
A common assumption is that as growth occurs and income increases pollution initially
rises, but as income increases beyond a level then society prefers a reduction in pollution,
causing emissions to fall. From Figure 8.6, an increase in income per capita would shift the

Pollution tax S
per unit (7)

N- MD,

Figure 8.6 Optimal pollution.

Source: Copeland and Taylor (2003).
TRADE AND THE ENVIRONMENT 161

supply curve for pollution upwards as higher incomes increases the WTP and thus marginal
damages. This is shown in Figure 8.6 as the new equilibrium at (t,, Z,). This demonstrates
the theoretical possibility of an EKC in an open economy. A complication is that growth can
shift both G, and [Link] simultaneously, leading to outcomes that may also increase pollu-
tion. This analysis is found in Copeland and Taylor (2003: ch. 3).

8.3 Empirical Evidence on the Trade Effects of

Environmental Regulation

There have been numerous studies of links between trade and the environment (for reviews,
see Copeland and Taylor, 2003; Sheldon, 2006). Figure 8.7, from Frankel and Rose (2002),
summarizes the links and possible causality, and poses a set of hypotheses relating to links
between trade and the environment.

1. The link between trade openness and real income per capita is expected by comparative
advantage to be strongly positive.
2. As GDP increase represents an increase in output, this is expected to have a positive effect
on pollution (a detrimental effect on environmental quality) due to the scale of production.

Location, Factor
culture endowments

Environmental
regulation

Environmental
quality

Figure 8.7 Possible causal links between trade and the environment.
Source: Based on Frankel and Rose (2002: 9).
162 ECONOMIC TOOLS FOR THE ENVIRONMENT

3. Ifa country achieves a high income per capita, this raises the public’s demand for
environmental quality, and if the country’s political institutions (the ‘polity’) are
effective, leads to tighter environmental regulations. At the same time, pollution per
unit of output may be modified by the composition effect and reduced by the technique
effect. A negative effect of growth on the environmental is expected at low incomes,
whereas the positive effect dominates at high incomes.
4. Ata given level of GDP per capita, are more open economies better or worse for the
environment? This may be viewed as the main trade effect and a test of the effects of
globalization. One hypothesis is a so-called regulatory chill, where governments are
reluctant to introduce new environmental regulation that will make domestic industry
less competitive. This may become a race-to-the-bottom hypothesis, where countries
progressively relax their environmental regulations in competition for inward
investment. An alternative hypothesis is the gains from trade hypothesis (Frankel and
Rose, 2005). Here, countries use open trade to reduce the detrimental effects on the
environment oftheir domestic firms and also to obtain market goods (for an alternative
hypothesis known as the California effect, see Box 8.1).
5. The pollution haven hypothesis predicts that trade openness leads to multinational
companies switching production to developing countries with lax environmental
regulations. This entails capital moving from one country to another.
6. The Porter hypothesis predicts that a tightening of environmental standards will induce
firms to become more efficient and engage in technological advance and innovation
(Porter, 1990) (see Box 8.2).

Copeland and Taylor (2003) use the theoretical model above to develop an econometric
model of pollution that tests whether there is any evidence for the scale, composition, and

The political scientist David Vogel (1995) has coined the phrase ‘the California effect’ to describe the
situation in which environmental standards for a group of trading countries tend to converge upon those
of the country with the highest standards. The 1970 US Clean Air Act Amendments allowed California to
enact stricter automobile emission standards than the rest of the USA. In 1990, Congress brought national
emission standards up to Californian levels and California adopted tighter standards still. American
automobile manufacturers produce vehicles to the California standard to sell to that market, and also in
anticipation that the standards in all states will be increased to the Californian level.
The term ‘California effect’ is used to describe a much broader phenomenon where tighter regulatory
standards are matched in competing countries. The economic explanation of this is that as one country's
product standards are improved, then domestic producers have an initial competitive advantage in the
market, so that other countries have an incentive to increase their standards so that their producers can
compete.
Vogel advocates a laissez-faire policy to product standards where there are incentives for countries to
increase their product standards to the levels required by the richer, greener nations. However, this effect
is only relevant in cases where the environmental problem can be resolved by improving product
standards.
D THE ENVIRONMENT 163

Box 8.2 The Porter Hypothesis

In an article in Scientific American, Porter (1990) challenges the notion that tight environmental policies
inevitably lead to increased production costs, which reduce the competitiveness of a country’s industry
and therefore reduce its export competitiveness. Instead, he claims that tighter environmental standards
trigger innovations that may increase a firm's competitiveness and outweigh short-run costs to firms of
complying with the regulation. The so-called ‘Porter hypothesis’ is supported by evidence from case
studies, which shows that some firms operating under strict environmental regulations have shown
relatively high levels of performance. The empirical evidence in Tobey (1990) and Jaffe et al. (1995)
indicates that there is very little evidence to indicate that tighter environmental regulations significantly
reduce competitiveness, which may be largely explained by the small proportion of costs associated with
complying with pollution standards.
The two reasons why Porter's hypothesis may be correct: first, environmental regulations make firms
aware of opportunities for changing production activities in ways not previously identified; second, firms
subject to stricter environmental standards than their foreign competitors may be at a competitive
advantage when environmental standards are tightened in their competitors’ markets.
Arguments against the Porter hypothesis include: first, why should rational firms need to be prompted
by environmental regulation to find new techniques that improve their competitiveness; second, if
environmental regulations are not tightened, the firm operating under the lax environmental standard
has a permanent competitive advantage.
At present, there is no theoretical justification (Xepapadeas and de Zeeuw, 1999) for the Porter
hypothesis and no conclusive empirical evidence. However, it received widespread attention because it
provided an attractive model to policy-makers attempting to justify tighter environmental regulations to
industrialists.

technique effects. The dependent variable to be explained is the concentration (parts per
million) of sulphur dioxide pollution in a selection of cities from 1971 to 1995. Sulphur
dioxide is chosen as an important pollutant that has detrimental effects on human health
and causes environmental damage through acid rain. Sulphur dioxide is also correlated
with a range of other air pollutants, such as nitrogen oxides and particulates.
The general equilibrium effects mentioned above are measured by a set of proxy variables;
thus the scale effect is measured as the GDP per square kilometre, to measure the intensity of
economic activity. Composition effects are measured by the capital-to-labour ratio; thus if
(K/L) is relatively large, this would tend to favour pollution-intensive industries. The tech-
nique effect is measured indirectly as an income effect. The results are given in Table 8.3 in
elasticity form; that is, as the percentage increase in sulphur dioxide emissions for a | per cent
increase in the value of a variable. Elasticities are calculated at the mean values.

Table 8.3 Elasticities of SO, emissions

Elasticity

Scale effect (measured by GDP per km?) OSS

Composition effect (capital/labour) 0.993
Technique effect (per capita income) —1.577
Trade intensity (exports plus imports)/GDP =01894
i LUE UEEE EERE

Source: Copeland and Taylor (2003: 25, table 7.4).

164 ECONOMIC TOOLS FOR THE ENVIRONMENT

We can see that a 1 per cent increase in the scale of economic activity leads to a 0.315 per
cent increase in sulphur dioxide concentrations. The technique effect indicates a strong
policy response, as income increases on average lead to a reduction in pollution. Similarly,
a country that has higher trade (exports plus imports) as a large proportion of GDP tends to
have lower emissions. The analysis lends little support to the pollution haven hypothesis,
which predicts that with trade liberalization, rich countries will shift dirty industries to poor
countries with lax environmental policy. Instead, the composition effect indicates that dirty
industries are shifting to the developed countries. This is because these industries are capital
intensive, and thus a factor endowment effect is dominating a pollution haven effect. The
overall conclusion of the study is that pollution as measured by sulphur dioxide has been
reduced by more open trade. However, the authors also conclude that there is no simple
relationship between trade, income per capita, and pollution.
The work by Copeland and Taylor (2003) depends critically on the pollutant selected, the
time period(s), and the countries/jurisdictions included. Frankel and Rose (2005) identify a
potential issue in that pollution and trade are determined simultaneously, and therefore it
is not possible to determine directly how trade affects the environment. Using data from
1990, the authors estimate the following function:

Environmental damage = f(GDP percapita,(GDP per capita),

Trade intensity, Polity, Land percapita).

Environmental damage stands for a range of pollutants and environmental effects, such
as deforestation. GDP per capita is a measure of income, the purpose of the squared term
being to pick up an EKC effect, whereby as income increases beyond a point environmental
damage starts to decline. Trade intensity is estimated as the sum of imports plus exports
divided by GDP, and Polity measures how democratic a country is. Finally, Land per capita
is a measure ofacountry’s capacity to assimilate pollution.
The results for three pollutants SO,, NO,, and particulate matter (PM) all lend strong
support to an EKC, in that the square term on GDP is negative and statistically significant:
thus countries start to reduce these pollutants once per capita income reaches a critical
point. The key focus of this analysis is on trade openness: here, the trade intensity tends to
reduce SO, and NO, and has no effect on PM. When the environmental damage variable is
CO,, trade intensity has a negligible effect on emissions, but the EKC never ‘turns down’,
According to the authors, this indicates that a global agreement will probably be needed to
address CO, emissions, as there is no indication that as incomes increase, CO, emissions
will be reduced by domestic environmental regulation.
The results discussed are not able to test for a pollution haven effect; that is, if net exports
of dirty goods are deterred by tougher environmental regulations in developed countries.
The pollution haven hypothesis predicts that pollution-intensive industries will relocate to
developing countries as environmental regulation is made more stringent in the developed
countries (Taylor, 2004). The presence of a pollution haven effect could be a precursor to
pollution-intensive firms shifting overseas. There is evidence for the pollution haven effect
(Levinson and Taylor, 2008) from data for Mexico, Canada, and the USA. The results show
that industries with the largest increase in abatement cost also had the largest increase in net
TRADE AND THE ENVIRONMENT 165

For a multinational firm, a decision to disinvest in a home market and shift production to another country
involves a number of considerations. To illustrate their theoretical model, Eskeland and Harrison (2003)
consider the example of a steel producer. The government introduces tighter environmental regulations;
this shifts the firm's average cost up and will reduce profits in the short run. In the long run (when capital
investments can be made), the firm will consider investing in new steel furnace capable of meeting the
emission standards or will switch investment to an overseas market. This decision will depend upon all
the factors that determine the firm's profitability, including the costs of environmental compliance.
In their empirical analysis, Eskeland and Harrison (2003) take patterns of foreign investment in four
developing countries: Mexico, Morocco, Ivory Coast, and Venezuela. They explain foreign investment as a
function of variables such as abatement costs (in the investing country), market size, wage rates, and
regulatory barriers. The results from a regression analysis yield no evidence that foreign investment in the
selected developing countries relates to abatement costs in industrialized countries that are the source of
foreign investment. Further, they find that foreign plants (as a result of foreign investment) are
significantly more energy efficient and use cleaner types of energy than the domestic companies. The
authors conclude that:

The relationship between investment and regulation is not as simple as assumed in a naive model. It
depends on a number of factors, the combined effects of which may be positive, zero or negative. We
also find that foreign firms are less polluting than their peers in developing countries. This does not in
any way mean that ‘pollution havens’ cannot exist, or that we should cease to worry about pollution in
developing countries. However, our research does suggest that policy makers should pursue pollution
control policy focusing on pollution itself, rather than on investment or particular investors.

(op. cit.;: 22)

The pollution haven hypothesis remains a controversial concept in economics, in politics, and amongst
environmentalists.

imports. This suggests that there is the potential for firms to take production of ‘dirty goods’
to other countries with lower abatement costs (for a study with a different conclusion, see
Box 8.3).
However, Taylor (2004), in reviewing a range of other empirical studies (Ederington
et al., 2004; Elbers and Withagen, 2004; Fredriksson and Mani, 2004), concludes that there
is no strong evidence for the pollution haven hypothesis, and that the decisions by firms to
locate in a country depend on a range of factors—such as labour costs—that are probably
more important than the costs of pollution control.

8.4 International Trade Agreements and the Environment

8.4.1 International trade agreements
Multilateral trade negotiations under the General Agreement on Tariffs and Trade have
their origins after the Second World War, when memories of the damaging protectionism
of the 1930s led the UK and the USA to plan for more stable trade relations. The Bretton
Woods Conference in 1944 proposed setting up three international organizations, the
International Monetary Fund (IMBP), the International Trade Organization (ITO), and
166 ECONOMIC TOOLS FOR THE ENVIRONMENT

the International Bank for Reconstruction and Development (IBRD). While negotiations
were taking place on the form of the ITO, a group of countries recognized the need for
immediate reductions in tariffs. The USA took the lead by drafting a general agreement on
tariffs and trade, and later this was agreed to by twenty-three countries, as the GATT.
Despite subsequent negotiations, the ITO never came into existence, leaving the GATT as
the most important framework for trade relations. In 1995, 100 countries accounting for
80 per cent of world trade were signatories to the GATT. An additional twenty-nine
countries abided by GATT rules.
In 1995, the GATT was replaced by the World Trade Organization, which inherited from
the GATT the following objectives: first, to provide a forum for multilateral trade negotia-
tions; second, to provide a framework for eliminating trade barriers; and third, to provide
agreed rules to reduce unilateral trade-restricting action. These objectives are pursued by a
set of thirty-eight Articles, which form the basis for resolving disputes. These Articles
embody three basic principles:
e Non-discrimination, or the most-favoured nation (MFN) clause—which binds WTO
signatories to treat all sources of imports for a good equally. Thus if one country
reduces the tariff on timber from another country, it must extend this new tariff to all
countries. This principle acts as a disincentive to bilateral trade agreements.
e Reciprocity—the principle that if one country agrees to reduce tariffs, then another
country should agree to reduce tariffs, in order to leave their bilateral balance of trade
unchanged.
e Transparency—a principle that, in most cases, involves the elimination of quantitative
trade restrictions in favour of tariffs. It is argued that tariffs should be clear to exporters
and domestic consumers.

When the GATT was established, there was no explicit reference to environmental issues,
and it is only recently that the GATT and WTO have been asked to adjudicate on barriers to
trade justified on environmental grounds. The GATT/WTO approach is best illustrated with
some examples. Most product-related environmental policies do not conflict with WTO
rules provided that they apply equally to domestic and imported products. This is the princi-
ple of the most favoured nation; thus if Germany requires domestic cars to carry catalytic
converters and seat belts, it can impose these standards upon importers, but it cannot require
imported cars to comply with more stringent standards. Countries can also adopt a range of
policies to protect the domestic environment, including restrictions on emissions to air and
water, but they cannot extend these standards to imported products unless they physically
affect the final product. An imported product should not be treated differently from an iden-
tical domestic product on the grounds that the production process is different.
This basic principle has led countries wishing to restrict trade on the grounds of process
and products to invoke Article XX, which allows for exceptions to general GATT principles
where ‘necessary to protect human, animal or plant life or health’ (Article XX(b)) and where
relating to the conservation of an exhaustible natural resource, if such measures are made in
conjunction with restrictions on domestic production or consumption (Article XX(g)). The
application ofthis article is informative in the case of the dolphin-tuna dispute between the
USA and Mexico.
TRADE AND THE ENVIRONMENT 167

In eastern tropical areas of the Pacific Ocean, schools of yellowfin tuna are often found
swimming beneath dolphins. When the tuna is harvested with purse seine nets, some dol-
phins become trapped and often die unless released quickly. The US Marine Mammal
Protection Act set out fishing methods that the American fishing fleet must comply with
and which extend to other countries fishing in the American zone of the Pacific Ocean. A
country exporting tuna to the USA must prove that it meets the dolphin protection stand-
ards; otherwise, an embargo will be placed on all fish exports from that country.
In particular, the USA banned tuna exports from Mexico and also a number of interme-
diary countries handling Mexican tuna en route to the USA on the grounds that Mexican
fishing methods were not dolphin friendly. Mexico requested a GATT panel in February
1991 and this reported in September 1991. The panel concluded that the USA could not
embargo imports of tuna products from Mexico simply because Mexican regulations on
methods of tuna fishing were less strict than American regulations. However, the USA
could apply its regulations to the quality of the product. It also concluded that GATT rules
did not permit one country to use trade policy as a means of coercing another to enforce its
domestic laws in another country.
This ruling is important in that it reinforced the principle that products should be judged
on their quality alone, not on the process used in their production. However, the panel
deemed it acceptable for the USA to allow advertising to identify brands of tuna as being
‘dolphin-friendly’. Mexico’s complaint was upheld by the panel, but the panel’s report was
not adopted and the disagreement was eventually resolved bilaterally between the USA and
Mexico. Trade disputes due to domestic environmental regulation have also arisen in the
European Union: for an example, see Box 8.4.
In 1995, the World Trade Organization was established as the successor to the GATT,
providing the framework under which multilateral trade negotiations take place. In the pre-
amble to the Marrakech Agreement (1994) that set up the WTO, reference is made to sus-
tainable development and environmental preservation. The WTO has become more involved
in resolving conflicts where trade policy and environmental concerns interact. To this end,
the WTO has established a Trade and Environment Committee to bring environmental and
sustainable development issues into the mainstream of WTO work (WTO, 2004). It remains
to be seen whether it is feasible for a trade organization to become involved in environmen-
tal policy. The WTO makes it clear that it has a limited role in environmental policy: “WTO
Members recognize, however, that the WTO is not an environmental protection agency and
that it does not aspire to become one. Its competence in the field of trade and environment
is limited to trade policies and to the trade-related aspects of environmental policies which
have a significant effect on trade’ (WTO, 2004: 6).

8.4.2 Multilateral environmental agreements and trade

The theoretical models presented in Section 8.2 predict that trade restrictions offer a sec-
ond-best, or sub-optimal, solution to local, transboundary, and global environmental prob-
lems. The first-best or optimal solution is found by either taxing emissions or introducing a
tradable permit system. Multilateral environmental agreements should be coupled with
optimal national environmental policies to achieve the agreed standard. This principle
holds for both transboundary and global pollution problems.
168 ECONOMIC TOOLS FOR THE ENVIRONMENT

Article 30 of the 1957 Treaty of Rome aimed to foster the freedom of trade between the member states
of the European Community by preventing ‘quantitative restrictions on trade’, but Article 36 of the same
treaty states that trade may be restricted on the grounds of‘public morality, public policy or public
security’ or for the ‘protection of health and life of humans, animals or plants’. Thus it allowed countries to
restrict trade so long as it wasjustified on one of these counts. So when is a domestic environmental
policy little more than a form ofdisguised protectionism?
In 1981, Denmark enacted legislation requiring all beer and soft drinks to be sold in a range of reusable
containers accredited by the Danish environmental protection agency. Containers used by exporters to
Denmark often did not meet the strict Danish recycling requirements and their sale was prohibited. The
foreign companies complained that the extra costs involved in modifying their containers and arranging
collection and transportation in Denmark would reduce their competitiveness. Manufacturers took their
case to the European Commission, who ruled that the Danish law contravened Article 30. Not satisfied by
Denmark's modification of the law in 1984, in December 1986 the European Commission took the case
to the European Court of Justice. This step was taken because there was a fear that countries would justify
protectionist laws on environmental grounds.
In September 1988, the Court ruled that Denmark's deposit and return system was legal under Article
36, because there was no alternative means of reducing the amount of waste. However, Denmark was
required to remove the restrictions on the type of container: any container could be used so long as it
was recycled.
This ruling was highly significant: for the first time the Court had sanctioned an environmental
regulation that restricted trade. This led the German government to enact very strict recycling laws, which
were also restrictive to trade. The implication was that, even within the Single European Market, countries
could introduce legislation that restricted trade so long as it wasjustified on environmental grounds. For
a fuller account, see Vogel (1995: ch. 3).

Unfortunately, there is little evidence that national economic policies—even in rela-

tively green developed countries—are close to being optimal, and in many developing
countries environmental policies may be weak or ineffectively enforced. The first-best
optimum is therefore unattainable and regulators are compelled to choose from among
second-best policies an option that, at least partially, achieves environmental objectives.
This argument goes some way to explaining why multilateral environmental agreements
include trade restrictions as the primary policy instrument. Examples include the
Montreal Protocol and CITES, discussed below. Trade flows, because they involve move-
ment across national boundaries, provide an opportunity for monitoring and restricting
the flow of materials. This may curtail the environmental damage due to a particular
product in the absence of effective national policies. We have also seen in Chapter 7 how
difficult it is to set up voluntary international agreements for the control of global pollut-
ants or the protection of global public goods, since countries have an incentive to free-
ride on such agreements. Trade sanctions provide a means of encouraging participation
in agreements and penalizing signatories that step out of line. The following policies pro-
vide examples.
The United Nations Convention on International Trade in Endangered Species of Wild
Fauna (CITES) was ratified in 1973. It regulated trade in species that are threatened with
extinction (Appendix | species) and those species that may become extinct unless trade is
TRADE AND THE ENVIRONMENT 169

strictly regulated (Appendix 2 species). Trade is regulated by a system of export and import
permits: the requirement for an import permit before an export permit is issued is an
attempt to prevent exporters from sending live animals or animal products to countries that
are not party to the convention. This policy aims to reduce the incentive to exploit rare or
endangered species through trade, but can be seen as a second-best substitute for effective
national policies that protect species and their habitats within countries.
The 1987 Montreal Protocol is an application of the 1985 Vienna Convention regarding
substances that cause depletion of the ozone layer. The parties to the original protocol
agreed to reduce chlorofluorocarbons (CFCs) by 50 per cent by 1999. Negotiations in
London in 1990 led to agreement that production of CFCs should cease in 1996. Further
meetings in Copenhagen (1992) and Montreal (1997) increased the list of substances sub-
ject to restrictions (for details, see [Link] The
Copenhagen (1992) meeting agreed not to export or import restricted substances and to
ban imports of products containing CFCs by 1993. This policy includes both agreements
over domestic production and consumption of these products, and is also backed up by
trade measures that prevent countries that have not ratified the treaty from becoming
potential markets for these products. This is an example of where a trade policy is used as an
effective complement to multilateral environmental agreements and effective domestic
policies. Trade policy is only used to prevent non-parties to the agreement eroding the effec-
tiveness of the restrictions agreed, and signatories from defaulting on their commitments.
The Basel Convention on the Control of Transboundary Movements of Hazardous Waste
and their Disposal allows restrictions to trade in hazardous waste. It forces countries to
focus on national solutions to industrial waste problems and protects export countries from
being used as dumping grounds. Thus the convention establishes that exports of waste
should only be with the importing country’s permission (as distinct from importing com-
panies). If an exporting country believes that waste will not be disposed of in an environ-
mentally sound manner, then it should not be exported. Trade with non-signatory countries
is not allowed (UNEP, 2011).

Summary
Trade flows have a profound effect on national economies and must therefore have a signifi-
cant effect on national environments. Simplistically, trade increases economic growth, pro-
duction, and consumption, and due to the fact that more production requires more natural
resources, an increase in trade is likely to lead to an increase in pollution. However, increased
trade also leads to a specialization of production, which may occur in countries where
inputs are used more efficiently or in ‘pollution havens’, countries with lax environmental
standards. The empirical evidence suggests that the effects of stringent environmental
standards on a country’s competitiveness is rather slight, and that countries with lax envi-
ronmental standards will not necessarily become pollution havens because firms place more
importance upon factors other than environmental costs when deciding where to locate.
From the theoretical and empirical analysis, there is no clear justification for restricting
trade on environmental grounds. Free trade for most countries is welfare-increasing and, so
long as countries have effective domestic environmental policies, there is no reason why it
170 ECONOMIC TOOLS FOR THE ENVIRONMENT

should be more damaging to the environment than restricted trade. Trade policies are
almost always sub-optimal instruments for protecting a country’s environment and should
be replaced by domestic environmental policies. The question arising from this is thus why
are trade restrictions used as a major component of environmental policies? First, environ-
mental justifications for trade restrictions are sometimes little more than covert protection-
ism. Second, in the case of international environmental agreements, they are a sub-optimal
alternative where the failure of countries’ domestic environmental policies may have conse-
quences on the global environment. Examples include, the Montreal Protocol, CITES, and
the Basel Convention on Hazardous Waste. In all these cases, using trade restrictions might
be the best approach that is available.
Finally, in a world in which countries compete both for inward investment and for global
market shares, it is not surprising that environmental policies are sometimes used as ways
of helping to achieve these objectives. But this could be consistent both with an increasing
level of environmental protection (competing for ‘green’ consumers or green inward invest-
ment) and a lowering of standards, as for the pollution havens hypothesis.

Tutorial Questions
8.1 You have been invited to take part in a debate with the proposal that ‘Globalization is
Bad for the Environment’. What would your three main points be if you were to argue
for the proposal? Which three arguments would you have against the proposal?
8.2 Distinguish between the pollution haven hypothesis and the Porter hypothesis in
relation to trade and the environment.
8.3 When is it justifiable to use trade tariffs and trade quotas as a substitute for domestic
environmental policy?
8.4 Distinguish between the scale effect and the composition effect of trade liberalization.
8.5 What is the link between trade and the environmental Kuznets curve?

References and Additional Readings

Cline, W. (2004). Trade Policy and Global Poverty Contributions to Economic Analysis and Policy
(Washington, DC: Institute for International 3(2), Article 3.
Economics).
Eskeland, G.S., and Harrison, A.E. (2003).
Copeland, B.R., and Taylor, M.S. (2003), “Trade ‘Moving to greener pastures? Multinationals
growth, and the environment} Journal of and the pollution haven hypothesis, Journal of
Economic Literature 62: 7-71. Development Economics 70: 1-23.
— and — (2004), Trade and the Environment: Esty, D.C. (1994). Greening the GATT: Trade,
Theory and Evidence (Princeton, NJ: Princeton Environment, and the Future (Washington,
University Press). DC: Institute for International
Ederington, J., Levinson, A., and Minier, J. (2004). Economics).
“Trade liberalization and pollution havens, Frankel, J.A., and Rose, A.K. (2002), ‘Is trade good
Advances in Economic Analysis and Policy 4(2), or bad for the environment? Sorting out the
Article 6. causality, NEBR Working Paper No. 9201.
Elbers, C., and Withagen, C. (2004). ‘Environmental — and —— (2005). ‘Is free trade good or
policy, population dynamics and agglomeration, bad for the environment? Sorting out the
TRADE AND THE ENVIRONMENT 171

causality, Review of Economics and Statistics 87: Runge, C.F. (1995). “Trade pollution and
85-91. environmental protection, in D.W. Bromley
Fredriksson, P., and Mani, M. (2004). ‘Trade (ed.), The Handbook ofEnvironmental Economics
integration and political turbulence: (Oxford: Blackwell).
environmental policy consequences), Sheldon, I. (2006). “Trade and environmental policy:
Advances in Economic Analysis and Policy 4(2), a race to the bottom?’ Journal ofAgricultural
Article 3. Economics 57(3): 365-92.
GATT Panel (1987). ‘United States—taxes on Smith, A. (1776). The Wealth of Nations, book IV: 2
petroleum and certain imported substances (New York: Modern Library edition).
1987. Report of the Panel, 17 June 1987; GATT Sraffa, P. (1951-73). The Works and Correspondence
Doc.L/[Link] 348/136 (Geneva: GATT).
of David Ricardo, 11 vols (Cambridge: Cambridge
Grossman, G.M., and Krueger, A.B. (1993). University Press).
‘Environmental impacts of aNorth American free Taylor, M.S. (2004). “‘Unbundling the pollution haven
trade agreement; in P.M. Garber (ed.), The US-
hypothesis, Advances in Economic Analysis and
Mexico Free Trade Agreement (Cambridge, MA: Policy 4, Article 8.
The MIT Press): 13-56.
Tobey, J.A. (1990) “The effects of domestic
Hudec, R.E. (1996). ‘GATT legal restraints on the use environmental policies on patterns of world
of trade measures against foreign environmental trade: an empirical test, Kyklos 43: 191-209.
practices; in J. Bhagwati and R.E. Hudec (eds.),
UNEP (United Nations Environment Programme)
Fair Trade and Harmonization: Prerequisites
for
(2011). Basel Convention, [Link]
Free Trade? (Cambridge, MA: The MIT Press):
convention/[Link] (accessed 24 June 2011).
95-174.
Vogel, D. (1995). Trading Up: Consumer and
Jaffe, A.B., Peterson, S.R., Portney, P.R., and Stavins,
Environmental Regulation in the Global Economy
R.N. (1995). ‘Environmental regulation and the
(Cambridge, MA: Harvard University Press).
competitiveness of US manufacturing: What
does the evidence tell us?’ Journal of Economic WTO (World Trade Organization) (1997). “Taxes
Literature 33: 132-63. and charges for environmental purposes—border
tax adjustments, Committee on Trade and the
Jaunky, V.C. (2011). “The CO, emissions—income
Environment, WT/CTE/W/47 (Geneva: WTO).
nexus: evidence from rich countries, Energy
Policy 39: 1228-40. — (2004). Trade and the Environment (Geneva:
WTO).
Kander, A., and Lindmark, M. (2006). ‘Foreign trade
and declining pollution in Sweden; Energy Policy — (2006). ‘India etc. versus US: “Shrimp-turtle” 5
34: 1590-9. Environment: Disputes 8, [Link]
english/tratop_e/envir_e/edis08_e.htm (accessed
Krugman, P.R., and Obstfeld, M. (2009).
1 May 2011).
International Economics: Theory and Policy, 8th
edn. (Boston, MA: Pearson Addison-Wesley). — (2010). ‘Australia—Measures affecting the
importation of apples from New Zealand, Dispute
Levinson, A., and Taylor, M.S. (2008). ‘Unmasking
Settlement: Dispute DS367, [Link]
the pollution haven effect’, International Economic
english/tratop_e/dispu_e/cases_e/ds367_e.htm
Review 49: 223-54.
(accessed 1 May 2011).
McGuire, M.C. (1982). ‘Regulation, factor rewards
Xepapadeas, A., and de Zeeuw, A. (1999).
and international trade} Journal of
Political ‘Environmental policy and competitiveness:
Economy 17: 335-54. the Porter hypothesis and the composition of
Petersmann, E., and Pollack, M.A. (2003). capital’? Journal of Environmental Economics and
Transatlantic Economic Disputes: The EU, the US, Management 37: 165-82.
and the WTO (Oxford: Oxford University Press).
Porter, PR. (1990). ‘America’s green strategy,
Scientific American 264: 168.
Part Il

Applying the Tools

The Economics of
Climate Change

We first explore the idea of climate change as a global environmental risk. We then discuss
the challenge of finding an international agreement on how to control climate change risk.
The next two sections examine the costs and benefits of international co-operation to con-
trol carbon emissions. Finally, we discuss the economic trade-off that exists in climate strin-
gency and flexibility to hit this given emissions target.

9.1 Introduction

Imagine an invisible quilt covering the earth. Its warmth allows us to grow food, build shel-
ters, and clothe ourselves. But a potential problem exists. Natural scientists warn us that the
earth is getting warmer, and human actions have impacted the climate to its detriment:
developing land, raising livestock, and burning fossil fuels are disrupting the planet’s atmos-
phere—and the consequences could be devastating. The scientists base their argument on
observed trends. First, the earth has warmed by 0.5°C, or 1°F, over the past 100 years. At the
same time, atmospheric concentrations of greenhouse gases have increased by about 30 per
cent over the past 200 years (see Table 9.1). These greenhouse gases consist principally of
CO,, methane, and N,O; the latter two are expressed as ‘carbon equivalents’ when we meas-
ure a country’s contribution to global warming.
Science has made a connection between these trends—or more precisely, scientists can-
not reject the hypothesis that human-driven climate change exists today. This view is
reflected by the scientists working for the Intergovernmental Panel on Climate Change
(IPCC)—the international panel that has been assessing climate change for past two dec-
ades. The IPCC has concluded that the balance of evidence points towards human influ-
ence on the global climate. Many natural scientists and policy-makers continue to advocate
for a worldwide reduction in greenhouse gas emissions. The implications of climate change
policy are significant: restrictions on carbon emissions will affect everyone on the planet.
Today, climate change dominates environmental policy discussions. People need to
understand and address the risks posed by climate change and the choices we have to reduce
them. Economics offers a unique perspective to help frame the discussion about the costs
and benefits of climate change policy. Investments to protect human and environmental
176 APPLYING THE TOOLS

Table 9.1 Total carbon dioxide emissions of Annex 1 parties in 1990, for the purposes
of Article 25 of the Kyoto Protocol
a

Party Emissions (Gg) Percentage

Australia 288,965 Di

Austria 59,200 0.4

Belgium 113,495 0.8
Bulgaria 82,990 0.6
Canada 457,441 23
Czech Republic 169,514 ez
Denmark 52,100 0.4
Estonia SE 0.3
Finland 53,900 0.4
France 366,536 ay
Germany 1,012,443 7A
Greece 82,100 0.6
Hungary sve} 0.5
Iceland IND 0
reland SOTA) 0.2
taly 428,94] 3.1
Japan 1,173,360 85
Latvia 22,976 0.2
Liechtenstein 208 0
Luxembourg 11,343 0.1
Monaco 7\ 0
Netherlands 167,600 ane
ew Zealand 25,530 0.2
Norway pojooe 0.3
Poland 414,930 3
Portugal 42,148 0.3
Romania 171,103 Ie
Russian Federation 2,388,720 174
Slovakia 58,278 0.4
Spain 260,654 ie
Sweden 61,256 0.4
Switzerland 43,600 0.3
UK 584,078 43
USA 4,957,022 36.1
Total 3,728,306 100.0

Note: Data based on the information from the thirty-four Annex 1 parties that submitted their
first national communications on or before 11 December 1997, as compiled by the secretariat
in several documents (A/AC.237/81; FCCC/CP/1996/12/Add.2 and FCCC/SB/1997/6). Some of
the communications included data on CO, emissions by sources and removals by sinks from
land-use change and forestry, but since different ways of reporting were used these data are
not included.
Source: Report of the Conference ofthe parties on its third session, held at Kyoto from 1 to 11
December 1997 FCCC/CP/1997/7/Add.1.
THE ECONOMICS OF CLIMATE CHANGE 177

health from climate change can be thought of as ‘planet insurance’. Planet insurance cap-
tures the idea that societies need to spend money today to help reduce future risks and dam-
ages. This chapter examines how we can use economics to understand how invest in planet
insurance cost-effectively.

9.2 A Global Environmental Risk

Life on earth is possible because certain gases trap sunlight in our atmosphere and keep us
warm—as in a greenhouse. Carbon dioxide (CO,), mainly released from burning fossil
fuels, is one such greenhouse gas. But in excess, those gases may work against us, holding in
too much heat, blocking outward radiation, and altering our climate. Scientists warn that
such changes could affect agricultural yields, timber harvests, and water resource produc-
tivity. Results might include a rise in sea level, saltwater contamination of drinking water,
and more storms and floods. Human health could be threatened by more heat waves and
spreading tropical diseases. Defining the risk of such outcomes is crucial for good climate
policy. Reducing the risk of climate change suggests that we must wean ourselves off of the
fossil fuels we have used to build our stock of wealth (also see Krugman, 2010).
Good policy should also distinguish between stock and flow pollution. Stock pollution is
concentration—the accumulated carbon in the atmosphere, like water in a bathtub. Flow
pollution is emissions—the annual rate of emission, like water flowing into the tub. Because
risk comes from the total stock of carbon, our focus should be on projected concentration
levels. Greenhouse gases remain in the atmosphere decades before they dissipate, so differ-
ent rates of emission could generate the same concentrations by a given year; policy-makers
have options regarding how they hit a given concentration target.
One policy option to reduce emissions, as identified by economists, is to start slowly and
then increase the rate of emission reductions after several decades. That would allow for a
natural rate of capital depreciation and the replacement of high-carbon energy sources—
such as coal—with low-carbon sources such as wind and solar. The ‘broad, then deep’ path
is recommended by many researchers and policy-makers: broad participation by both
developed and developing countries, and a gradual emission reduction path to achieve a
long-term concentration target (Shogren, 1999). Olmstead and Stavins (2012) again stress
these points in their discussion of effective post-Kyoto international climate policy. They
argue that three elements are essential: a structure that incorporates both developed and
developing nations to participate, a focus on a gradual time path for emissions targets, and
the use of flexible incentive systems to reduce costs and promote cost-sharing.
Good policy should also account for alternative risk-reduction strategies—mitigation,
adaptation, and insurance. Mitigation involves investments to reduce the probability of
damages due to carbon emissions; mitigation is a public good determined by the sum of all
nations’ efforts to reduce carbon. Mitigation is a public risk-reduction strategy in which the
benefits of reduced risk accrue to all nations Adaptation involves investments to reduce the
severity of realized damages. Adaptation is a private good in which the benefits of reduced
severity accrue to one nation—and usually to certain sectors of this nation (World Bank,
2010). Insurance involves investments that transfer wealth from good to bad states of nature
given a bad event has been realized.
178 APPLYING THE TOOLS

Climate policy acknowledges that the climate is a global public good: everybody uses the
same one. It is the sum of all the carbon emitted around the globe that matters. This is crucial
because the biggest emitters of greenhouse gases will change over the next few decades.
Today, the industrialized world accounts for the largest portion of emissions; soon, transi-
tional economies in China and India will be the world’s largest emitters. That is why interna-
tional co-operation is essential.
ry

9.3 The Challenge of International Coordination

The world community began responding in 1979 with the First World Climate Conference,
held in Geneva and sponsored by the World Meteorological Organization. In 1988, the
IPCC held its first meeting to begin to address the state of scientific knowledge on climate
change. The IPCC consists of some 2,500 scientists, natural and social, who are charged with
evaluating potential impacts and finding cost-effective solutions. At the Rio Earth Summit
in 1992, 154 countries singed the United Nations Framework Convention on Climate
Change—the signatories agreed to stabilize greenhouse gas emissions voluntarily at 1990
levels by the year 2000.
By the mid-1990s, most observers knew that the Rio emission reduction target was
unlikely to be realized. In 1997, in Kyoto, Japan, 150 countries signed the Kyoto Protocol,
which binds the major industrialized nations (i.e. the 38 Annex 1 countries) to emissions
reductions below 1990 levels by 2008-12. Basic provisions were made for international
trading of emissions allowances and for sinks, but no agreement was reached about the
responsibilities of developing countries or financial incentives to them, and no concrete
measures for enforcement of the agreements were determined (on the Kyoto Protocol, see
Box 9.1). The Kyoto Protocol became international law in 2005 after Russia’s ratification
pushed emission reductions over the 55 per cent mark.
In 2007, the IPCC released its Fourth Assessment Report (FAR), stating that human-
driven climate change is ‘unequivocal’ (see Pachauri and Reisinger, 2007). The Copenhagen
Accord emerged in 2009. One of the Accord’s key achievements was to unite the United
States and China—the two largest emitters of greenhouse gases—in a commitment ‘to
reduce global emissions so as to hold the increase in global temperature below 2°C’. But the
Accord is not a legally binding agreement, and there was no global target for emission
reductions by 2050.
Today, achieving meaningful international co-operation remains an ‘insurmountable
opportunity’. Even though nations have a common interest in climate change, many are
reluctant to reduce carbon voluntarily. They realize that they cannot be prevented from
‘free-riding’: benefitting from reductions in emissions whether they contribute or not. Free-
riding is complicated further in developing countries, where clean water and a stable food
supply seem more urgent than climate change policy. The developing world also has less
financial and technical capacities to act, and different perceptions about what constitutes
equitable distributions ofeffort.
Economists use game theory as a tool to examine the climate coordination problem, as we
saw in Chapter 7. The typical approach to model strategic interaction in climate policy is to
think of it as a public good (or prisoner’s dilemma) game. Since a credible ‘climate police
CS OF CLIMATE CHANGE 179

Box The Kyoto Protocol on Climate hange ”

The Kyoto Protocol represents a major international environmental agreement—one of the most
significant attempts to coordinate voluntary and binding global action on climate change policy. About
150 countries met in Kyoto, Japan in December 1997 at the Third Conference of the Parties (COP-3) to
the United Nations Framework Convention on Climate Change (UNFCC) (see United Nations, 1992).
Their task was to create a legally binding international agreement for climate protection—the Kyoto
Protocol. The Kyoto Protocol (United Nations, 1997) was the culmination of years of negotiations to
strengthen the first international climate change treaty signed by over 160 countries at the 1992 Earth
Summit in Rio de Janeiro. The original treaty, the UNFCC, called on industrial nations to voluntarily
reduce their greenhouse gas emissions to 1990 levels by 2000.
What does the Kyoto Protocol say?
Targets and timetables (Article 3). The protocol set a legally binding target for thirty-nine of the world’s
most developed countries to reduce greenhouse gas emissions in aggregate by 5.2 per cent from a 1990
baseline for the period 2008-12. The targets are differentiated by nation, ranging from an 8 per cent
reduction (the EU) to a 10 per cent increase (Iceland) from 1990 levels. The USA agreed to a target of
7 percent reduction; Japan a 6 per cent reduction (see Table 9.2). The goal was for each party to show
demonstrable progress towards meeting its target by 2005.
Nations can act jointly to hit their target (Article 4). The Protocol lets a group of nations form a multi-
country ‘bubble’, in which the group has an overall target to reach. Each nation inside the bubble has its
own commitment to the rest of the group. The bubble met the EU’s demand that it should be able to
comply as a group. The bubble does require the EU to adjust its commitment if its membership enlarges.
Greenhouse gases (Article 3: Annex 1). The Protocol covers six greenhouse gases—carbon dioxide,
methane, nitrous oxide, hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulphur hexafluoride
(SF6)—as a ‘basket’. The latter three use a 1995 baseline instead of 1990. The inclusion of the six gases
allows for some flexibility in reaching the target. Reductions in one gas can be used to substitute for
reductions in other gases.
Emission trading (Article 16). The Protocol allows for emission trading among the nations to fulfil their
commitments. An emission-trading programme provides greater flexibility for a nation to achieve its target.
The joint implementation/clean development mechanism (Articles 6 and 12). Joint implementation (JI) is
when one nation gets credit for implementing a project to reduce carbon emissions in another country.
A new device, the clean development mechanism (CDM), was developed for joint projects with
developing nations through the payment of a special administrative fee by developed nations.
Carbon sinks. The protocol allows for carbon sinks—land and forestry practices that remove carbon
emissions from the atmosphere. Sinks could play an important role for some nations because they represent
a low-cost option. Sinks are ambiguously defined in the Protocol, and will be a challenge to measure.
No harmonization ofactions. The Protocol allows each nation to figure out its own best strategy to
meet its commitment. Not everyone sees this as a good thing: some critics have argued that the world
would have been better served by a common action rather than a common target.
What didn’t the Kyoto Protocol achieve?
Developing country participation. No agreement was reached in Kyoto on what commitments they
should assume to reduce their greenhouse gas emissions. But everyone agrees that climate protection
requires the participation of the developing countries, because by the middle ofthe next century, they
are predicted to generate the largest share of carbon emissions. Developing nations have no incentive to
reduce their economic growth. In 2011, China was the largest emitter just ahead of the USA, but its per
capita emissions are about a seventh of those in the USA. A Chinese delegate captured the sentiment
underlying the opposition: [W]hat they [developed nations] are doing is luxury emissions, what we are
doing is survival emissions.’ Compensation might be required to induce necessary participation of these
emerging economies.
180 APPLYING THE TOOLS

Table 9.2 Kyoto Protocol Annex 2. Quantified emission limitation

or reduction commitment (percentage of base year or period)

Australia 108
Austria ; 92
Belgium SV
Bulgaria’ 92
Canada 94
Croatia? 5
Czech Republic? 92
Denmark 92
Estonia? a2
European Community 92
Finland 92
France 92
Germany 92
Greece 2
Hungary# 94
celand 110
reland 92
taly 92
Japan 94
Latvia? 92
Liechtenstein 92
Lithuania? 92
Luxembourg 92
onaco 92
etherlands 92
New Zealand 100
Norway 10]
Poland 94
Portugal 92
Romania? wi
Russian Federation? 100
Slovakia’ 92
Slovenia® 92
Spain 92
Sweden 32
Switzerland 92
Ukraine? 100
United Kingdom 92
United States 93
er
a ee ae a aN ee Sd
* Countries in transition to a market economy.
THE ECONOMICS OF CLIMATE CHANGE 181

force’ does not exist to enforce an international environmental agreement, this implies that
an agreement must be voluntary and self-enforcing. But recall that in a public good game
each player has an incentive to deviate from the co-operative solution—each country wants
to ‘free-ride’ on the abatement actions of others (or each prisoner gives up the other pris-
oner). These free-riders capture all the benefits of climate protection and pay none of the
costs. In this case, the Nash equilibrium—the outcome in which no one has a unilateral
incentive to deviate—is for all countries to free-ride.
In reality, what we have witnessed are partial international environmental agreements—
a coordination game with multiple equilibria (ranging from good to bad) rather than a pub-
lic good game with a single worst-case outcome (see DeCanio and Fremstad, 2011). If
coordination games better capture climate policy, the open question is whether it is better
to have a partial co-operation, in which some nations do a lot and others do nothing
(co-operation-limited), or to have full co-operation, in which all nations do a little but not
the optimal amount (co-operation-lite). Which scenario leads to the most total abatement?
In our example in Figure 9.1, it is co-operation-lite that has the greatest total abatement, but
the opposite could also hold. This issue deserves more attention.
Game theory suggests that sub-optimal coordination can be alleviated if conforming
nations can retaliate against violators with trade sanctions. But the force of this deter-
rence is blunted in several respects. First, a nation’s incentive to deviate from the agree-
ment depends on how it views a short-term gain from cheating compared to the long-term
losses from punishment. Nations that must deal today with other problems, such as the
supply of clean water, might discount the threat of sanctions. Second, conforming nations
must see a gain in applying punishment; otherwise, their threats are not to be believed.
And since many forms of sanctions exist, nations need to select a mutually agreeable
approach—not a trivial negotiation. Alternatively, one could argue that the process of
negotiation itself can help reinforce mutual expectations for co-operation. This perspec-
tive recognizes that unilateral confidence-building moves by some countries could inspire
like-minded actions by others, and communication itself can reinforce positive
expectations.

Full co-operation

~di
\
Co-operation-lite

Co-operation-limited 7

Free-riding
the
of
abatement
total
optimal
Per
cent
Abatement level

Figure 9.1 Voluntary contributions to abatement.

182 APPLYING THE TOOLS

Good international climate policy should also address the implementation of cost-
effective risk-reduction strategies, such as carbon taxes or carbon emission trading.
Carbon taxes fix the cost of carbon, and allow the quantity of emissions to be determined
by the private sector. Emission trading fixes the quantity of emissions and allows people
to trade emission permits at a price set by the market. Such flexible mechanisms may
increase the likelihood of international co-operation.
First, consider the carbon tax. A carbon tax add8 a fee to the price of fossil fuels according
to their relative carbon content. The tax could be collected in various ways: as a severance
tax on domestic fossil fuel output plus an equal tax on imports; as a tax on primary energy
inputs levied on refineries, gas transportation systems, and coal shippers; or further down-
stream to homeowners and car or truck owners. The further upstream the tax is levied, the
less carbon ‘leaks’ out through uncovered activities such as oilfield processing. A fossil fuel
tax also would not be that difficult to administer, given the existing tax collection apparatus
in the USA and the European Union. Several countries now have a carbon tax: Denmark,
Sweden, Norway, and the United Kingdom. The UK carbon tax was imposed on electricity
generation in the 2011 budget, and is set initially at a level of £16 per tonne of CO,, rising to
£30 per tonne in 2020. This tax is called the ‘carbon floor price’, since it operates in tandem
with the EU carbon trading system. Policy-makers use this carbon tax to change the relative
prices between renewable and non-renewable energy resources; electricity generation via
wind and solar now become cheaper relative to non-fossil fuel sources.
A carbon tax provides incentive to emitters to reduce emissions when the marginal abate-
ment cost was less than or equal to the tax, as we saw in Chapter 2. A tax can stimulate several
responses. Firms might reduce their tax exposure by reducing CO, emissions. Fossil fuel
users would have an incentive to improve energy efficiency, use less carbon-intensive fuels,
and consume less of the goods and services produced in carbon-intensive ways. Taxes would
trigger the diffusion and development of new technologies that emit less carbon. The tax is
cost-effective because it sets up the institutional structure so that the marginal abatement
costs are the same across all sources—there are no gains from trade remaining to be had. One
downside is that a carbon tax does not guarantee a specific emissions reduction goal.
A carbon tax could be extended to other greenhouse gases. The appropriate tax on natural
gas entering the pipeline system could account for leakage, and the greater relative potency
of methane. Levies also could be placed on methane releases from coal mines and landfills,
and on hydrochlorofluorocarbons, on the basis of their expected venting to the atmosphere
through sources such as automobile air-conditioners. Extending taxes to agricultural
sources of methane is conceivable, but given the decentralized and difficult-to-measure
nature of these sources, such an extension might be problematic in practice.
Second, we consider a tradable permit system; also called cap-and-trade. The Kyoto
Protocol establishes quantitative targets for reductions in the emissions of carbon dioxide
by signatory countries. Many countries have promoted the use of tradable permit markets
to achieve these targets. Why? If economics has one key insight, it is that value is created by
trade. High-cost firms can buy permits to emit carbon from low-cost firms—both gain from
the exchange. Tradable permits are attractive since they create mutual gains between trad-
ing partners, guarantee a fixed level of emission reductions, and produce less political flak
than green taxes. As Box 9.2 explains, an extensive carbon trading system was introduced by
the EU in 2005.
AICS OF CLIMATE CHANGE 183

The EU Emissions Trading Scheme was launched in January 2005, covers about 45 per cent of UK
emissions, and is mandatory for power generators, metal and minerals industries, and pulp and paper
plants (see Ellerman and Buchner, 2007). The system covers 12,000 CO, producers across the whole EU.
They use the National Allocation Plans (NAP) to determine baseline allocation, with grandfathering under
Phase 1 (2005-7) for 95 per cent of permits, and 5 per cent allocated by auction. These NAP are
supposed to be consistent with individual-country Kyoto targets. We are currently in Phase 2, from 2008
to 2012, which will then be followed by Phase 3 (2013-20). The EU system is the largest permit trading
scheme ever, and one that includes the option to trade across borders, and it is linked to ‘flexible
mechanisms’ under Kyoto. German sources (ZEW) predicted that EU-wide costs of hitting Kyoto targets
would be higher by €79 billion without permit trading:

— Phase1 (2005-7) — Phase3 (2008-12)

€/tonne
CO,

Source: Ellerman and Buchner (2007).

A big difference in Phase 2 is a reduction of between 5 per cent and 10 per cent in the emissions
permits granted. City analysts believed that this would lead to a big increase in the market price of
carbon. Deutsche Bank expected forward prices to rise from the 2007 level to €35. These predictions
were made before the recent global recession, however, which has seen prices fall (the current price is
€13 per tonne). The system also allows for banking and borrowing of permits between Phases 2 and 3.
More auctions are planned in Phase 3, and Phase 3 will see the scheme extended to commercial airlines
flying in European airspace.

Some economists, however, have questioned whether cap-and-trade is the best approach
given the uncertainty attached to the costs of reducing CO). They have promoted the use of
a carbon tax. A nation sets a tax (or price) on emissions of CO,, or on actions that produce
the fossil fuels that generate the CO, emissions. Producers respond to the tax by seeking out
the lowest means ofproduction, internalizing the costs of CO, emissions. Several European
nations have introduced carbon taxes during the 1990s and 2000s, including Denmark,
Finland, France, the Netherlands, Norway, Sweden, Switzerland, and the UK. In the sum-
mer of 2012, the Australian government will introduce a new carbon tax aimed at reducing
carbon emissions, which implies that firms and people will probably have to reduce their
use of fossil fuels.
184 APPLYING THE TOOLS

If no uncertainty existed over future control costs, tradable quantity controls and carbon
taxes would produce similar outcomes. But there is considerable uncertainty over how big
carbon control costs will be in the future due to three factors: (i) we have little experience
with such large cuts in emissions; (ii) we do not know what future technological options will
be; and (iii) we do not know what the ‘do nothing’ level of emission will be, relative to which
achievements are measured and targets set.
Pizer (2002) constructed a scenario in which uncertainty exists over control costs, and
that these costs turn out to be greater than thought. Under a permit system, emissions stay
constant, but costs of control rise. Under a tax system, emissions are cut by less, even though
the price per tonne stays constant. Permit systems result in more uncertainty about costs
than emission reductions, whilst this situation is reversed with taxes. Under certainty, emis-
sions by 2010 for either a $80 per tonne carbon tax or a permit market with 8.5 gigatonnes
or total allowances are about equal.
Pizer tested the strength of these alternatives by simulating 1,000 different predictions
based around IPCC calculations. He finds emissions are below the 8.5 gigatonnes of carbon
level in 75 per cent of cases with the tax, but exceed it in the remaining cases. The permit mar-
ket ensures that emissions never go above 8.5 gigatonnes. The same simulation of possible
future scenarios shows the costs of the permit scheme to be in the range of 0.0-2.2 per cent of
global gross domestic product (GDP), a much larger range than that for the tax, at 0.2-0.6 per
cent. The variation of control costs is much greater under the permit system than with a tax.
Which policy option should we choose? Pizer argues that it depends on what we believe
about the damage. If there is some threshold beyond which further CO, emissions will
impose high (and maybe irreversible) costs, the greater certainty over emission levels that
comes with permit markets is preferable. If, instead, damage rises smoothly with increasing
emissions, the threats are not so bad, and we prefer the greater certainty over control costs
that comes from taxes. This preference for taxes over permits is reinforced when one
remembers that it is not current emissions that most worry us about climate change, but the
overall stock of greenhouse gases in the atmosphere, which changes slowly.
Nordhaus (2008) makes the case for an internationally harmonized carbon tax over a per-
mit system. His argument is that a carbon tax is likely to be a more efficient incentive device
because of its conceptual simplicity. A harmonized carbon tax is a method to help coordinate
policies across countries. The tax will hold the advantage over permits if the goals are to
maintain the flexibility needed to promote economic growth, to minimize inefficiencies due
to highly non-linear damages, to avoid the volatility that can arise in permit prices, and to
encourage the idea of a ‘double dividend’—to raise revenues through the carbon tax while
reducing taxes on other goods or inputs. The potential downside of a harmonized carbon tax
is that implicit carbon prices will differ across countries due to different opportunity costs.
This suggests that richer nations will have to make large transfer payments to poorer coun-
tries to offset any differences in carbon prices. Landis and Bernauer (2012) estimate that for
a global carbon price of $35 per tonne of CO,, the transfer payments could range from $15
billion to $48 billion per year. As a comparative benchmark, the Development Assistance
Committee of theOECD provided $134 billion in official development assistance in 2011.
But carbon taxes are politically unpopular, and unlikely to be widely used. What to do?
Economists have argued for a hybrid system—one that combines taxes and permits. In this
system, the government first allocates a number of limited life permits freely, and then
THE ECONOMICS OF CLIMATE CHANGE 185

makes additional supplies available at a price—a ‘trigger price’. This trigger price works like
a tax on emissions or a safety valve, and can be increased over time if we want to tighten up
progressively on climate policy.

9.4 The Benefits and Costs of International Co-operation

The conventional economics approach to assessing climate change policy is to calculate the
benefits and costs of action or inaction. As noted by Pindyck (2011), most economic analyses
of climate change policy compare costs and benefits in five steps: (a) defining a ‘business as
usual’ (BAU) carbon emissions benchmark against which we can compare any policy
options; (b) the potential temperature changes that might arise from staying on the BAU
benchmark path, either at a global average or at a regional level; (c) the estimated losses to the
world economy (GDP) due to the higher temperatures from the BAU path—some estimates
include human adaptation, some do not; (d) the estimated costs associated with reducing
BAU emissions to some reduced target level of emissions; and (e) presumptions about how
society values these changes today and off into the future.
The benefits from climate policy are determined by what losses we avoid with lower con-
centrations of carbon: severe weather patterns, hobbled ecosystems, less biodiversity, less
potable water, loss of coastal areas, rises in mean temperature, more infectious diseases such
as malaria and cholera. Climate change could benefit some parts of agriculture and forestry
with longer growing seasons and more fertilization. These benefits from a climate agree-
ment are most likely to accrue to the future generations in developing nations because their
economies depend relatively more on climate (e.g. agriculture, forestry, and fishing). As
stressed by Schelling (1997), climate policy boils down to a wealth transfer from today’s
industrial nations to future generations in emerging economies.
These gains (or losses) can be categorized into four broad sets that are increasingly diffi-
cult to quantify: the avoided losses to market goods and services, non-market goods, ancil-
lary effects, and catastrophes. People judge the benefits of climate protection as the
incremental reduction in human and environmental risks compared to the ‘business as
usual’ baseline. Under BAU, modellers have estimated that carbon concentrations might be
expected to double pre-industrial levels within the next half century, with mean tempera-
tures predicted to rise by about 1°C by 2050, and 2.5°C by 2100. With Kyoto, concentrations
are still likely to double, with temperatures increasing by about 0.1°C by 2050 and 0.5°C by
2100.
By most estimates produced in the 1990s and early 2000s, climate change could reduce
gross world product by around 1 or 2 per cent. The impact on GDP in the USA has been
estimated to be between plus or minus 1 per cent. Most industries in the developed nations
are separate from climate: less than 3 per cent of US livelihoods, for instance, are earned in
agriculture and other climate-sensitive activities. If one included the potential non-market
damage, the market and non-market benefits to the USA might be at most about 2 per cent
of GDP.
Tol (2012) recently reviewed the estimated range of total economic impacts due to global
warming (estimated between 1994 and 2006). He finds that most estimates suggest that a
doubling of atmospheric concentrations would lead to moderate economic decline, a
186 APPLYING THE TOOLS

decrease of 1 or 2 per cent of GDP. The overall pattern that emerges from these economic
estimates is that of moderate economic gains with a modest increase in temperature, but
then significant economic losses with more substantial temperature increases, over 3°C.
These estimates include the 2006 Stern Review on the Economics of Climate Change,
which changed the political debate on the range of damages to the global economy. Without
action to reduce global emissions, the Stern Review estimated that world GDP could fall by
5 per cent every year for the long run; potentially, under the worst-case scenario, GDP loss
might be as high as 20 per cent per year. The Stern Review raised the political costs of doing
nothing. The main conclusion was that the benefits (i.e. avoided damages) of extreme and
immediate climate protection outweigh the costs. The estimated damages could be reduced
if the global community took early and strong action to reduce emissions, actions that
would cost about 1 per cent of world GDP. Krugman calls this action the ‘climate policy big
bang’ approach.
Critics of the Stern Review challenge both the estimated costs and benefits of climate
action. They argue that the estimated future damages are too high, given the report’s use of
unjustifiable low social discount rates. Economists use a discount rate to compare economic
impacts that arise at different times. A social discount rate captures how today’s generation
values the benefits and costs that accrue to future generations—an ethical statement about
how we value present and near-present costs relative to the distant future. The Stern Review
argued that a low social discount rate (e.g. not much more than zero) was ethical and there-
fore appropriate. Critics such as Nordhaus argued that such a low rate was inconsistent with
the implied discounting rates that are consistent with today’s marketplace.
Critics also argued that the costs of climate protection are too low—more reasonable
estimates would raise the costs of mitigation. The costs of mitigation are lower if one makes
favourable assumptions about rapid technological change and institutional responsive to
implement energy efficiency policy. If people and institutions are assumed to readily adopt
new low-carbon technologies without a price shock, the costs of mitigation are much lower.
Many new low-carbon technologies exist already; the open question is how quickly these
options will penetrate into the economy. Economists believe that a price shock in energy is
a necessary requirement to speed up adoption of new technologies.
In addition, there are two topics that arise in the non-market arena that are likely to
trigger debates over costs and benefits of climate change: human health and ecosystem/
endangered species services. Potential threats to human health include thirty diseases
and infections new to medicine, such as E. coli, hantavirus, and HIV, plus old scourges
such as cholera, plague, yellow and dengue fever, tuberculosis, and malaria. How do we
quantify such threats? It is another challenge to estimate the social value of ecosystem
services and endangered species. Despite the extraordinary analytical difficulties associ-
ated with measuring the social value of preserving each species, determining at least a
range for these values is essential if we are to make judgements about the benefits of
preservation.
One way to increase the size of the benefits of the Kyoto Protocol is to add in the potential
ancillary benefits that might come from discouraging fossil fuel consumption. The Kyoto
Protocol would reduce emissions of such air pollutants as carbon monoxide, sulphur and
nitrogen oxides, and toxic trace pollutants in exhaust gases, thereby reducing their damage
to health, visibility, materials, and crops. Studies in Europe and the USA have estimated that
187

Box 9.3 Stated Preference Evidence fromthe US

In a recent stated preference study of 1,651 US households, Lee and Cameron (2008) estimated the
willingness to support/pay for ‘large’ mitigation policies that maintain current climate. They investigate
whether this WTP depends on: (i) how seriously people take the ‘do nothing’ scenario; (ii) how much
mitigation costs; and (iii) who pays these costs (how they are distributed), both domestically and
internationally. They implement the question of who pays (domestic) by looking at how costs are divided
across energy taxes, income taxes, falling investment returns, and higher consumer prices; they
implement who pays (international) by considering global cost shares between (i) India and China,
(ii) ‘other developing countries’, (iii) the USA and Japan, and (iv) other industrialized countries.
Their results confirm ex ante expectations that the WTP depends on perceived severity of impacts of
climate change; on how much of the costs are borne by increased energy taxes; on how much the USA
and Japan pay as their share of global costs; and on which sectors experience the biggest effects (e.g.
agriculture, health, ecosystems). For example, the average WTP per household per month in higher
energy taxes to reduce the risk of ‘moderate harm’ or ‘substantial harm’ is as follows: (a) $6 or $456 if the
USA/Japan pay 100 per cent; and (b) $271 or $728 if the USA/Japan pay 31 per cent and India/China pay
17 per cent. Overall, across the entire sample, the mean and median WTP if the USA/Japan pay all costs
are $151 and $62 per month, respectively.

the non-climate benefits might be as large—or larger—than the direct benefits from avoid-
ing climate change. Box 9.3 shows some results from a study of the willingness to pay (WTP)
of US citizens to mitigate carbon emissions, and shows that these depend on which other
countries join in the effort, and on which sectors are likely to experience the biggest impacts.
Finally, although modellers typically presume that climate change will be gradual—a
slow and steady rise in temperature or precipitation—some have raised the spectre of catas-
trophe. They suggest that the risk of a sudden rupture is real; for example, a structural
change in ocean currents or the melting of the Western Antarctic ice sheet. The problem is
that researchers do not have any reasonable estimates of the odds that these events will
come to pass, but making informed policy judgements requires knowing those odds.
There has been a range of estimated costs of climate protection. Some studies suggest that
the world could reduce emissions at negligible cost; others call climate change policies an
‘economic disarmament’. In the early 2000s, a report from the US White House stated that
the costs to the USA ofmeeting its Kyoto target were ‘likely to be modest’ if reductions were
efficiently pursued with domestic and international emissions trading, joint implementa-
tion, and the Clean Development Mechanism (a system in which developed nations can buy
the carbon reductions in developing nations). By ‘modest’, the report means an annual GDP
drop ofless than 0.5 per cent (roughly $10 billion); no expected negative effect on the trade
deficit; increased gasoline or petrol prices of about 5 cents a gallon in the USA; lower elec-
tricity rates; and no major impacts on the employment rate.
But other estimates suggest that the US GDP could take an annual hit of nearly 3 per cent,
or about $250 billion a year, with intra-nation emission trading. Also, the trade deficit
would increase by billions of dollars; gasoline prices would increase by 50 cents a gallon;
electricity prices would nearly double; and two million US jobs would disappear. The net
global costs have been estimated at over $700 billion, with the USA bearing about two-
thirds of those costs.
188 APPLYING THE TOOLS

The effects of international climate policy on world trade patterns are not well under-
stood. Many leaders envision a ‘pollution haven’ hypothesis, as outlined in Chapter 8: a
domestic industry relocates to a developing-country ‘haven’ that has weaker regulations on
emissions. This scenario seems unlikely on economic grounds. Except for the biggest pollut-
ing industries, the costs of complying with environmental regulations are a small fraction of
total costs, and are outweighed by international differences in labour costs, capital costs,
material costs, and exchange rate changes. The differences between developed nations’ envi-
ronmental regulations and those of most major trading partners are not that big. Besides,
developed-nation firms build state-of-the art facilities abroad regardless of the host nation’s
environmental regulations.
Leaders also fear that carbon policy will affect the demand for domestic energy-intensive
goods and cause the trade balance to deteriorate. But studies have not supported this idea.
A related notion is the ‘leakage effect’—cuts in domestic emissions are offset by shifts in
production and increases in emissions abroad. The early research range of estimates on
carbon leakage: unilateral emissions-reduction policies in OECD countries predict leakage
rates of anywhere between 3.5 and 70 per cent. A recent study by Aichele and Felbermayr
(2012) has arrived at a more definitive and pessimistic story. They examine the carbon foot-
print for countries that committed to the Kyoto Protocol. A carbon footprint is defined as a
measure that accounts for all carbon emissions generated by a nation’s citizens, regardless
of where the good was produced. Calculating the carbon footprint between 1995 and 2007,
they estimate that the Kyoto Protocol generated significant carbon leakage. Countries com-
mitted to Kyoto have reduced domestic emissions, but given the relocation of production,
they have not reduced their carbon footprint. This result supports the ‘broad, then deep’
idea for international climate agreements. Again, this suggests broad participation by devel-
oped and emerging economies, with a gradual emission reduction path over a long period—
what Nordhaus calls the ‘climate-policy ramp’.
Finally, cost estimates are likely to be on the low side for several reasons. Models presume
the most efficient possible climate control programme. The models assume that the control
programme is announced early and maintained indefinitely, even though governments will
be hard pressed to maintain consistent control over the decades. Many models focus on
long-term equilibrium and do not address the short-run adjustments, such as the oil shocks
ofthe 1970s, which could raise cost estimates by a factor of between one and four. This fac-
tor would cause Kyoto to reduce GDP by 1-10 per cent from baseline. Compare that to the
2 per cent of GDP that the USA now spends on all environmental programmes combined.

9.5 Economic Issues Underlying Benefit and Cost Estimates

How one accounts for the benefits and costs of international co-operation depends on what
one believes about the nature of three elements that underlie climate protection: the chance
of a catastrophe, the degree of flexibility in finding low-cost solutions for mitigation and
adaptation, and the origins of technological advance. We now take a brief look at each of
these issues,
If you believe that catastrophe is imminent, or at least more likely than previously
believed, emission reductions need to start now. If you do not, it is harder to justify the likely
THE ECONOMICS OF CLIMATE CHANGE 189

costs of rapid climate protection without global trading. Reliable information is needed to
help people understand the nature ofclimate change. Science does not know with certainty
which regions will get warmer or cooler, which will get wetter or drier, which will get storm-
ier or calmer. Climate policy discussions also consider whether a model has accounted for
the likelihood that a change in the ecosystem will be discontinuous—a catastrophe. Most
modellers acknowledge that their models do not address the potential of structural change
such as discontinuous shocks, like a sudden shift in the Gulf Stream or an unravelling of
natural systems from biodiversity losses.
The risk of climate change catastrophe has triggered a debate about how to evaluate the
benefits and costs of climate change under uncertainty over climate variability. This is the
so-called ‘fat tails’ argument. A fat tail reflects the idea that the probability of an extreme-
impact event is not as rare as believed under the ‘business as usual’ climate scenarios. A ‘fat
tail’ is the term used to capture the idea that societies cannot rule out the possibility of an
extreme event due to climate change. So-called “deep structural uncertainty exists about the
nature of climate system and its variability. This structural uncertainty means that the
chance of extreme-event damages are not zero—there is a positive chance of an absolute
catastrophe due to climate change. In this case, standard cost-benefit analysis is unhelpful,
because the benefits of avoiding infinite damages will always outweigh the costs of climate
protection (see Weitzman, 2011).
In addition, extreme events and structural uncertainty about climate change induce a
behavioural response from regular people—we all tend to overestimate low-probability
events. Experience tells people little about low-probability risks such as climate change.
They rely on outside sources of information to help them make judgements about the likeli-
hood that a bad event will come to pass. If that information stresses severity without giving
some notion of the odds, people systematically bias their risk perceptions upwards.
Numerous studies have revealed that people overestimate the chance that they will suffer
from a low-probability/high-severity event; for example, a nuclear power accident.
Second, the costs to meet a policy depend on how quickly society wants to change its
energy systems and capital structure. A stringent, inflexible carbon policy will induce
greater economic burden than a loose, flexible policy, since more flexibility allows firms
greater agility to search out the lowest-cost alternatives. Estimates suggest that any agree-
ment without the flexibility provided by trading will at least double the costs.
Flexibility means the ability to reduce carbon at the lowest cost, and three issues are rel-
evant. We first need to work through how the trading system would be designed before
alternative policies can be evaluated. The business of defining the rules for flexibility incen-
tive systems is wide open, and experimental economists could play an important role in
reducing the uncertainty. This holds for joint implementation and the clean development
mechanism as well.
We need to address the role of ‘carbon sinks’, which remain the wild card in the search for
flexible, low-cost solutions. A sink is a process that destroys or absorbs greenhouse gases,
such as the absorption of atmospheric carbon dioxide by trees, soils, and other types of
vegetation. In the USA, forests are an important terrestrial sink, since they cover about
750 million acres. A few studies have found that carbon sequestration through sinks could
cost as little as $25 per tonne in the USA, although this figure is likely to vary substantially
across types of forest, across land types, and according to how forests are managed (Read
190 APPLYING THE TOOLS

et al., 2009). The United Nations REDD programme (Reducing Emissions from
Deforestation and Forest Degradation) is another ongoing effort to create financial value
for carbon storage, especially in forests in developing countries. Started in 2008, the REDD
programme is designed to provide technical assistance to implement strategies to reduce
deforestation (also see Chapter 7). But serious uncertainties remain about how to measure
and account for estimates of net carbon sequestered in forests, whilst enhancing forest car-
bon sequestration might have mixed effects on other public goods associated with forestry,
such as biodiversity (Caparros et al., 2010).
We also need to consider that the existing tax system might accentuate the costs of cli-
mate protection. Labour and capital taxes distort behaviour because they reduce employ-
ment and investment levels to below what they would have been otherwise. If we add on a
carbon tax that discourages consumption and production, we further reduce employment
and investment, which then exacerbates the labour and capital tax distortions, perhaps by
as much as 400 per cent. One could reduce these extra costs by channelling the revenue from
the carbon tax, if any existed, to reduce the labour and capital taxes.
Finally, the costs of climate protection depend on what one chooses to believe about the
origins of technological diffusion. Some people argue that those origins are rooted in non-
price responses: people will do the right thing for the right reason. Economists disagree with
such optimistic scenarios—the majority of people do not adopt technologies that cost more
just to protect the environment. People have other more immediate needs. Economists see
the origins of technological advance as driven by changes in relative prices. Even if new
technologies are available, people do not switch unless prices induce them to switch. People
behave as if their time horizons were short, perhaps reflecting their uncertainty about future
energy prices and the reliability of the technology. High initial investment costs also slow
down the adoption of new technologies—for example, replacing all household light bulbs at
once with energy-efficient lights—whilst transaction costs also set barriers on the take-up of
low carbon technologies by households.
If people adopted new technologies—such as compact fluorescent light bulbs, improved
thermal insulation, heating and cooling systems, and energy-efficient appliances—without
a price shock, carbon emissions could be eliminated at lower cost (see Gillingham et al.,
2009). The open question is whether people can be ‘nudged’ into adopting climate-friendly
technologies without a sustained price increase in energy. While some people adopt new
technologies on their own, economists estimate that higher energy prices are typically asso-
ciated with significantly more adoption of energy-efficient equipment. For example,
Chakravorty et al. (1997) argue that if historical rates of cost reduction in the production of
solar energy are maintained (30-50% per decade), more than 90 per cent of the world’s
remaining coal will never be used. In their simulations, the world economy makes the tran-
sition to solar from coal and oil even without a carbon tax. Global temperatures will increase
by 1.5-2.0°C by around 2050, and will then decline to pre-industrial levels.

9.6 A Flexibility-Stringency Trade-off

In the end, the economics ofclimate change rests on the trade-off between the stringency of
a country’s emissions reduction path and how much flexibility this country has to reach its
THE ECONOMICS OF CLIMATE CHANGE 191

:No emission trading

stringent
Less
———>

of
climate
Stringency
agreement
a

Full world trading

More flexibility ————
Flexibility of aclimate agreement

Figure 9.2 The flexibility-stringency trade-off.

target. A stringent, inflexible carbon policy creates a greater economic burden than a loose,
flexible policy. If the goal is to keep costs down to an acceptable level, the push for a more
stringent target requires flexibility in how a nation can achieve the target. If market or insti-
tutional imperfections restrict flexibility, policy-makers can loosen the stringency of the
target to keep the costs down. Figure 9.2 illustrates this stringency—flexibility trade-off; the
iso-cost curves represent different combinations of flexibility and stringency that generate
the same given cost to the economy.
Now consider the policy options from a post-Kyoto agreement, using points A and B in
the figure. Both points are on an iso-cost curve representing a relatively low cost to the
economy. Point A could represent China going into post-Kyoto negotiations. Point B
represents the same total cost to the economy, except that now emissions trading is pro-
hibited. This point is consistent with the negative reaction of many developed and devel-
oping countries towards emissions trading. To maintain the same cost as at point A, point
B must involve a weaker target. If you lose flexibility, you have to give up some stringency
to keep costs fixed. What emerged instead from the original Kyoto process was something
like point C, with both less policy flexibility than some nations initially sought and a
somewhat more ambitious emissions control target. Point C is on an iso-cost curve, with
higher cost to the economy than points A and B.

Summary
Climate change is a historical fact, and its connection to human actions is better understood
now. Today, economists do not debate over the need for climate policy; rather, we make
cases for different levels of stringency in abatement of carbon emissions and when these
policies should begin to go from loose to strict targets. Economics offers insight into how
192 APPLYING THE TOOLS

people value these alternative reductions in climate risk—how much they are willing to
sacrifice in today’s consumption for less climate risk in the future. An understanding of this
trade-off is necessary to help design more cost-effective climate change policy. Cost-effective
policy requires that we better understand how countries can coordinate their actions, and
how private citizens adapt to changing climate, because adaptation affects the effectiveness
of international mitigation efforts.
In the end, estimating the costs and benefits of climate policy depends on what you
choose to believe about four points: whether people are increasing the real risk of a climate
catastrophe; how much flexibility people will have to find low-cost solutions (e.g. cap-and-
trade); whether people can coordinate actions for global participation to avoid a shell game
of moving carbon emissions around the globe; and how responsive people are to adopting
new technologies, with and without changes in the relative price of high- and low-carbon
energy. If you are a pessimist on climate risk but an optimist that people will react rationally,
the economics of climate change suggest that people should take strong action right now—
the climate policy big bang. If you are catastrophic risk optimist but a realist about human
actions, economics does not say ‘do nothing’—it suggests a slower policy response, in which
we follow a climate policy ramp-up over a long period to reduce fossil fuel use.

Tutorial Questions
9.1 Describe the major economic benefits of reducing climate change risk.
9.2 Describe the key drivers that will either increase or decrease the costs of climate
policy around the globe.
9.3 Suppose that you are asked to put a price on carbon emissions. Explain why you
prefer a global carbon tax or a global cap-and-trade system.
9.4 How does the idea of a ‘fat tail’ extreme event affect the economics of climate change?
9.5 Explain the challenges that countries face when considering how to coordinate over
an international environmental agreement such as the Kyoto Protocol.

References and Additional Readings

Aichele, R., and Felbermayr, G. (2012). ‘Kyoto and the DeCanio, S., and Fremstad, A. (2011). ‘Game theory
carbon footprint of nations, Journal ofEnviron- and climate diplomacy, Ecological Economics,
mental Economics and Management 63: 336-54. [Link]
Blinder, A. (1997). ‘Needed: planet insurance, New $0921800911001698.
York Times, 22 October. Ellerman, D., and Buchner, B. (2007). ‘The
Caparros, A., Cerda, E., Ovando, P., and Campos, P. European Union Emissions Trading
(2010). ‘Carbon sequestration with reforestation Scheme: origins, allocation, and early
and biodiversity, Environmental and Resource results, Review of Environmental Economics
Economics 45: 49-72. and Policy 1: 66-87.
Chakravorty, U., Roumasset, J., and Tse, K. (1997).
Gillingham, K., Newell, R., and Palmer, K. (2009).
‘Endogenous substitution among energy ‘Energy efficiency economics and policy, NBER
resources and global warming, Journal ofPolitical Working Paper No. 15031.
Economy 105: 1201-34.
THE ECONOMICS OF CLIMATE CHANGE 193

Heal, G. (2009). ‘Climate economics: a meta-review Read, D.J., Freer-Smith, P.H., Morison, J.I.L., Hanley,
and some suggestions, Review of Environmental N., West, C.C., and Snowdon, P. (eds.) (2009).
Economics and Policy 3: 4-21. Combating Climate Change: A Role for UK Forests.
Krugman, P. (2010). ‘Building a green ‘The Synthesis Report, (Edinburgh: Stationery
Office).
economy, New York Times Magazine, 7
April, [Link] Schelling, T. (1997). “The costs of combating global
magazine/[Link]?pagewanted=all warming, Foreign Affairs, November/December:
(accessed 11 November 2011). 8-14.

Landis, EF, and Bernauer, T. (2012). “Transfer Shogren, J. (1999). The Benefits and Costs of the
payments in global climate change’, Nature Kyoto Protocol (Washington, DC: American
Climate Change 2: 628-33, [Link] Enterprise Institute).
com/nclimate/journal/vaop/ncurrent/full/ — and Toman, M. (2000). ‘Climate change
[Link] (accessed 1 June 2012). policy; in P. Portney and R. Stavins (eds.), Public
Lee, J.J., and Cameron, T.A. (2008). ‘Popular support PoliciesforEnvironmental Protection, 2nd edn.
for climate change mitigation: evidence from a (Washington, DC: Resources for the Future):
general population mail survey, Environmental 125-68.
and Resource Economics 41(2): 223-48. Stern, N. (2006). Review on the Economics of
Nordhaus, W. (2008). A Question of Balance:
Climate Change, HM Treasury, UK, October,
Weighing the Options on Global Warming Policies [Link] [Link]/+/
(New Haven, CT: Yale University Press).
[Link]
[Link] (accessed 30 May 2012).
Olmstead, S., and Stavins, R. (2012). “Three key
Tol, R. (2012). ‘On the uncertainty about the
elements of a post-2012 international climate
total economic impact of climate change;
policy architecture, Review of Environmental
Environmental and Resource Economics (online).
Economics and Policy 6: 65-85.
United Nations (1992). United Nations Framework
Pachauri, R.K., and Reisinger, A. (eds.) (2007).
Convention on Climate Change (New York:
Climate Change 2007: Synthesis Report.
United Nations).
Contribution of Working Groups I, II and
III to the Fourth Assessment Report of the —— (1997). Kyoto Protocol to the Convention on
Intergovernmental Panel on Climate Change Climate Change (New York: United Nations).
(Geneva: IPCC). Weitzman, M. (2011). ‘Fat-tailed uncertainty in the
Pindyck, R. (2011). “Uncertain outcomes and economics of catastrophic climate change;
climate change policy; Journal of Environmental Review of Environmental Economic Policy 5:
Economics and Management 63: 289-303. 275-92.

Pizer, W. (2002). ‘Combining price and quantity Wigley, T., Richels, T., and Edmonds, J. (1996).
controls to mitigate global climate change, “Economic and environmental choices in the
Journal of Public Economics 85: 409-34. stabilization of atmospheric CO, concentrations,
Nature 379: 240-3.
Plambeck, E., and Hope, C. (1996). ‘An updated
valuation of the impacts of global warming; World Bank (2010). Economics ofAdapting to Climate
Energy Policy 24: 783-93. Change (Washington, DC: The World Bank).
This chapter gives an economic analysis of the market and non-market value of forestry.
This includes natural forests, modified natural forests, and timber plantations. The reason
why we treat forests in a separate chapter is that they are critically important in providing a
range of ecosystem services, especially as a host to biodiversity and a store of carbon.
e Wereview the costs and benefits of conserving natural forests.
e Weanalyse why natural forests are being cleared.
e Weconsider how the optimal area of forest might be determined.
e We discuss local and international policy responses to the problem ofdeforestation.

10.1 Introduction

The destruction of large areas of natural forest, especially tropical rainforest, has become a
major environmental issue. Many environmentalists believe that natural forests are being
converted to other land uses too rapidly, and that this process is causing profound damage
to the world’s environment and ecosystems. In addition to forest loss, there is also wide-
spread forest degradation as a result of fragmentation and the disturbance of vegetation.
A different view of deforestation is that it is a necessary and inevitable part of the
development process: land under natural forest is converted to more economically pro-
ductive uses including agriculture, commercial forestry, and urban expansion. The coun-
tries with high rates of deforestation are repeating a pattern of development followed in
medieval Europe, seventeenth- to nineteenth-century North America and nineteenth- and
twentieth-century Australia, where forests were cleared on a massive scale for agriculture,
mining, and urban uses.
From an economist’s viewpoint, natural forests may be excessively depleted due to a fail-
ure of market price signals to reflect the wide range oflocal, national, and global benefits that
natural forests provide. In contrast, the benefits of deforestation are expressed through tim-
ber markets and agricultural commodity markets, and without government intervention,
FORESTS 195

rates of deforestation may continue to be above the social optimal level in many countries.
Government policies can also increase deforestation rates; for example, by subsiding invest-
ment and tying land ownership to land clearing. The benefits of conserving forestry are as
follows.

10.1.1 Local benefits

For instance, tropical forests are home to many millions of people, including numerous
indigenous groups. The forests provide a range of outputs, including timber and other wood
products as well as non-timber forest products (NTFP) such as edible fruit, oils, latex, fibre,
and medicines, which can be consumed or sold. Estimates of the number of people who
utilize NTFP are between 0.955 and 1.455 billion (Scherr et al., 2004). They utilize the forest
in many ways. The largest proportion, 50-75 per cent, are those who manage remnant forest
for subsistence and income as part of a farm. Others, approximately 25 per cent, depend on
the forest for supplementary and ‘safety-net’ income, especially when their main source of
income, possibly agriculture, is adversely affected by drought, for instance. People that
depend largely on the forest as a source of income for hunting and gathering are between
4 and 6 per cent of the total. Standing forests through renewable NTFP are capable of yield-
ing higher net returns per hectare than timber (Peters et al., 1989; Shackleton et al., 2011).
Local people often make their living practising shifting cultivation and collecting a wide
variety of forest products to sell in local markets. The income of poor landless farmers
depends in many tropical regions on the availability of forest resources for shifting cultiva-
tion, bushmeat, and supplies of fuelwood. This type of benefit is not restricted to tropical
rainforests: many natural and partially modified forests across the world provide similar
market and non-market benefits. In Box 10.1, we discuss the nature and extent of ecosystem
services derived by farmers in the Bragantina region of Brazil.

10.1.2 National benefits

At a regional level, forests perform complex ecosystem functions, including the regulation
of flows of surface and groundwater, and the protection and enrichment of soils through
reduced erosion and nutrient recycling. Forests act as a sponge, releasing water at a steady
rate and evening out variable precipitation rates (for a discussion and a more complex story
that indicates that forests are good at moderating minor floods, but have little effect on
major flood events, see FAO, 2005). Once a rainforest is cleared, rainfall runs off the land
more rapidly, potentially causing increased flooding and soil erosion. Tropical forests also
maintain a balance of species by providing pest control services, and they regulate surface
temperatures and local and regional climates through evapotranspiration.

10.1.3 Global benefits

Greenhouse gas emissions. Forests make a significant contribution to climate regulation

through their role in the global carbon cycle. Forests store carbon in vegetation and the soil
and take carbon out of the atmosphere. Deforestation and forest degradation release stored
carbon into the atmosphere as carbon dioxide emissions. Forest growth sequesters carbon
196

Secondary regrown tropical rainforest is important in many areas of Brazil, where the original forest has
been cleared many years ago, but farmers engaging in slash-and-burn agriculture allow forest fallow; that
is, a period when cultivation ceases and the forest is allowed to regenerate. Forest fallow provides many
of the same market and non-market ecosystem services as the original forest: it sequesters carbon,
reduces soil erosion, increases soil fertility, and reduces the damage due to flood events. Klemick (2011)
interprets forest fallow for between 3 and 8 years as providing private benefits to the farmers with forest
fallow on their land and positive externalities, in the form of reduced flood frequency and intensity, to
neighbouring farmers that share the same water catchment.
Bragantina is a shifting cultivation region of the Brazilian Amazon. The author combined farm survey
data with satellite land cover data to estimate a production function to explain the value of farm output.
In a simplified version, the production function is as follows:

Farm output $ = f(on-farm fallow (ha), all-other farms fallow,

cultivated land, labour, fertilizer).

The study estimated a Cobb-Douglas production function that regresses the logarithm of output on
the logarithm of all the variables. This means that the coefficients of the regression can be interpreted as
elasticities. The results find that a 1 per cent increase in the area of fallow on farm increases the value of
output by 0.135 per cent (Klemick, 2011: 101, table 3, model (3)). The positive external benefit derived by
the farm from a 1 per cent increase in the upstream fallow area is 0.355 per cent.
Determining the optimal allocation of land to fallow requires an investment analysis that accounts for
the opportunity cost of land in fallow (in terms of lost output) and the opportunity cost of labour
(required to clear the fallowed land at the end of the fallow period). Farmers allocate land efficiently
between cultivation and fallow if the marginal benefits of cultivated land minus the marginal costs in
terms of land clearing, forgone soil quality, forest products, and positive externalities to other farms
equal zero.
Evidence from Bragantina using the results from the estimated production function is that, if anything,
farms over allocate land to fallow and take account of the positive externality that they have on other
farmers. This slightly surprising result may be due to the fact that Bragantina farmers have well-
established private tenure of their land and are therefore willing to invest in soil conservation through
forest fallow and, possibly inadvertently benefit their downstream neighbours through the hydrological
services provided by forest fallow.

from the atmosphere. Globally, the forest sector emits approximately 5.8 Gt of carbon diox-
ide a year from deforestation. The Eliasch Review (Eliasch, 2008) estimates that halving
deforestation rates by 2030 would reduce global greenhouse gas emissions by 1.5-2.7 Gt of
CO, per year. The same review estimates that halving emissions from the forestry sector
from 2005 to 2030 could be worth between US$17-33 billion per year if forests were included
in global carbon trading. The net present value of benefits is estimated at US$3.7 trillion
between 2010 and 2200. This figure does not include the many other local, regional, and
global benefits of forest ecosystems.
Biodiversity store. Tropical rainforests are a rich source of biodiversity: they cover 7 per
cent of the earth’s surface but contain 50 per cent of all species (Brown and Pearce, 1994).
The gene base present in tropical rainforest is a commercially exploitable natural resource,
which contains biologically active compounds with medicinal properties (see Chapter 12).
FORESTS 197

10.2 Forest Distribution and Losses

Forests can be classified according to tree type and physiognomy (overall physical structure
or development stage) and phenology (concerning phenomena, such as leaf fall, and their
relation to climate). The most aggregated classification, based on climate and tree type, is
into: temporal and boreal needleleaf; temporal broadleaf and mixed; tropical moist and
tropical dry; and parkland (Groombridge and Jenkins, 2002: 81). Trends in deforestation
and reforestation vary regionally and according to forest type.
This section summarizes the main trends in the world’s forests. The state of the world’s
forests is assessed every five years by the Global Forest Assessment (see, e.g., FAO, 2010). In
2010, the world’s total forest area was estimated at slightly over 4 billion hectares, with just
five countries—the Russian Federation, Brazil, Canada, the United States, and China—
accounting for more than half of the total area (53%).
The loss of primary (undisturbed) forest and naturally regenerated forest remains alarm-
ingly high. FAO (2010) estimates this as at around 13 million hectares of forest lost each
year from 2000 to 2010. The land is either converted to agriculture or lost due to natural
causes such as bushfires. This is actually a reduction in the even higher rate of loss that
occurred in the 1990s, of 16 million hectares in each year. The loss of 13 million hectares of
forest entails a major loss in biodiversity, as it is often the undisturbed forest that is lost.
Offsetting some of these losses is an expansion in natural forests and an increase in planta-
tion forests. This gives a net annual forest loss of 5.2 million hectares.
The distribution of forest area and forest loss is given in Table 10.1. This shows that although
the rate of net forest loss is slowing, especially in South America and Asia, the rate of loss is still
greatest for the tropical nations that host the forests with the highest level of biodiversity.
One issue with considering data such as that presented in Table 10.1 is that sums across
different forest types simplify the patterns of forest loss and gain. Plantation forest may be a
close substitute for primary and natural regenerating forest in terms of timber and some
ecosystem services, such as moderating soil erosion and the local hydrological system, but a
poor substitute as a host for biodiversity. Thus the continued loss of primary forest in Brazil,
at an average rate of2.3 million ha per annum between 2005 and 2010 (FAO, 2010: table 8),
and its corresponding loss in biodiversity is not compensated for by a similar increase in, for
instance, the area of planted forest in China (2.0 million ha expansion per annum, 2005-10)
(FAO, 2010: table 9).

10.3 Why is Rainforest Lost? Economic Theories of

Deforestation
Rudel and Roper (1997) assess two competing economic explanations of deforestation, the
frontier model and the immiserization model. The frontier model assumes that capital invest-
ment is the main determinant of deforestation, with the timber industry playing a key role. The
immiserization model assumes that increasing population and impoverishment is driving the
conversion of forestry to agriculture. We now consider each of the models in more detail.
198 APPLYING THE TOO LS

L90-S0'0-9¥'0- y0'L €€0- 800 090 £00 SS0-

4A/euQO0'L
0LO0Z-S00z
Ly- 9€S'L- LSS'7Z
ZE8'L- L66- OLL Lr 87Z CO

-
ayes $9°0-S0'0-Sv 67 alive 900 060 Z00 LO
% 0-
asueYD

4K/ey000'L
s00Z-0002
jenuUYy Crs’ Lp- 665 S00'€€9€- 786 6S 8rL LCE
L- l=

79'0-Ceov Lg'0 L0= 600 180 su Z0'0-

% 0-
4A/e4000'L
0007-0661
06S-LE9'|- Z9L'Lety
Lv8'L- LL8 €S ZE 9€-
e=

ease
0”s'SS6
eunear 0S9'7Z0' 966'L68
L6Z'€S0'LLZ7v'OLl 9LZ'90€'Z
pLy'e@ BLL'66L'Z
Ev
Ayjuno> L'9S8

LISL9% Bg0'8ze_ 979'ySZ

yL8'8Z ELE'p6Z LOO'SOO'L
€€6'9 196849 p8ElL6L
OLOZ

= ss =
(ey 6LO'6LOLL'SEE LPS'LhZ
6L9'9LZ OSL'LOO'L
LZE'OET-—S—“isé*EL'LOE
87L'9 E7819 SPZ'96L
000'L)s00z

Bese ss“ = ~=—s

OGH'SSZ
y@7 revere SLB'9ZZ_~—s 6€7'866
yey'9 EBO'LL9.—sL8E'86L
ysa4040002
6L
SBULYD

= = =
‘(€ pur | Sajqey ‘saiqeL feqo|d ‘OL0Z) OW :a2nos

EZL'SB€08'6SE BUL'60Z
ZLE'POE ET'SZE Z06'S
LLY’'686 pL'9LO vr
066L L'86L
Jsa1OJBale PUR

§=—BDy'W POLY
[EUOIZaY

BIS
WayINOS[PJUSD 1S29-YINOS

BIL BOUBWY
puke PUR
aIGeLLOL

UJSYLON puke ueaqque)

Wa}sey Welse BIS adouny eBIUeZO
Pay YINOS YON
1Se3
6Lb'€-
POLY
¥9S'80L
89r'lL69
6lr'7Z9
L8r'LEO'e
67
OLr'e-
£90'p-
0-
95
0S8€Z'6rL 9S0'80L'L
CL
LL0
Lr
68r'
Zr
LOZ
088'Zr
ELS'Eb
SEL
ZE0
/ZL
PUP
[BNJUSD
670=~
BISY
W9}Sa/\
Blige
Nig
ZLS'76S
ELy'Z8L'€
S6S-
OL0-
yOL'OLS
8h0'¥8S
8r0
€69'|
670
eis al Ov-
OLL'9ZS 66r'6L
LOL'ZS
yle-
99'L-
Lye-
LLL'SZ
086'LZ
Spl'0z
SlL-
6yc-
[esJUSD
eoUalWy 967502
€6€S0L
PLZZ
ELS'
682-
0'0-
LO'0-
6L
pue
|esjUeD
Baa
YON
su
L6r'SOL
€8€'80Z=
ELb'b-
LSe'p98
OvO'ESL'L
ELZ'b-
Sr'0-
~ZE'v06
857788
0-
6y
18S'€-
Ly'0-
96
vS
YINOS
BOLEWY
PEV'EL
ZEC
EZE'B-
07'0-
Ly8'b-
891'S80'r
¥96'090'r
ZL'0-
Lgs'¢-
66€'89L'r
vLO-
PHOM
O90'EED'Y
FORE 199

The frontier model identifies networks of entrepreneurs, companies, and small farmers as
the chief agents of deforestation. Together, this group has sufficient power to raise private
capital and obtain assistance from the state to open up regions for timber extraction, settle-
ment, and deforestation. For an initial period, while roads are being established and land-
ownership is contested, farmers have a strong incentive to clear land rapidly to state a claim
or to extract resources before they are appropriated by others. This is an example of an
open-access resource (for a more detailed analysis, see Chapter 7). Investment-driven
deforestation will open up whole regions to settlement and agriculture; however, once this
process has taken place, there may be a pattern of more gradual deforestation as farmers
incrementally expand their cultivated area.
In this form of deforestation, population growth increases the size of the surplus labour
force that is willing to enter remote regions and clear land. Gross national product (GNP) is
a measure of the funds available to invest in large capital projects such as road-building and
timber companies. Timber companies contribute to deforestation directly by harvesting,

The Amazon encompasses the world’s largest moist forest region, at 5.5 million km?. Of this, roughly
3.8 million km? lie within Brazil. The Brazilian forests are home to as many as one-fifth of the world’s
plant and animal species. Brazil has the world’s largest remaining tropical forest, and by far the largest
area of annual deforestation.
In Brazil, most of the extensive deforestation can be traced directly to government-financed
programmes and subsidies. Cattle-ranching has been the foremost cause of forest conversion, with small
farm settlements (mainly government promoted) the second largest cause of deforestation. Small farms
have accounted for about 11 per cent of the Amazon’s total deforestation up to 1983. Other large
projects developed partly to alleviate Brazil's foreign debt crisis have deforested more areas. These
include several huge hydroelectric investments such as the Tucurui Hydroelectric Project, which cost
USS$4 billion, and flooded 2,160 km? offorested land.
The government of Brazil has engaged in massive efforts to colonize its tropical forests with small
farms; for example, along the Transamazon Highway. The Northwest Development Programme
(Polonoroeste) encompasses the entire state of Rondonia and part of the Mato Grosso. The government
undertook to demarcate plots and establish land titles, and by mid-1985 the responsible agency had
awarded 30,000 titles, mostly for 100 ha farms. Settlers only paid a nominal fee for their land (US$1) and
could recover their relocation costs by selling timber. Once they had obtained title to the land, they
become eligible for subsidized agricultural credits. It has been estimated that subsidized settlers cleared
almost 25 per cent more forests than those not benefiting from government programmes (Repetto,
1988).
Polonoroeste was intended to promote sustainable farming systems, based on tree crops, and to
include environmental protection. Instead, unguided colonization took place. Settlers poured into the
state of Rondonia increasing its population at an average rate of 14 per centa year from 1980. The wave
of settlers in Rondonia cleared and burnt the forest, and accelerated the rate of deforestation from an
area equating to 3 per cent of Rondonia in 1980, to an area that accounting for 24 per cent by 1985
(Repetto, 1988). The deforestation was in part a direct result of existing government policies. First,
clearing the land was required as evidence of ‘land improvement’. The settlers were then able to claim
title to an amount of land in direct proportion to the area of forest cleared. Poor settlers who did not
benefit from tax incentives offered at the time gained more by clearing the land, selling it to large
cattle-ranchers, and moving on.
200

and indirectly by building roads, which are then used by farmers and ranchers to gain access
to remote areas (see also Box 10.2).
The immiserization model assumes that individual farm household decisions are the
source of deforestation. In a country with low levels of GNP, foreign debt and tight fiscal
policies lead to a low level of growth in non-farm employment and income. The lack of
off-farm opportunities and increasing household size mean that the opportunity cost of
labour is low and this provides the household with an incentive to expand the cultivated
area to more marginal land areas. The immiserization model has strong Malthusian over-
tones (see Chapter 6), in that Malthus predicted that population growth will continue until
the family drives its per capita food consumption down to subsistence levels, leading to a
‘dismal’ equilibrium where the family just meets it subsistence needs. Walker (1993) con-
cludes that deforestation due to poverty amongst farmers tends to be more local and incre-
mental than the frontier model, in which whole regions are opened up to deforestation (see
also Box 10.3).
In the immiserization model, economic growth, as measured by GNP, will reduce the rate
of deforestation in two ways: growth increases off-farm opportunities so that labour moves
out of agriculture, and this reduces the incentives for deforestation. A country’s debt burden
affects deforestation in two ways. First, the government’s requirement for foreign exchange
to pay debts encourages expansion of export sectors, which include cash crop production
and timber. Second, the devaluation of local currency that often accompanies a debt crisis
tends to increase the profits of timber companies and leads to increased investment and
expansion of logging activities. Finally, an increase in the rural population leads to the
movement of agriculture to more fragile marginal land.
Testing the models. On the basis of a cross-country study, Rudel and Roper (1997) con-
clude that the evidence gives more support to the immiserization model than the frontier
model. The main justification for this conclusion is that economic activity, as measured by

of Immiserization

The Ivory Coast has lost tropical rainforest at the rate of 300,000 ha per year during the 1980s and of an
original area of 16 million ha, only 3.4 million ha remains. Deforestation is due to an increased demand
for land for shifting cultivation. This demand is driven by population growth, rising poverty, and
ill-defined property rights over forest land. Deforestation has led to soil erosion, a reduction in
agricultural productivity, and the siltation of waterways. Attempts by government to define protected
areas have been frustrated by the continued loss of forest to slash-and-burn agriculture.
In their paper, Ehui and Hertel (1989) estimate the costs of deforestation through average agricultural
yields. As deforestation proceeds, average yields across cultivated land decline due to soil erosion anda
reduction in soil fertility. The socially optimal equilibrium is where both the forest area and the average
agricultural yield are constant. The outcome is highly sensitive to the discount rate. For instance, if the
discount rate is 3 per cent (r= 0.03), the optimal forest area is 5.4 million ha, which implies some
reforestation. If, however, the discount rate is 11 per cent (r= 0.11), then the optimal forest area is
2 million ha. In a country that is prone to economic instability, high interest and discount rates tend to
be the norm. On the basis of this analysis, without government protection the rainforest area in the Ivory
Coast will continue to decline.
201

the GNP per capita, is not found to be a significant determinant ofdeforestation: this counts
against the frontier model as an explanation. However, the choice between the two models
is not clear-cut and explanations of deforestation vary from country to country. In
low-income countries with small remnant forest areas—for instance, the Ivory Coast—
population growth appears to increase the rate of deforestation. In countries such as Brazil,
with large blocks of forest, the frontier model is more appropriate. It is also possible to dis-
cern a progression, where initial deforestation is due to capital investment in roads, which
is then followed by agricultural expansion that may be driven by population growth.
Different pressures for deforestation may wax and wane, depending upon the cycles in a
country’s economy. If the economy is growing strongly, then this leads to frontier expan-
sion and reduced dependence on subsistence agriculture. If the economy moves into reces-
sion, then frontier development declines, but deforestation for subsistence agriculture
increases as a large proportion of the labour force turns to agriculture for a livelihood as the
availability of urban jobs declines.
To generalize this relationship, Rudel and Roper predict the relationship, shown in
Figure 10.1, between the proportion of the forest area logged and GNP:
® Stage 1. At low levels of GNP, deforestation is due to impoverished peasant farmers
clearing more land to meet their subsistence food needs. Funding is not available to
finance an expansion in the frontier and open up whole regions for deforestation. As
the GNP increases and off-farm labour opportunities improve, then the rate of
deforestation declines as labour moves out of agriculture.
e Stage 2. As economic development proceeds, investment capital becomes available
from government and the private sector, and the frontier is opened up by investment in
roads and other infrastructure.

Deforestation Stage 1 | Stage 2 Stage 3

GNP per capita

Figure 10.1 The deforestation environmental Kuznets curve.

202 APPLYING THE TOOLS

e Stage 3. The level of wealth increases such that consumers have a demand for forest
protection, and the labour force is less dependent on agriculture due to the
development of jobs in manufacturing and service sectors.
Rates of deforestation may thus follow an environmental Kuznets curve (EKC). The pos-
sibility of a positive relationship between environmental protection at high income levels is
a recurring theme that has already arisen in Chapter:6 and Chapter 8 in relation to trade.
Other studies have confirmed the results of Rudel and Roper. There is evidence from cross-
country studies by Ehrhardt-Martinez et al. (2002) and Shandra (2007) to support the pres-
ence of an EKC: the former find an inverse relationship between the rate of deforestation
and income per capita in a country, whilst the latter finds that, on average, higher rates of
population growth increase the rate of deforestation.

10.4 Tropical Deforestation and Poverty

The UN established the Millennium Development Goals in 2000 to address key measures of
development, with a deadline of 2015 for the world community to achieve these goals. The
goals relate poverty and hunger alleviation, education, gender equality, child health, HIV/
AIDS prevention and treatment, environmental sustainability, and global partnership
(largely concerned with opening markets to exports from developing countries). In the con-
text of this chapter, a key question is: what role does forestry have in fostering development
and, specifically, the reduction of poverty? In a review paper, Sunderlin et al. (2005) con-
clude that this is a complex question, as in some cases conserved forests reduce poverty and
in others deforestation, by increasing the availability of agricultural land, reduces poverty
and hunger.
Who depends on tropical rainforests? Some forest dwellers are traditional/indigenous
peoples and their poverty is often ‘primordial’ (Sunderlin et al., 2005) and is not due to eco-
nomic change. By ‘primordial’, the authors mean that the forest dwellers were relatively
impoverished prior to large-scale deforestation and economic development. Incoming
migrant farmers migrate to forest areas to increase their incomes and in this case the forest
provides potential for increasing income. Forests are in some regions a refuge for landless
people, as the forests are often treated as open-access resources.
For all these groups, the ability of forestry to provide income depends upon how the peo-
ple interact with the forestry resource. The use of forestry can be classified as hunting and
gathering, shifting cultivation, and sedentary cultivation. These uses can be seen as a pro-
gression from livelihoods that require a high forest density (hunting and gathering) through
a moderate level of forest cover (shifting cultivation), to low levels of forest cover (sedentary
agriculture).
Forests have the potential to contribute to market and non-market based determinants
of poverty reduction. Timber sales worldwide (2003-7) had an annual value of US$100
billion (FAO, 2010: 120). In 2005, the value of non-timber forest products was estimated
at US$18.5 billion (FAO, 2010: 120). Values from timber production do not always ben-
efit the poor, as they typically do not have ownership ofthe resource. Timber production
FORESTS 203

Protected Fore

In a detailed spatial analysis from Thailand, Sims (2010) assesses whether people living in a region
with a relatively high proportion of protected forest areas are poorer compared with regions with a
relatively low proportion of protected areas. One problem with undertaking this analysis is that
historically protected areas have been established in regions with low development potential and
high levels of poverty. Once this tendency has been corrected for in a regression analysis, the results
indicate that a high proportion of protected area tends to reduce poverty rather than increase it. Sims
considers this to be due to an increased income from tourism offsetting the costs of restrictions on
land use.

is capital and skills intensive, and this can exclude impoverished forest dwellers from
employment in the industry. The more significant beneficial effects of forestry to farmers
are through the positive externality due to increased ecosystem services and access to
non-timber forest products.
It may be that the extraction of timber generates the highest short-term monetary
(market) income and it is relatively straightforward to assign ownership to timber
through timber concessions. This is in marked contrast to retained forest that provides
non-monetary (non-market) ecosystem services to groups benefitting through hunting
and gathering and shifting agriculture. These groups may be relatively poor and thus
even a non-market valuation would tend to return a low willingness to pay (WTP)—
even though the forest resources account for a significant part of their income. One pol-
icy solution proposed by Sunderlin et al. (2005) is to strengthen community ownership
over the forest. This may mean that these non-timber values are taken into account in the
decision to clear an area of forest. Community ownership may also benefit the relatively
poor, although evidence on this is mixed (for a review, see Shackleton et al., 2011) (see
also Box 10.4).

10.5 Forestry Management

In this section, we consider two problems: first, how should forests be managed when the
standing forest is a commercial enterprise that is managed for its timber values alone; and,
second, how should the standing forest be managed when it also has an external (non-tim-
ber) value?
Forest managed for timber value. The problem of how long trees should be left to grow
(i.e. how long the forest rotation should be) to maximize the present value of profits for
multiple rotations was solved in 1847 by a German forester called Martin Faustmann. As a
stand of single-aged trees grows through time, so does its timber value V(T). This is shown
in the top half of Figure 10.2. The value oftrees is set against their opportunity cost; that is,
the return on money tied up in trees that could earn a rate of return r if invested elsewhere.
The bottom half of Figure 10.2 shows the rate of return on trees as the percentage growth in
value. This is given as the marginal growth in value (MV) divided by the value of the trees.
204 APPLYING THE TOOLS

timber
V(t),
value
$
G(t)
value
non-timber

values
cent
per
non-timber
and
value MV+G,
timber
of
Rate
return
on
Se oes | V
PS SS SSS = === MV+G,

Figure 10.2 The optimal timber rotation.

This value is compared with rand when the two are equal, as at a in Figure 10.2, it is optimal
to fell the forest and sell the timber. At this point, the present value of the forest is maxi-
mized. Note that this is a shorter time than T,,,,,, the time when V(T) reaches a maximum.
The original Faustmann formula includes an adjustment for the site value; that is, the oppor-
tunity cost of the land, in terms of delaying the start of all future rotations. Ifthe interest rate
rises, then the optimal rotation length falls; for instance, at r,, the length of the rotation
reduces to T,.
Forest managed for timber and non-timber value. If the forest has a non-timber value—
for instance, in terms of its recreation value, or its value as a biodiversity refuge—then this
changes the socially optimal forest rotation. Instead of maximizing the present value of
timber, the forest is managed to maximize the present value of timber and non-timber val-
ues. Unlike the timber value, which is only realized at the time when the forest is clear-felled,
the non-timber value is given as a flow of benefits in each year. The line G,(t) in the upper
half of Figure 10.2 shows the value of these non-timber benefits. Note that the non-timber
benefits G,(t) continue to increase with the rotation length, reflecting the fact that wildlife
habitat values tend to increase with the length ofthe rotation. This contrasts with the timber
value V(T), where the forest is assumed to reach an age at which it starts to have a declining
timber value due to tree death and senescence. The optimal rotation, at which the rate of
return on the total value equals the discount rate ((MV+ G,)/V) =r, at discount rate r,, is
now longer than the yield-maximizing rotation T,,,, at T;. It may be optimal to never har-
vest the timber because of the size of the non-timber benefits; this is illustrated for a level of
non-timber benefits G, in Figure 10.2.
FORESTS 205

10.6 How Much Natural Forest Should be Preserved?

10.6.1 The optimal area of natural forest for a region

In most regions, the loss of native forest is either entirely or partially irreversible. Natural
forests have much in common with non-renewable resources like oil where extraction per-
manently depletes the stock. Once a forest has been cleared and the land converted to agri-
culture or abandoned, soil changes and the long time period required for the forest to return
to its original state means that the natural forest is effectively lost for good. This is a simpli-
fication, since in some regions partial regeneration is possible and some of the biodiversity
and ecosystem services of the original forest might be recovered in a period of 10-30 years.
The annual benefits from a forest comprise three components. The aggregate market ben-
efits (net of costs) of the annual area of forest felled q, given as a function h(q, x,), where x,
is the remaining forest area at time f. The market net benefits derive from the net benefits of
logging and increased agricultural productivity, as more land becomes available for agricul-
ture. For any given level of deforestation, the marginal benefits to the farmer of clearing
decreases with the cumulative area logged. This is explained by forest clearance moving to
land of progressively lower agricultural value. The benefits of preserving forest derive from
two sources. Local benefits include the market and non-market benefits derived from forest
products and flood protection b!(x,). Global benefits include the benefits of biodiversity,
which include the possibility of discovering new pharmaceutical products, and the exist-
ence values to citizens of other countries derived from knowing that the forest is protected,
this benefit being given by the function b°(x,). Both benefit functions are an increasing func-
tion of the forest area, with a declining marginal benefit. Thus as the cumulative area cleared
increases, the marginal benefit of preserving the last hectare increases. Together, the net
benefits of preservation are as follows:

D(x; ) = b' (x, ) ais bE (oy):

A socially optimal forest area is one in which the marginal benefits of clearing equals the
present value of the marginal benefits of preservation. The marginal benefits of clearing in
terms of timber and agriculture are assumed to be generated over a relatively short period:
immediately for timber and over a relatively short period, say five periods, for agricultural
land. The benefits of preservation are assumed to be forever. Therefore, the present value of
preservation benefits is given by the formula b(x,)/r, where ris the discount rate (see Box 7.3
about discounting), as in the Krutilla~Fisher model of wilderness conservation. The social
optimum is where the present value of the marginal benefit of harvesting, mh, equals the
present value of the marginal benefits of preservation, mb:

mh = mb/r = (mb! + mb°)/r. (10.1)

The social optimum is illustrated in Figure 10.3. The initial forest area is x9. The opti-
mum solution that accounts for the local and global benefits of preservation is at c, with a
conserved forest area of x* where the condition in equation (10.1) holds. If, however, there
is no mechanism to account for (to pay forest owners for), the global values of rainforest
206 APPLYING THE TOOLS

Forest area

Figure 10.3 The equilibrium natural forest area.

preservation, the equilibrium is b at a lower level x**. If the forest is open access and there
are no effective policies for its protection in place, the equilibrium is driven down to x,
where mh = 0. Note, from equation (10.1), that the discount rate is a key parameter: if this
is high, then the present value ofpreservation is reduced. An example ofthis is given in Box
10.3, for the Ivory Coast. The discount rate on which firms base their decisions will be
biased upwards if there are imperfections in the capital market or the government, in plan-
ning natural resource use, adopts a discount rate that does not reflect society’s true rate of
time preference. There is also no incentive for the country to account for the benefits that
other countries derive from the forest. This problem has much in common with the prob-
lems ofproviding public goods discussed in Chapter 2. Putting all of these factors together
goes a long way to explaining why forest losses across the globe have been excessive and
represent a market failure.

10.7 Policies for Rainforest Conservation

10.7.1 International policy

Forestry has been an international policy priority since the 1992 United Nations Conference
on Environment and Development (UNCED). UNCED led to a Non-legally Binding
Authoritative Statement of Principles for a Global Consensus on the Management,
Conservation and Sustainable Development of all Types ofForests, also known as the ‘Forest
Principles’, and Chapter 11 of Agenda 21, ‘Combating Deforestation’. This was followed by
the Intergovernmental Panel on Forests (IPF), from 1995 to 1997, and the Intergovernmental
Forum on Forests (IFF), from 1997 to 2000, both under the UN Commission on Sustainable
207

Development. These have led to the IPF/IFF Proposals for Action. The IPF/IFF proposals
are not legally binding, but they do provide organizations in participant countries with a
rationale for improving national forestry policy.
The UN Forestry Forum (UNFF) was established by ECOSOC Resolution/2000/35 to
build on IPF and IFF processes. The resulting Non-legally Binding Instrument on All Types
of Forests was adopted in 2007 by the UN General Assembly (Resolution 62/98). The pur-
pose of this resolution is as follows:
a. To strengthen political commitment and action at all levels to implement effectively
sustainable management of all types of forests and to achieve the shared global
objectives on forests.
b. To enhance the contribution of forests to the achievement of the internationally agreed
development goals, including the Millennium Development Goals; in particular, with
respect to poverty eradication and environmental sustainability.
c. To provide a framework for national action and international co-operation.

10.7.2 The Global Environment Facility (GEF), REDD, and REDD+

The Global Environmental Facility (GEF) was established in 1991 to provide funds to
address global environmental issues. Since 1991, the GEF has allocated US$9.5 billion of
core funding and leveraged approximately US$42 billion in co-financing (GEF, 2012a).
Most of the Focal Areas for GEF 5 are directly and indirectly related to natural forest protec-
tion (GEF, 2012b) (see also Box 10.5).
The UN-REDD Programme is the United Nations Collaborative Initiative on Reducing
Emissions from Deforestation and Forest Degradation (REDD) in developing countries.
The programme was established in 2008 to provide funds to developing countries to prepare
projects that reduced greenhouse gas (GHG) emissions from deforestation. REDD also
involves the Food and Agriculture Organization of the United Nations (FAO), the United

in Brazil

The GEF project, the Amazon Region Protected Areas Program, has three specific objectives:

1. To identify and create new strict protected areas.

2. The effective establishment of these new areas.
3. To develop long-term sustainable management tools and mechanisms for effective protection within
all Amazonian strict protected areas.

The project emerges from a Government of Brazil commitment to extend strict protection in the
Amazon to cover at least 10 per cent (37 million ha) of the biome (370 million ha). Although it is expected
that all costs associated with the project will be of global benefit and incremental, GEF grant funds will act
as seed capital to catalyse additional funds. The total cost of the project from GEF and other sources was
US$82 million (GEF, 2012b).
208

Carbon Offset Project

The Noel Kempff Mercado Climate Action Project (NKMCAP) is a deforestation avoidance project
established in 1997, in the Santa Cruz Department of north-eastern Bolivia. It is a pilot project under the
Kyoto Protocol Clean Development Mechanism and also an example of how REDD and REDD+ may
operate. The project, by buying out logging concessions, expanded the area of the Noel Kempff National
Park by 70 per cent (Pereira, 2010), with the primary aim of reducing carbon emissions by storing carbon
that would have been released by timber harvesting and other forms of deforestation (NKMCAP, 2008).
NKMCAP is a partnership between the Bolivian government, the Nature Conservancy (an international
NGO), a Bolivian NGO (Fundacion Amigos de la Naturaleza), and four multinational energy corporations
(American Electric Power, British Petroleum Amoco, and PacifiCorp). The energy companies contributed
to the US$11 million cost of the project. In return, they received carbon offsets. A community of 237
indigenous families were supported to apply for land title and also received other economic benefits,
including work in the park and work in the ecotourism industry. This project is an example of where a
payment for ecosystem services is combined with carbon abatement.

Nations Development Programme (UNDP), and the United Nations Environment

Programme (UNEP) (UN-REDD, 2008). Box 10.6 gives an example of how the REDD
mechanism might work, with an example of where payments for environmental services are
partially derived from carbon offsets.
Agreement was reached at the COP16 (Conference of Parties to the Convention on
Biodiversity) in Cancun, Mexico, in 2010 to reinforce REDD through the REDD+ agree-
ment. This scheme extended the REDD agreement to include multiple benefits of reduced
deforestation, ecosystem services, and biodiversity protection, in addition to GHG emis-
sion reduction:
REDD+ goes beyond deforestation and forest degradation, and includes the role of conserva-
tion, sustainable management offorests and enhancement of forest carbon stocks.
(UN-REDD, 2010)

The funding for REDD+ is around US$30 billion and this represents a substantial invest-
ment in conserving forestry in developing countries.

10.7.3 How effective are international policies?

Are international treaties effective? On the basis of the current rate of deforestation, which
is still running at above 13 million ha a year, the short answer to this question is ‘no’. The
longer answer is that, while deforestation rates remain very high, there are some reasons to
be optimistic. The FAO (FAO, 2010: ch. 8) notes that international policies feed through
into an increase in the number and scope of national forestry programmes and policies.
Independent evidence comes from Schofer and Hironaka (2005), who find, from a statisti-
cal analysis of deforestation, that participation in international institutions such as the
UNFF has led to a significant estimated reduction in the rate of deforestation.
In the same paper, the authors hypothesize that ‘institutions affect outcomes when they
are: 1) highly structured; 2) when they penetrate actors at multiple levels of the social
FORESTS 209

system; and 3) when they are persistent over time’ (ibid.: 25). The last of these conditions
applies as the UNCD brought deforestation to centre stage almost 20 years ago. As with
the biodiversity conventions discussed in Chapter 12, there is a danger that:
states routinely sign treaties to support global environmental norms, conformity to external
norms may be ceremonial, offering environmental protection in form but not in substance.
(ibid.: 26)
There is evidence of an increase in the enactment of national forest policies since 2000, with
around 70 per cent of countries enacting new forest programmes between 2005 and 2009
(FAO, 2010: 154). It is argued by FAO that this relatively large number of new items of leg-
islation have been driven by commitments to the UNFP.
There remains a fundamental problem that tropical rainforest is largely controlled by
developing tropical nations, whilst many of the benefits of preservation—including carbon
sequestration and biodiversity conservation—are global. The developed nations have the
technological capacity to take advantage of gene material, with commercial potential for
new crops and medicines that may be discovered in preserved rainforests.
International policy to protect rainforests has a strategic dimension, as does any problem
that involves the provision of a public good (Barrett, 2005). There is a strong incentive for
countries to free-ride: that is, benefit from reduced deforestation without paying their share
of the costs. Figure 10.3 illustrates a problem of strategic interaction. With no intervention
from the international community, then the optimum is at b, since this is where only
national benefits are recognized. However, this is an undesirable outcome and represents
the situation if there is no binding agreement to pay tropical nations to protect more of their
forest. If countries can agree to a side payment to the tropical nations, this payment must be
at least the area abc, which represents the tropical nations’ net costs in terms of the reduc-
tion in the benefits of clearing. The additional benefits to the international community
would be dgef. The tropical country may negotiate hard to make the side-payment as large a
share of the additional benefits as possible. A complication in this process is that some
countries may try to avoid paying their share of the side-payment, in the hope that others
will pay and allow them to benefit for free (for a more detailed analysis of this issue, see
Sandler, 1993).
The policy response to rainforest loss has been at the international, national, and local
level. National policies are the most important, as they determine the incentives that farm-
ers and timber companies face when taking decisions to clear an area of forest. International
policies have had a limited impact so far; this is largely because they express broad objec-
tives, without the resources needed by the tropical nations to agree to reduce the rate of
deforestation. However, international payments for forest conservation under REDD+ may
go some way to addressing this failure in future, by increasing financial transfers to tropical
countries conditional on reduced rates of deforestation and better forestry management.

10.7.4 National and local policies

International initiatives to conserve tropical forests depend critically upon national and
local policies within tropical nations. The objective of national policies should be one of
reducing the adverse effects of market failure and government failure (McKean, 1965).
210 APPLYING THE TOOLS

Government failure, where government intervention actually makes matters worse rather
than better, includes subsidies for forest clearing, the link between land title and clearing,
and price support for agricultural outputs such as palm oil. Areas in which government
policy may eliminate market failure include placing taxes on deforestation, establishing
effective property rights over forest land, creating conservation areas, and banning or regu-
lating clearance in sensitive areas.
Deforestation remains a major global environmental problem because of government
and market failure. The most significant market failure is the failure to account for national
and global non-market values. A further problem in many poor countries is that govern-
ment policy may be only partially implemented. International funds might be provided for
rainforest conservation but, due to corruption and weak administrative arrangements, may
not be spent effectively.

Summary
The international community agrees that tropical rainforests are valuable—they regulate
the local environment, shelter biodiversity, and store carbon. Market failure, market dis-
tortions, missing markets, and ineffective and unstable government in many tropical
countries have led to rates of deforestation that are greater than socially optimal rates of
deforestation.
Those who own or who have access to the forests will cease deforestation only when it is
profitable to do so (Mendelsohn, 1994). This can come about through national governments
offering incentives to stop deforestation. In turn, this may only happen if there are funds
from the international community to account for the global public goods provided by tropi-
cal rainforests.

Tutorial Questions

10.1 What are the main drivers of forest loss in Brazil and the Ivory Coast?
10.2. When should a natural forest never be cleared?
10.3 Why should the developed countries consider compensating the poor nations to stop
deforestation?
10.4 Why are non-timber forest products often not taken into account when governments
decide to allocate logging concessions?
10.5 Is there any evidence for an environmental Kuznets curve for forestry?

References and Additional Readings

Barrett, S. (2005). Environment and Statecraft: The Brown, K., and Pearce, D. (eds.) (1994). The
Strategy of Environmental Treaty-making (Oxford: Causes of Tropical Deforestation (London:
Oxford University Press). UCL Press).
FORESTS 211

Ehrhardt-Martinez, K., Crenshaw, E.M., and Jenkins, Peters, C.M., Gentry, A.H., and Mendelsohn, R.O.
J.C. (2002). ‘Deforestation and the environmental (1989). ‘Valuation of an Amazonian rainforest,
Kuznets curve: a cross-national investigation of Nature 339: 655-6.
intervening mechanisms, Social Science Quarterly
Repetto, R. (1988). The Forest for the Trees?
83: 226-43.
Government Policies and the Misuse of Forest
Ehui, S.K., and Hertel, T.W. (1989). ‘Deforestation Resources (Washington, DC: World Resources
and agricultural productivity in the Céte d'Ivoire, Institute).
American Journal ofAgricultural Economics 71:
Rudel, T., and Roper, J. (1997). “The paths to rain
703-11.
forest destruction: cross-national patterns of
Eliasch, J. (2008). Climate Change: Financing Global tropical deforestation, World Development 25:
Forests (The Eliasch Review) (London: HMSO). 53-65.
FAO (Food and Agriculture Organization) (2005). Sandler, T. (1993). “Tropical deforestation: markets
Global Forest Resources Assessment, Main Report and market failures, Land Economics 69(3):
(Rome: UN Food and Agriculture Organization). 225-33.

— (2010). Global Forest Resources Assessment, Scherr, S.J., White, A., and Kaimowitz, D. (2004). A
Main Report (Rome: UN Food and Agriculture New Agenda for Forest Conservation and Poverty
Organization). Alleviation: Making Markets Work for Low-income
GEF (Global Environmental Facility) (2012a) ‘What Producers (Washington, DC: Forest Trends).
is GEF?’, [Link] Schofer, E., and Hironaka, A. (2005). “The effects
(accessed 15 November 2012). of world society on environmental protection
—— (2012b). The Greenline (July 2012), [Link] outcomes, Social Forces 84: 1-25.
[Link]/gef/greenline/july-2012/progress- Shackleton, S., Shackleton, C., and Shanley, P. (eds.)
towards-impact-review-outcomes-phase-i- (2011). Non-timber Forest Products in the Global
amazon-region-protected-areas-pr (accessed Context (Heidelberg: Springer-Verlag).
15 November 2012). : Shandra, J. (2007). “The world polity and
Groombridge, B., and Jenkins, M.D. (2002). World deforestation: a cross-national analysis,
Atlas of Biodiversity: Earth’s Living Resources in International Journal of Comparative Sociology
the 21st Century (Berkeley, CA: University of 48(1): 5-28.
California Press). Sims, K.R.E. (2010). ‘Conservation and
Klemick, H. (2011). ‘Shifting cultivation, forest development: evidence from Thai protected
fallow, and externalities in ecosystem services: areas, Journal of Environmental Economics and
evidence from the Eastern Amazon, Journal of Management 60: 94-114.
Environmental Economics and Management 61: Sunderlin, W.D., Angelsen, A., Belcher, B., Burgers,
95-106. P., Nasi, R., Santoso, L., and Wunder, S. (2005).
McKean, R.N. (1965). “The unseen hand in ‘Livelihoods, forests, and conservation in
government, American Economic Review 55: developing countries: an overview, World
496-506. Development 33: 1383-402.
Mendelsohn, R. (1994). ‘Property rights and tropical UN-REDD (2008). UN collaborative programme
deforestation, Oxford Economic Papers 46: 750-6. on reducing emissions from deforestation and
forest degradation in developing countries
NKMCAP (Noel Kempff Mercado Climate Action
(UN-REDD), [Link]
Project) (2008). Project summary, [Link]
Portals/15/documents/publications/UN-
[Link]/ourinitiatives/urgentissues/global-
REDD_FrameworkDocument.pdf (accessed
warming-climate-change/places-we-protect/
15 November 2012).
noel_kempff_case_study_final-[Link] (accessed
6 June 2011). UN-REDD (2010). About REDD+, [Link]
[Link]/aboutredd/tabid/582/[Link]
Pereira, S.N.C. (2010). ‘Payment for environmental
(accessed 15 November 2012).
services in the Amazon forest: how can
conservation and development be reconciled?’ Walker, R. (1993). ‘Deforestation and economic
Journal of Environment and Development 19: development, Canadian Journal of Regional
171-90. Science 16: 481-97.
The Economics of
Water Pollution
cece cee cere reece reese ree sesencasneseesese bcc ccc mrss c cece eevee sesnsecesesessscessecsssece

In this chapter, we will explore the economics of water pollution. Most of the chapter is
taken up with analysis of the costs and benefits of reducing water pollution.

11.1 Introduction

What is ‘water pollution’? For a natural scientist, water pollution is the discharge of a sub-
stance into a water-body (a river, stream, lake, or estuary, or an underground aquifer),
which changes the functioning of the system. For example, the input of organic wastes to a
river speeds up biological processes, and in the process uses up oxygen. The input of ammo-
nia may be directly toxic to fish. Phosphate inputs to a lake lead to nutrient enrichment of
the aquatic ecosystem, and can result in a build-up of toxic algae and lead to a change in
plant and animal communities.
For an economist, however, we need to know more—precisely, we need to know whether
water pollution adversely affects at least one person’s well-being (e.g. people affected by
sewage levels at their local beach) or adversely affects production (e.g. commercial fishers
suffer reduced catches as a result of an oil spill). This is because pollution only becomes an
external cost, and a source of market failure, if people suffer from its impacts. We classify
water pollution problems into two types, each with different policy implications:

e point-source pollution; and

e non-point pollution.

Point-source pollution is defined as emissions that enter water-bodies from an easy-to-

identify single source, such as a pipe from a factory or the outfall from a sewage works.
These kinds of emissions can be monitored at low cost, and linked to the actions ofindi-
vidual firms or households. Non-point pollution, in contrast, enters water-bodies in a dif-
fuse manner, such as surface run-off from farmers’ fields or from forests, or as pollution
seeping into groundwater aquifers. Non-point pollution emissions are hard to measure,
since they cannot be traced to a single pipe or point of entry to the water-body. It is hard to
THE ECONOMICS OF WATER POLLUTION 213

assign responsibility for the emission of non-point pollutants to individual firms or

households.
Water pollution may originate from many different sources, for example:

e from industry, in the form of point-source discharges of heavy metals, organic wastes,
and other pollutants;
e from sewage treatment works or direct sewage outfalls;
e as acid drainage waters from old mines;
® as leachate from landfill sites;
e as pathogens washing off from fields in which cattle are kept;
® asrun-off of oils and solvents from city streets;
e asrun-off of fertilizers and pesticides, and soil erosion from farmland and forests; and
® as accidental (unplanned) spillages—for instance, from oil tankers.

In what sense is pollution harmful? Pollutants may have their main effect on aquatic
organisms by reducing the dissolved oxygen (DO) content of water. The amount of oxygen
dissolved in water is important for the support of fish. Alternatively, pollutants may be
directly toxic: for example, if pesticides or chlorine are spilt in a river. Pollutants can also
change the acidity of a river or lake, making it impossible for certain organisms to survive
over time. Pollution can change water temperatures, or increase bacterial levels to the detri-
ment of human health. Finally, pollutants can change the nutrient balance of a water-body,
making it over-rich, a process known as eutrophication.
The impact of emissions on water quality may vary in space and time. Consider the dis-
charge of sewage into the sea. The impact is measured in bacterial counts, and will depend
on tide movements, temperature, and sunshine, and the overall direction of water circula-
tion. This means that a given quantity of sewage discharged at two different points on the
same coastline may have differing effects on water quality at a local beach. Another example
is how organic discharges affect local water-quality levels (measured in DO) in an estuary.
Again, the impact of one ton of effluent discharged at one point on the estuary will vary
according to where it is measured (e.g. upstream or downstream). Pollutants whose impact
varies spatially are called non-uniformly mixed. Some water pollutants do not have this
property: for instance, if the concern is with nitrate levels in a lake, it is not important where
in the catchment nitrate originates—cutting any source will have roughly the same impact
on eutrophication. Whether pollutants are uniformly or non-uniformly mixed is important
for the design of policy, since this determines whether our control policies need to discrimi-
nate between polluters according to their location.

11.1.1 Trends over time

In many early-industrializing countries such as the United Kingdom and Germany, most
water-quality indicators have shown improving trends over the past 40 years. This follows
periods of decline associated with earlier industrialization and urbanization. For instance,
the earliest legislation passed in England (the 1876 Rivers Act) followed on from the closure
of Parliament during the ‘Great Stink on the River Thames in London in 1858. Gross
214 APPLYING THE TOOLS

discharges from industrial and municipal sources (sewage works) have now largely been
brought under control, as a result of legislation such as the Control of Pollution Act (1974)
in the UK, the Water Framework Directive of the European Union, and the Water Quality
Act (1965) in the United States. In the UK, for example, the length of rivers classified as
‘grossly polluted’ fell from 2,000 km in England and Wales in 1958 to 800 km in 1980. Since
then, progress has slowed. Many localized difficulties remain, whilst diffuse-source pollu-
tion has risen in significance worldwide (Foley et al., 2005) (see also Box 11.1).
Non-point pollution was the reason why many rivers failed to achieve a ‘Good Ecological
Status’ in England and Wales in 2006, and it is the main cause of water-quality problems in
the USA (Ribaudo, 2009). Table 11.1 shows the different reasons why water-bodies in
Scotland were classified as ‘at risk’ of not achieving Good Ecological Status, from a 2005
audit. The main non-point pressures vary across type of water-body. Pollution of ground-
water aquifers by pesticides, nitrates, chlorinated solvents, and chlorine-rich water from old
mines, and the presence of newer, exotic pollutants such as TBT (tri-butyl tin) and poly-
chlorinated biphenyls (PCBs), has raised relatively recent concerns. In contrast, in many
developing countries, water pollution from ‘traditional’ point sources such as industry and
sewage works is still a significant problem.

11.2 The Costs of Water Pollution Control

Market failures require corrective action (see Chapter 2). Water pollution is a classic exam-
ple of a market failure. What could governments do to reduce water pollution, whether it be

Prior to 1800 and rapid urbanization, Scotland's rivers were clean and healthy. However, by 1850, rivers
such as the Clyde and the Almond were rendered foul by a combination of sewage and industrial waste
and this trend continued into the twentieth century. Significant efforts in restoring Scotland's rivers did
not start until 1965, but since then, progress has been considerable. The improvement has been due toa
combination of causes; principally Scotland's shrinking heavy industrial base and the enforcement of new
legislation.
A comprehensive assessment of trends in water quality since the mid-1970s is provided by SEPA
(2010). Based on a system of ‘harmonized monitoring stations’, they look at trends in a large number of
water-quality parameters. They also consider spatial and seasonal variations in water quality. The main
conclusions reached are that’... improvements in many aspects of water quality have been delivered
through environmental regulation, cleaner technologies, improved sewage treatment and changes in
agricultural practice. This is most notable for parameters such as biochemical oxygen demand,
ammoniacal nitrogen, lead and sulphate, which have all generally declined in river waters. Other
parameters showed a more complex pattern of regional and seasonal trends’ (SEPA, 2010: 7). Total
phosphorus levels (due to a mixture of non-point pollution from farming and point-source pollution
from sewage treatment work, for example), show improving trends in many parts of the country, but
rising levels in northern rivers. The report suggests agricultural intensification as one reason for this trend.
See [Link]
[Link]/.
THE ECONOMICSOF WATER POLLUTION ZAS

Table 11.1 Sources at risk of not achieving Good Ecological Status for different water-body types, Scotland, 2005
a eS ee ee A Ke ee ee
Pressure Rivers Lakes Estuaries Coastal waters Groundwater
Point discharges 14.2 12.3 45 24.7 16.0
Diffuse pollution 24.3 184 45 see 19.8
Abstraction 24.6 SG) es Not applicable 10.4
Physical changes Bes 38.8 40 9.6 Not applicable
Overall percentage of water-bodies at risk 455 54 575 28.7 23.6

Source: SEPA (2005).

from point or non-point sources? Government intervention to reduce pollution can take
three basic forms:

e Regulation. This works by setting technology or performance standards. For water

pollution, a performance standard could be a quantitative limit on how much
pollution firms were allowed to discharge to a river per day, enforced by law.
Technology standards might include the specification of minimum
technological standards for sewage works (e.g. the installation of nutrient
stripping technology).
e Voluntary means. Households could be asked to use low-phosphate laundry products
in a catchment suffering from excess nutrient levels in its streams and rivers; firms
could be asked to achieve improvements in their emissions as their contribution to a
catchment management plan.
e Economic incentives. This could include a tax on emissions of organic wastes to an
estuary; or a tradable permit scheme for such wastes.

The costs of achieving a water-quality standard (such as a minimum concentration of

dissolved oxygen, or a maximum concentration of ammonia) depends on which govern-
ment intervention is used. In Chapter 2, we saw that aggregate reduction in pollution at
least cost depends on the marginal costs of abatement (MACs) being equalized across
sources; only economic incentives can achieve such an outcome (in the absence of an all-
powerful, fully informed regulator who can impose the least-cost pollution on polluters by
regulation). For a target reduction in some measure of water pollution in a lake, achieving
the least-cost outcome means that:

e Marginal abatement costs are equalized between different categories of sources: for
example, between farmers, paper mills, and sewage works, if all are sources of nutrient
pollution.
e Marginal abatement costs are equalized across all firms or households within a
particular source category—that is, for example, across all farmers.

The costs of reaching an aggregate pollution target will exceed the theoretical minimum
if differences exist between and within source categories. Ifa system of performance stand-
ards means that MACs vary significantly across factories all discharging some pollutant,
costs savings will emerge if firms could trade the right to emit—as occurs in a tradable
216 APPLYING THE TOOLS

permit scheme—so to take advantage of these differences. Suppose for example, BK Paper
has marginal abatement costs of €2,000 per tonne at its current level of emissions and Ace
Paper has marginal abatement costs of €5,000 per tonne. If BK cuts its emissions by an
extra 100 tonnes and sells permits to Ace, Ace can increase emissions by 100 tonnes, and
the total costs to the economy of achieving the target fall. A pollution tax can have a similar
effect, as we have seen: both BK and Ace can increase or cut emissions until MAC is equal
to the tax rate for each firm (see Figure 11.1).
One problem with using economic incentives for water pollution is when non-uniform
mixing exists:

e When pollutants are non-uniformly mixed, a single tax rate is inefficient. This happens
because the tax is levied on emissions rather than on their environmental impact
(Muller and Mendelsohn, 2009). Firms that cause more damage per unit of emissions
should be taxed at a higher rate than firms that cause lower per-unit-of-emission
damage. At the limit, correcting this problem involves a unique tax rate for each firm.
More pragmatically, suggestions have been made for ‘banded tax rates’ —where taxes
vary according to the stretch of an estuary in which a firm is located, for instance—to
try to control for non-uniform mixing to at least some degree.
e When pollutants are non-uniformly mixed, allowing permits to trade at a one-for-one
rate may result in local violations of water-quality standards. Imagine two firms are
thinking of trading. In Figure 11.2, firm A is a potential buyer from firm B. But because
A is located upstream of B, each unit of emissions from A does more harm than each
unit from B. If A buys 100 permits from B, total emissions remain constant—but

MAC,

ey ep Emissions

Figure 11.1 The effect of a pollution tax (t) on emissions from two firms, A and B. Note that each firm
reduces emissions up to the point at which its marginal abatement costs (MAC) are equal to the pollution
tax rate.
THE ECONOMICS OF WATER POLLUTION 217

eFirmC
(no change)
VI
VVV

The Estuary

Se ee te
@FirmA eFirmB
buys 100 Hin
On ) 0 10 (sells 100)
—EE——————
|

miles

Figure 11.2 Permit trading in an estuary.

environmental damage rises, especially in the zone immediately downstream of A. This

situation arises for many cases of water pollution control, and several solutions have
been proposed. One is zonal trading, which might involve banning trades between A
and B, and only allowing A to trade with C. But the more trade is restricted in this way,
the lower the cost-saving potential of the scheme. Another idea is to use trading rules,
which would govern the rate at which A and B can trade. Suppose that A’s emissions
are twice as harmful per unit as B’s in terms of average water quality. An exchange rate
of 0.5/1 could be imposed on trading between the two. Under this scheme, exchange
rates would have to be calculated for all firms on the river, but this is possible using
water-quality models. It remains a challenge to establish viable and reasonable trading
ratios for the design of water-quality trading markets (Farrow et al., 2005).

Box 11.2 gives two empirical illustrations of the potential cost savings from introducing
tradable permit systems for water pollution control.

11.3 Non-point Source Water Pollution: A Difficult Problem

to Solve
We have pointed out that non-point pollution is an important source of water-quality prob-
lems in many countries. As point-source pollution is reduced, non-point emissions from
farmland and other types of land cover become more important for governments trying to
drive up water-quality standards. Worldwide, agriculture is now the main source of exces-
sive nitrate and phosphate levels in fresh water and coastal waters (Foley et al., 2005). Like
any form of pollution, non-point source (NPS from now on) pollution is amenable to man-
agement using regulation, voluntary measures, or economic incentives. One challenge for
NPS pollution control is the costs of measuring and monitoring emissions (such as soil
eroding from a hillside, or nutrients flowing into an estuary). An even more difficult task is
to attribute these to the actions of an individual land manager. We use nitrate pollution to
illustrate how the various policy options might work to control NPS pollution. Too high a
nitrate concentration in water-bodies can result in eutrophication, and consequent fish
deaths and loss of amenity. Farming is a major source of nitrates; artificial fertilizers and
livestock waste are therefore potentially polluting inputs. If a stable relationship could be
218

Little empirical evidence exists on the ability of tradable permit systems to deliver actual cost savings with
respect to water pollution control, since they have not be used that much (in stark contrast to their use in
controlling air pollution). The Fox River, Wisconsin trading system was the subject of initial studies that
suggested large cost savings from permit trading (O'Neil et al., 1983). In practice, however, the trading
scheme was hamstrung by regulations that only one trade ever occurred. Most ‘evidence’ for cost savings
from tradable pollution permits (TPPs) for water pollution control comes from simulation studies. In this
box, we give brief details on two such studies in two European countries.
Scotland: the Forth Estuary. The Forth Estuary, in central Scotland, is a tidal water-body that is subject to
many demands, including providing water for industrial cooling, for recreation, as a habitat for birds and
as a sink for waste disposal. Most wastes come from industry, notably from a large petrochemical
complex and from a yeast factory. A seasonal ‘sag’ in dissolved oxygen in the upper estuary due to
excessive pollution has been noted in many summers: this has a bad effect on salmon migrating
upstream. Control is exercised by the Scottish Environment Protection Agency, which uses performance
standards to regulate discharges of pollution from firms and sewage works. Hanley et al. (1998) report
results from a simulation exercise to study the potential cost savings from introducing a TPP system to
improve DO levels. They found that such a system could generate large cost savings over regulation,
although these were reduced once uncertainty over water-quality impacts was allowed for. For example,
a TPP system could achieve a 20 per cent improvement in DO in the most polluted part of the estuary at
one-ninth of the cost of uniform regulation. This large saving occurs because marginal abatement costs
vary greatly over firms at the current level of control. Under uncertainty, the TPP system generates higher
costs, but still achieves the target (in probabilistic terms) at a much lower cost than standards. These
results were obtained by combining an economic model of polluters, based on abatement costs, with a
water-quality model that allows for firms located in different parts of the estuary to have different
impacts on DO per unit of emission.
Coastal water quality in Sweden: the pulp and paper industry. Paper and pulp mills have traditionally
been seen as major sources of water pollution in many countries. In Sweden, more than 50 per cent of
total discharges of oxygen-depleting pollutants, and almost all discharges of chlorine-containing material
come from this industry. Most plants are located on Sweden’s east coast, so a lot of these discharges end
up in the sea. Discharges are regulated by non-tradable permits, set at the firm-specific level, as in the
Scottish case. Runar Brannlund and colleagues studied the likely impact on abatement costs of allowing
firms to trade these permits. Their results are based on a study of41 pulp mills over the period from 1986
to 1990 (Brannlund et al., 1998). They find that moving from no trading to trading increases industry-level
profits by around SEK1.2 billion in 1989, a 6 per cent rise, since firms with higher abatement costs can buy
permits from firms with lower abatement costs, who can profit from such sales. Some thirty-two firms
become permit sellers, and eighteen firms become buyers. The authors note that the permit system they
model could result in increased environmental damages, since although total emissions are the same in
the trading and no-trading cases, the impacts of these emissions may vary, since firms have different
environmental impacts per unit of emission due to their physical locations (i.e. this is a problem of a
non-uniform mixing pollutant).

found between inputs of nitrate fertilizer and nitrate levels in a polluted lake, control could
be exercised on this use of input. This means that we need information about the ‘pollution
production function’ (g), which relates inputs of fertilizer, N,, from source i to water-quality
levels at a range of monitoring points, Q;

Q, = g(N,,Z). (11.1)
THE ECONOMICS OF WATER POLLUTION 219

Here, Z represents all the other observable factors that determine nitrate concentrations
in a lake at a given point in time, such as rainfall, and the type ofcrops being grown. Imagine
that the state of knowledge is such that for a particular lake, we can estimate equation (11.1)
using water-quality modelling.
The policy alternatives are the same as with point-source pollutants: a tax could be
placed on N, or a tradable permits system set up for purchases/applications of N.
Alternatively, farmers could be regulated in terms of how much N they are allowed to
apply. In Figure 11.3(a), we show the impacts of a tax on nitrate fertilizer. The initial price
is P,, at which price farmers maximize profits by applying N, units, since at this point the
marginal cost of N (its price) is equal to its marginal benefits, measured by the demand
curve N°. Introducing a tax of t raises the price to (P,, + t), and nitrate applications fall to

(a)

x
£
& pttt--------------4
o [Pha beau alee
en Beenie LS
1
1
1
1
i]
1]
1]
!
i)
i]

No Ny
Quantity of nitrate fertilizer, N
(b)

oO
2
we
—
inn

2
lac}
=

Nitrate inputs

Figure 11.3 Using input taxes to reduce non-point pollution.

220 APPLYING THE TOOLS

Np. This reduction in fertilizer use is shown in Figure 11.3(b) as resulting in a reduction in
nitrate inputs to the river from N, to No, [Link] quality from Q, to Q,.
The demand curve N° can be interpreted as the farmer’s marginal abatement cost curve
for nitrogen use, since each unit less the farmer uses costs him in terms of forgone output
(as yields fall). The fall in output will vary across farms, due, for instance, to variable land
productivity, varying managerial skills, and variable climate. This means that the MAC
curve for nitrate use varies across farms, just as with point-source pollution. Given such
differences in MAC curves, an economic incentive approach such as a tax should give a
desired reduction in emissions at a lower aggregate cost than regulation. In this instance,
regulation would involve constraining all farms to apply no more than N, nitrogen to their
land. But if the N curve varies across farms, then this uniform regulation will be more
costly (less efficient) than a tax. We can also see that under the nitrates tax, the farmer incurs
a financial burden that exceeds the value of lost output, since the farmer must also pay the
tax on his fertilizer use. Alternatives to a tax on the input N are: (i) a tax on predicted emis-
sions; (ii) a tax/subsidy scheme based on ambient levels of nitrate in the lake, known as a
Segerson tax; and (iii) a tax on the outputs that the nitrate is used to produce. For a discus-
sion of (i) and (ii), see Shortle and Horan (2001).
This seems straightforward in theory, but many problems exist in practice. The impact
of each unit of nitrogen applied on nitrate levels in the lake will vary both within a given
farm (according, for example, to the time of year it is applied, the crop growing in the
fields, or the slope ofthe fields) and across farmers. The land management regime adopted
may matter more than the amount of fertilizer applied. This means the function Q = g(.)
will be complicated. Nitrate pollution is also non-uniformly mixed. This implies for many
areas of the country there is no problem: nitrate levels in many rivers and groundwater
fall below levels that cause trouble, for many reasons. In other rivers, ambient nitrate
levels exceed targets during wet periods (e.g. winter) when run-offis greater, but not dur-
ing dry periods. This suggests that a tax on nitrogen is too crude a policy measure to
achieve the specific, local objectives that typify many non-point pollution problems. We
need to look at other options, and see whether economic incentive properties can some-
how be maintained.
One option is to provide financial incentives for management practices that reduce
nitrate pollution. These include avoiding leaving land bare in winter, and avoiding nitrate
applications or livestock densities that are excessive. Farmers could be offered subsidies to
sign up voluntarily to management practices designed to reduce pollution in specific areas
of the country. This might yield cost savings over a system in which all farmers must con-
form to the same management restrictions, since those farmers who face the lowest abate-
ment costs are the most likely to sign under the voluntary scheme. Similar subsidy systems
are used for the control of soil erosion from farmland in the USA under the Conservation
Reserve Program. Studies suggest such management-incentive-based approaches are more
cost-effective than enforcing uniform management standards (see Shortle and Horan, 2001;
Shortle et al., 2012). Land use itself could be targeted for control, again either using eco-
nomic incentives or regulation.
Finally, ‘mixed-instrument’ approaches, which combine some elements of input taxes,
management restrictions, and subsidies for desirable land-use change, could out-perform
either a system based on either economic incentives only or regulation only. Aftab et al.
THE ECONOMICS OF WATER POLLUTION 22)

(2007) provide one illustration. They model policy options to reduce nitrate pollution in a
catchment in eastern Scotland. A model of different farm types in the catchment was con-
structed, and linked to a water-quality model that related land use to nitrate levels in the
river. The authors considered the following range of policy options:
e A tax on estimated nitrate run-off (estimated emissions)
e A taxon nitrate inputs
e Input quotas per hectare
e Limits on livestock density
® Set-aside requirements (a requirement to leave a certain fraction of farmland
uncultivated)

They considered mixes of incentives to capture how regulators use more than one policy
lever to achieve their objectives. The model was run to estimate the costs of hitting a target
maximum concentration of nitrates (specified as how many weeks per year the target was
allowed to be exceeded). Since nitrate pollution also depends on the weather, two scenarios
were considered: an average and a wet year. Policy options were ranked by costs, and the
results are shown in Table 11.2.
The pure economic instrument of an estimated emissions tax is always lowest cost.
However, it is dificult and controversial to implement such a tax in practice, since farmers
may object to the levels of emissions that were being attributed to them. A pure tax on
inputs is relatively low cost, but a mix of instruments, such as combining stocking density
restrictions with an input tax, can be better. In a wet year, the ranking changes slightly.
Measures effective in reducing nitrate run-off in wet conditions (e.g. setting land aside from
cultivation) are now relatively more attractive. These results suggest that regulators have
good grounds to combine an economic instrument with more regulatory approaches for
NPS pollution.
Another policy option for NPS pollution is point-non-point trading. In some water-
bodies, nutrient pollution originates from both point sources such as sewage treatment
works or factories, and from farmland. Allowing permit trading between point sources
saves costs. Finding an efficient balance of emission reduction activities between point and

Table 11.2 Ranking of options to achieve a water-quality target according to cost

Rank during an average year Rank during wet years

Estimated emissions tax | ]

Input tax 3 4
Input quota 4 5
SD restriction + input tax 2 8
Set aside + input tax 6 3
SD restriction, set aside + input tax 5 2
Set aside + SD restriction 7 7
Set aside 8 6

Note: SD = stocking density.

Source: Adapted from Aftab et al. (2007).
222

non-point sources requires that we either set a pollution tax that applies to both sectors, or
we allow them to trade emission reduction rights. A local authority might find it cheaper to
pay a group offarmers to reduce expected emissions from their land by, for example, switch-
ing to low input cropping, rather than investing in nutrient removal facilities at its treat-
ment plants.
The scope for trading emission reduction rights across point and non-point sources
depends on the extent ofdifferences in marginal abatement costs between them. In princi-
ple, point-non-point trading of emission reductions could deliver pollution reduction tar-
gets at lower costs than in the absence of such a system. Box 11.3 gives an example of such
systems. Several schemes that allow point-non-point trades are in operation in the USA,
Canada, and New Zealand. Their success has been rather limited, due to several factors.
These include regulatory uncertainty about how much emission reduction will be forth-
coming from a particular set of actions by farmers; and that emissions from farmland
depend on the weather and vary by season. It is a challenge to know the correct exchange
rates between point and non-point sources. For a full discussion of the pros and cons of
designing and implementing point-non-point trading schemes, see Shortle and Horan
(2001) and OECD (2010). For a recent simulation of the winners and losers under a point-
non-point trading scheme, see Lankowski et al. (2008).

A review of water-quality trading schemes related to agricultural pollutants around the world is given in
OECD (2010). Some twenty-six programmes worldwide where trading could occur were identified,
although trading had not actually occurred in all of these. Most schemes were found in the USA, but
similar policies also exist in Australia, Canada, and New Zealand, whilst trading programmes are being
considered in Finland and Sweden.
A point-non-point trading programme for phosphorus management for the South Nation River in
Ontario, Canada, is described. This allows fifteen municipalities and two industrial point sources of
phosphorus (P) to offset their requirements for pollution reduction by funding projects to reduce
emissions from agricultural sources. A trading rule requires an estimated 4 kg of P from agriculture to be
cut to offset 1 kg of P from point sources. The programme is thought to result in cost savings, since
reductions from agriculture, even with this trading rule, are cheaper than reductions from point sources
via enhanced treatment. All point sources covered by the scheme have chosen to use these agricultural
offsets, although there is no direct bargaining between farmers and point sources: trades are affected
through an agency that contracts with farmers to engage in emissions reduction programmes, and then
sells the offsets to point sources. Direct trading between point sources for pollution credits does,
however, occur in the Hunter River Salinity Trading Scheme in Australia, via an online trading platform.
Other examples of non-point trading schemes in the USA include the California Grassland Areas
Program (where trades are in reductions in selenium discharges from farmland), and the Greater Miami
Watershed point-non-point trading scheme, established in Ohio in 2005. The California scheme is
interesting since rather than being based on estimated emissions, it is based on actual selenium levels,
monitoring of which is made possible due to the nature of the irrigation system within which the scheme
works. The Ohio scheme features trading of nutrient reduction credits for nitrates and phosphorus
between farmers and regulated point sources in the catchment.
THE ECONOMICS OF WATER POLLUTION 223

11.4 Measuring Water-quality Benefits

Water-quality benefits can arise from the amelioration of many sources of impairment,
including:

e non-point pollution from farmland and forests;

e changes to river morphology;
e hydroelectricity production;
® storm water overflows;
® point-source pollution from industry and sewage treatment works; and
e habitat loss and conversion.

Benefits arise from improvements to river appearance/aesthetics and flows; improve-

ments in bankside vegetation; improvements to in-stream ecology; increases in biodiversity
(water birds, otters, amphibians); and reductions in bad odours and health risks (e.g. from
cyanobacteria: Hunter et al., 2012). These physical changes in water-bodies generate a range
of economic benefits, involving both increased use values—for example, to fishers or kayak-
ers Johnstone and Markandya, 2006; Hynes et al., 2009), or to those living near a polluted
water-body—and increased non-use values; for example, because of an improvement in the
survival probabilities of rare species. Many benefits will not be directly reflected in market
prices, so that non-market valuation methods must be used to estimate them.
Many methods have been used to measure the economic value of water-quality improve-
ments, dating back to early work by Smith and Desvouges (1986). A recent overview of
methods that can be applied within the context of the European Union’s Water Framework
Directive can be found in Birol et al. (2006). Consider three methods—contingent valua-
tion, choice experiments, and revealed-preference approaches.
Starting with work in the 1980s, the contingent valuation method has been a popular tool
to value changes in water quality (e.g. Mitchell and Carson, 1989). Loomis et al. (2000), for
instance, use a contingent valuation survey to measure the benefits of restoring ecosystem
service flows on the Platte River; Holmes et al. (2004) studied the benefits of ecosystem ser-
vice restoration for the Little Tennessee River. Both studies were used as part of a cost-
benefit analysis (CBA) of river restoration, showing in both cases that benefits outweighed
costs (see also Box 11.4). In the UK, contingent valuation has been used to estimate the
benefits of improvements to flow conditions in rivers subject to summer low flows (e.g.
Hanley et al., 2003).
The second method is choice experiments. This method has proven itself attractive to
water-quality researchers, since it enables separate values for different attributes of water
quality to be assessed (e.g. in-stream ecology, aesthetics, bankside vegetation). Such attrib-
ute values are seen as being useful from a management or policy context. Policy-makers
like the detailed picture ofthe relative benefits of water-quality management options to be
put together. Examples include Hanley et al. (2006), who looked at the value of improve-
ments to two rivers in eastern Scotland, and Smyth et al. (2009), who study the benefits of
224

Viscusi et al. (2008) report the results of anational stated preference survey aimed at measuring the
economic value of changes in US water quality over the period from 1994 to 2000. How to quantify
changes in water quality is one problem facing economists wishing to place monetary estimates on these
changes. Viscusi et al. use the US Environmental Protection Agency's classification of water quality, based
on what uses can be made of a particular water-body; for example, whether it is clean enough to swim in,
and whether fish caught there can be safely eaten, or whether it is only suitable for boating. In their stated
preference survey, they use two classes of water: ‘Good’ and ‘Not Good’. Good quality was described as ‘a
lake or river suitable for all uses’. ‘Not Good’ meant ‘a lake or river is an unsafe place to swim due to
pollution, has fish that are unsafe to eat, or supports only a small number of fish, plants and other aquatic
life’. Between 1994 and 2000, national water quality in the USA, defined on this basis, declined: there was
a 6.2 per cent decline in the area of lakes and miles of river classified as ‘Good’.
Some 4,527 members of the US public were questioned, using an Internet panel. Respondents were
asked to make choices between moving to one of two possible new regions. Each pair of regions differed
in terms of the annual cost of living and the percentage of lake acres and river miles that were of ‘Good’
quality. By observing people's preferences in repeated choice situations, the researchers could infer their
willingness to pay (WTP) for a 1 per cent increase in the percentage of waters near their (hypothetical)
residence that were of Good quality. On the basis of this, they estimated a mean figure of $196 per
household per year on the economic cost of the water-quality decline from 1994 to 2000; this is
equivalent to around $22 billion across all US households. In a benefit-cost analysis, this figure could be
compared with the costs of restoring this length of degraded lakes and rivers to ‘Good’ status.

management actions for Lake Champlain. Attributes used in the Smyth et al. study include
water clarity, beach closures, the spread of an invasive plant, and fish consumption adviso-
ries. An example ofachoice experiment task, for the Hanley et al. (2006) study, is included
as Table 11.3. Figure 11.4 shows in more detail how the varying levels of water-quality
attributes can be portrayed to respondents. Table 11.4 shows typical results in terms of
marginal values for improvements in river attributes, again from the Hanley et al. (2006)
study. Box 11.5 reports ona study using both choice experiments and contingent valuation
to study benefits of water-quality improvements in Iowa.

Table 11.3 An example ofa choice card

Do nothing

Option impact
Number of agricultural jobs lost or gained No loss, no creation Loss of five jobs Creation of two jobs
1 the local area
i = 5 = > -
Visual impact: number of months of low » Months 2 months 3 months
ow conditions in the river

Ecological condition of river Worsening Slightimprovement Big improvement

3 d
icrease In water rates per year EQ Ed £40

Please tick the option you prefer O L O

eeee ae ee
Source: Hanley etal, (2006),
THE ECONOMICS OF WATER POLLUTION Zoo

River life: fish, insects, plants

Poor ‘|
te
Moderate Good

&
ot a
era ra

Low number and variety Reduced number and High number and variety
offish, insects, and variety offish, insects, and of fish, insects, and plants:
plants: plants:

Coarse fish Coarse fish present, Salmon, trout

salmon and trout at risk as well as coarse fish
Tolerant species (water Sensitive species (lamprey, Sensitive species (lamprey,
hog louse and weed) crayfish, mayflies, native crayfish, mayflies, native
common plants) occasionally present plants) present

Figure 11.4 Explaining varying levels of an attribute.

Source: Stithou (2011).

A third type of valuation procedure that has been applied to water-quality improvements
involves revealed-preference methods. These methods are used to measure changes in rec-
reational use values. An example is the work reported by Johnstone and Markandya (2006),
who relate anglers’ choice of total fishing trips in a year, and where those trips were made,

Table 11.4 Example outputs from a choice experiment for two rivers in Scotland. The units are pounds sterling
per household per year. Bid vehicle: local water taxes. Figures in parentheses are 95% confidence intervals

Motray River Brothock River Pooled data

Local farm jobs Shey 3168 3.65

(2.38; 4.66) (2.41; 4.98) (2.81; 4.48)
River flow conditions, per month improvement 3.87 2.70 3.00
(2.52; 5.07) (0.90; 4.21) (1.74; 4.25)
Ecology, slight improvement 8.97 10.53 9.45
(5.41; 12.38) (4.57; 17.19) (6.25; 12.93)
Ecology, big improvement 24.03 28.26 2519
(18.53; 31.08) (19.65; 40.57) (21.10; 31.74)

Source: Hanley et al. (2006).

226

Water-quality Improvements

Contingent valuation and choice experiments have both been used to measure the benefits of water-
quality improvements. But do we obtain the same answers for a given improvement, irrespective of
which method of measurement we use? Christie and Azevedo (2009) investigate this issue, looking at the
specific case of improvements to water quality in Clear Lake, lowa. Clear Lake is the third-largest natural
lake in lowa, and is intensively used for recreation. However, non-point phosphate pollution from
agriculture has resulted in a decline in water clarity, an increase in algal blooms, and a fall in fish diversity.
A contingent valuation study was administered to 900 lowa residents. Three scenarios were used, each
describing a different change in water quality. Plan A involved a programme of measures that would
prevent further declines in water quality. Plan B would produce a ‘moderate’ improvement, and plan Ca
‘substantial’ improvement in water quality. The water quality was described in terms of water colour and
clarity, the number of algal blooms per year, water odour, and fish populations. These were also the
attributes used in the choice experiment, which was administered to 600 people. Respondents were told
that such improvements could only come at the expense of a rise in their taxes.
The following table shows the WTP estimates from the two methods:

Contingent valuation Choice experiment

($ per household) ($ per household)
Plan A—further declines -1,327 -2 341
Plan B—a moderate improvement 1,093 385
Plan C—a substantial improvement 1,642 2,692

As. can be seen, both methods show a welfare loss (negative WTP) for plan A, which the authors interpret
as how much worse off people would be if water quality declined below the current level. Both methods
give a welfare gain (positive WTP) for improvements in water quality, which are bigger for the larger
improvement. However, the two measures give significantly different mean WTP values for any given
change. Christie and Azevedo then test whether the preferences underlying these responses differ
between the choice experiment and contingent valuation. They do this by combining the three contingent
valuation scenarios into one data set. They find that there are no statistically significant differences
between the marginal values that people place on the four attributes used in the design (water colour and
clarity, number of algal blooms per year, water odour, and fish populations), which they see as
encouraging in terms of the ‘convergent validity’ between these two stated preference methods.

to water-quality parameters such as the number of taxa within the river, organic pollution
levels, habitat quality, and the number of fish species. Separate models were estimated for
lowland, upland, and chalk streams in England, based on a survey of anglers. They found
that both the number of angling trips and the distribution of these trips across sites were
significantly related to most of the water-quality measures used, although parameter signs
are sometimes unexpected. They then used the combined participation and site choice
model to measure welfare benefits (changes in consumer surplus per trip) for a 10 per cent
improvement in water-quality measures. Table 11.5 shows some of their results. However,
one problem with applying travel-cost models to valuing water-quality improvements is
that researchers find a high degree of multi-collinearity between measures of water quality,
os

Table 11.5 Economic benefits to anglers in England from water-quality improvements, as

given by change in consumers’ surplus per trip in pounds sterling for a 10 per cent improve-
ment in statistically significant river attributes

Upland rivers Lowland rivers Chalk rivers

Number of fish species 2.49
BOD -0.43
Ammonia =ONle
DO 2.09 0.29
Flow Wey 3.70 OS

Note: The mean consumers’ surplus across all three rivers under current conditions was £25 per trip.
Blank cells imply that the water-quality parameter was not significant in the choice model at 95%.
Source: Adapted from Johnstone and Markandya (2006).

making it difficult to identify individual effects. This was an issue in the Johnstone and
Markandya study.
Hedonic pricing (HP) is a revealed-preference approach that relates variations in water
quality to variations in house prices, as a way of measuring aspects of the benefits of
improvements in quality. Leggett and Bockstael (2000) look at the effects of varying faecal
coliform levels in coastal waters on property prices in Anne Arundel County, Maryland.
The irregularity of this coastline means that water-quality levels vary substantially within
the housing market. The analysis was based on house sales data of waterfront properties
from 1993 to 1997. The authors argue that faecal coliform counts are a good measure of
water quality to use in HP studies, since it is something that people are likely to care and
know about, especially if they engage in water-based recreation such as swimming and
boating, due to the health risks. High levels of faecal matter also make the water smell and
look bad, and pose health risks to users. Leggett and Bockstael explore different functional
forms for the hedonic price equation, including linear, double-logarithmic, semi-loga-
rithmic, and inverse semi-logarithmic. With the exception of the linear form, the measure
of faecal coliform concentration is significant and negative as an explanatory variable.
Welfare changes from reducing faecal coliform pollution are then estimated. Taking one
polluted stretch of coastline, where the levels range from 135 to 240 coliforms per 100 ml
of water, the authors find that property values would rise by $230,000 if levels were cut to
100 coliforms per 100 ml, a 2 per cent gain in value (based on the inverse semi-log hedonic
price equation).

11.5 Problems for Cost-benefit Analysis of Water-quality

Improvements

Two broad classes of problem exist when using cost-benefit analysis for water-quality issues:

1. Identifying the water-quality change to be valued.

2. Valuing the water-quality change, once identified.
228 APPLYING THE TOOLS

11.5.1 Identifying the water-quality change to be valued

Economists start by assuming that the change in water quality or quantity is well estab-
lished. But this is a strong presumption. Welfare economics operates on the assumption
that the ‘initial’ and ‘subsequent’ amounts of a public good are known, so that willingness to
pay can be measured. But two issues exist in applications to water-quality management, for
instance, in terms of assessment of a new policy to reduce water pollution. First,a CBA must
identify what changes would have taken place over time without this measure in place—the
‘business as usual’ scenario. This can be tricky, since many factors determine future water
quality. For example, the closure of factories due to an economic downturn could lead to
water-quality improvements independently of a clean-up programme, as could the require-
ments of other pieces of legislation. Whilst households might not care why their river has
become cleaner, the application of CBA to a programme of measures requires the analyst to
separate cause and effect.
Second, the impacts of a programme of measures on water quality will be uncertain.
The reduction of dairy herds should improve water quality in a catchment, but these pre-
dictions are made with error bounds implicitly or explicitly attached. Should respondents
in a WTP study be made aware of this uncertainty over the water-quality outcomes of,
say, a tax on fertilizer use, or the installation of buffer strips alongside watercourses?
Finally, researchers may be unsure how long an aquatic ecosystem will take to recover
from some pollution episode and subsequent remediation (e.g. liming of acidic lakes:
Banzhaf et al., 2006). This will affect the time profile of benefits that form part of the CBA,
and that constitute aspects of the benefits scenario within which respondents are asked to
state their preferences. How should this uncertainty be conveyed to those who might
benefit?

11.5.2 Problems in pricing the environmental change, once identified

Three problems will be reviewed here. First, the way in which water quality is measured by
scientists does not describe how it is perceived by ordinary people, either in terms of what it
is important to measure, or whether actual quality (in terms of health risks, say) corresponds
to perceived quality. Yet it is perceived quality that determines behaviour and helps deter-
mine WTP (Adamowicz et al., 2003). This mismatch between how water quality is measured
by scientists and regulators and how it is perceived by ordinary people creates problems for
the application of CBA to water policy and water management.
Second, aggregate benefits ina CBA depend both on per-person or per-trip measures of
water-quality improvements, and the number of people or number oftrips that are impacted
by this change. Economists tend to focus on the per-person or per-trip measurement prob-
lem, rather than the ‘how many people benefit?’ problem. One solution to this problem in
the water-quality context is the estimation of distance-decay functions, which try to find a
spatial relationship between who is willing to pay for an improvement and how far away
from the river and so forth they live (Bateman et al., 2006): Figure 11.5 shows examples of
two such functions for water-quality improvements to the River Mimram, England and for
the Manchester Ship Canal in England.
THE ECONOMICS OF WATER POLLUTION 229

(i) River Mimram

(£)
WTP

(0). ee T SS a Sa ipa raat

0 2 4 6 8 10 12 14 16 18 20
Distance from Mimram (km)

(ii) Manchester Ship Canal

20 74

a 15
a
= 10

5
—@— Average WTP
0 Gle=7enmeeh at laiwews PS Beds a alee T T 1

Lessthan Between Between’ Between Between Between

1 mile Jand2 2and5 5and10 10and 20 and
miles miles miles 20 miles 30 miles

Figure 11.5 Examples of distance-decay functions for English water-bodies.

Sources: Hanley et al. (2003); Hanley and Colombo (2008).

Summary
The economics of water pollution offers an interesting contrast between policy and economic
theory. Despite the arguments of economists, governments have been reluctant to adopt eco-
nomic instruments rather than regulation to reduce water pollution. Examples of water pollu-
tion taxes exist, but on the whole, economic instruments have gained less ‘traction’ in water
pollution control than in air pollution control. This might be because of the difficulties of
applying taxes or tradable permits to non-point sources of water pollution, and because of the
way in which even point-source emissions map on to ambient quality levels is complex and
locale-specific. In contrast, the economic ideas of cost-benefit analysis and benefits assess-
ment have gained relatively more prominence in both project and policy appraisal for water
pollution; for example, in the context of the EU Water Framework Directive (see Box 1 1.6).
230 APPLYING THE TOOLS

Improvements?

Water-quality improvements impose costs on sectors responsible for water pollution. Depending on the
market structure, a proportion of these costs will fall on consumers, Those paying the costs of clean-up
are likely to differ from those benefitting from better water quality. A nice illustration of
this concerns a
study by Fezzi et al. (2008), on the costs and benefits of improving water quality in rivers in Yorkshire. A
major source of nitrate pollution in this area is agriculture: measures to reduce nitrate pollution impose
costs on farmers in terms of lost profits. Benefits accrue to those who value water-quality improvements,
including recreational users of the catchment. Using a combination of Geographic Information Systems
and a range of environmental valuation methods, Fezzi et al. find that most benefits accrue to urban
households in the area. For an improvement in water quality required under the Water Framework
Directive, they find that benefits are in the order of £12.5 million, whilst the costs to farmers are in the
order of £5.5 million. So whilst (urban) households gain from the water-quality improvement, the costs
are felt by farmers as lost income due to their forgoing more profitable cropping activities. This study is a
good example ofthe use of integrated modelling to measure benefits and costs of water pollution
reductions, and on whom these benefits and costs fall.

Tutorial Questions
11.1 What are the main types and sources of water pollution? Why might the source of
the problem matter for policy design?
11.2 What problems would a regulator face in using a system ofpollution taxes to
improve water quality in a river, where the main problem is point-source discharges
from factories and sewage treatment works? What would be the main potential
benefits of using such a tax approach, relative to regulation?
11.3 How could economic instruments be used to reduce pollution from non-point
sources in an estuary? In what way does the nature of the pollution source impose
additional policy design problems for the government?
11.4 How could you use (i) choice experiments and (ii) hedonic pricing to estimate the
benefits of an improvement in river-water quality in an urban area?
11.5 What are the problems of undertaking a cost-benefit analysis of a planned
improvement in coastal water quality through improvements in sewage treatment?

References and Additional Readings

Adamowicz, W., Swait, J., Boxall, P., Louviere, J., Aftab, A., Hanley, N., and Kampas, A. (2007). ‘Co-
and Williams, M. (2003). ‘Perceptions versus ordinated environmental regulation: controlling
objective measures of environmental quality in non-point nitrate pollution while maintaining
combined revealed and stated preference models river flows, Environmental and Resource
of environmental valuation, in N. Hanley, D. Economics 38(4): 573-93.
Shaw, and R. Wright (eds.),
; The New Economics of
Banzhaf, H.S., Burtraw, D., Evans, D., and
Outdoor Recreation (Cheltenham: Edward Elgar).
Krupnick, A. (2006). “Valuation of natural
THE ECONOMICS OF WATER POLLUTION 231

resource improvements in the Adirondacks) Land improvements: distance—decay functions for use
Economics 82(3): 445-64. and non-use values, Journal of Environmental
Bateman, I., Day, B., Georgiou, S., and Lake, I. Management 68: 297-304.
(2006). “The aggregation of environmental benefit Holmes, T-P., Bergstrom, J.C., Huszar, E., Kask, S.B.,
values: welfare measures, distance decay and total and Orr, F. II (2004). ‘Contingent valuation,
WTP? Ecological Economics 60: 450-60. net marginal benefits, and the scale of riparian
Birol, E., Karousakis, K., and Koundouri, P. (2006). ecosystem restoration, Ecological Economics 49:
‘Using economic valuation techniques to inform 19-30.
water resources management: a survey and Hunter, P., Hanley, N., Czajkowski, M., Mearns,
critical appraisal’ Science ofthe Total Environment K., Tyler, A.N., Carvalho, L., and Codd, G.A.
365: 105-22. (2012). “The effect of risk perception on public
Brannlund, R., Chung, Y., Fare, R., and Grosskopf, S. preferences and willingness-to-pay for reductions
(1998). ‘Emissions trading and profitability: the in the health risks posed by toxic cyanobacterial
Swedish pulp and paper industry, Environmental blooms; Science of the Total Environment 426:
and Resource Economics 12: 345-56. 32-44.

Christie, M., and Azevedo, C. (2009). “Testing Hynes, S., Hanley, N., and O'Donoghue, C. (2009).
the consistency between standards contingent ‘Alternative treatments of the cost of time in
valuation, repeated contingent valuation and recreational demand models: an application
choice experiments, Journal of Agricultural to whitewater kayaking in Ireland? Journal of
Environmental Management 90(2): 1014-21.
Economics 60(1): 154-70.
Farrow, R.S., Schultz, M., Celikkol, P., and Van
Johnstone, C., and Markandya, A. (2006). “Valuing
river characteristics using combined site choice
Houtven, G. (2005). ‘Pollution trading in water
and participation travel cost models; Journal of
quality limited areas: use of benefits assessment
Environmental Management 80: 237-47.
and cost effective trading ratios, Land Economics
81(2): 191-205. Lankowski, J., Lichtenberg, E., and Ollikainen, M.
(2008). ‘Point/nonpoint trading with spatial
Fezzi, C., Rigby, D., Bateman, I.J., Hadley, D.,
heterogeneity, American Journal of Agricultural
and Posen, P. (2008). “Estimating the range of
Economics 90(4): 1044-58.
economic impacts on farms of nutrient leaching
reduction policies, Agricultural Economics 39: Leggett, C.G., and Bockstael, N. (2000). “Evidence
197-205. on the effects of water quality on residential land
prices, Journal of Environmental Economics and
Foley, J.A., DeFries, R., Asner, G.P., Barford, C.,
Management 39: 121-44.
Bonan, G., Carpenter, S.R., Chapin, FS., Coe,
M.T., Daily, G.C., Gibbs, H.K., Helkowski, Loomis, J., Kent, P., Strange, L., Fausch, K., and
J.H., Holloway, T., Howard, E.A., Kucharik, Covich, A. (2000). ‘Measuring the total economic
C.J., Monfreda, C., Patz, J.A., Prentice, I.C., value of restoring ecosystem services in an
Ramankutty, N., and Snyder, P.K. (2005). ‘Global impaired river basin: results from a contingent
consequences ofland use} Science 309: 570-4. valuation survey, Ecological Economics 33:
103-17.
Hanley, N., and Colombo, S. (2008) ‘Benefits of
potential water quality improvements on the Mitchell, R., and Carson, R. (1989). Using Surveys
Manchester Ship Canal, Unpublished report, to Value Public Goods: The Contingent Valuation
Economics Division, University of Stirling. Method (Washington, DC: Resources for the
Future).
—, Colombo, S., Tinch, D., Black, A., and Aftab, A.
Muller, N.Z., and Mendelsohn, R.O. (2009). ‘Efficient
(2006). ‘Estimating the benefits of water quality
improvements under the Water Framework pollution control: getting the prices right,
American Economic Review 99(5): 1714-39.
Directive: are benefits transferable?’ European
Review ofAgricultural Economics 33: 391-413. OECD (2010) Water Quality Trading in Agriculture
(Paris: OECD).
—., Faichney, R., Munro, A., and Shortle, J. (1998).
‘Economic and environmental modelling of O’Neil, W., David, M., Moore, C., and Joeres, E.
pollution control in an estuary; Journal of (1983). ‘Transferable discharge permits and
Environmental Management 52: 211-25. economic efficiency: the Fox River} Journal of
Environmental Economics and Management 10:
—, Schlapfer, E, and Spurgeon, J. (2003).
346-55.
‘Aggregating the benefits of environmental
232 APPLYING THE TOOLS

Ribaudo, M. (2009). ‘Non-point pollution regulation constrained environment, Environmental Science

approaches in the U.S’ In J. Albiac and A. Dinar and Technology 46(3): 1316-25.
(eds.), The Management of Water Quality and Smith, V.K., and Desvouges, W. (1986). Measuring
Irrigation Techniques (London: Earthscan): Water Quality Benefits (Boston, MA: Kluwer
83-102. Nijhoff).
SEPA (Scottish Environment Protection Smyth, R.L., Watzin, M.C., and Manning, R.E.
Agency) (2005). Scotland River Basin District: (2009). ‘Investigating public preferences for
Characterisation and Impacts Analyses Required managing Lake Champlain using a choice
by Article 5 of the Water Framework Directive— experiment; Journal of Environmental
Summary Report (Stirling: SEPA). Management 90: 615-23.
— (2010). Trends in Scottish River Water Quality Stithou, M. (2011). “The economic value of
(Stirling: SEPA). improvements in the ecology ofIrish rivers due
Shortle, J.S., and Horan, R. (2001). “The economics of to the Water Framework Directive, Unpublished
nonpoint pollution control; Journal ofEconomic PhD dissertation, Economics Division, University
Surveys 15(3): 255-89. of Stirling.
Shortle, J.S., Ribaudo, M., Horan, R., and Blandford, Viscusi, K., Huber, J., and Bell, J. (2008). “The
D. (2012). ‘Reforming agricultural nonpoint economic value of water quality, Environmental
pollution policy in an increasingly budget- and Resource Economics 41(2): 169-87.
The decision about what species and habitats to protect and how they should be protected is
a major issue for society, as terrestrial and marine habitats hosting high levels of biodiversity
are being irreversibly lost or degraded on a massive scale. In economic terms, biodiversity is
becoming scarcer as some species are lost and others become more vulnerable to extinction.
The ecosystems that we depend upon to provide life-support functions are losing the species
that maintain these functions. We approach this topic by addressing a series of questions.
e What is biodiversity?
e What is the objective of biodiversity conservation?
e Howis biodiversity valued, especially in terms of its contribution to ecosystem services?
e And how can we design policies to ensure consumers, farmers, foresters and fishermen
have an incentive to protect biodiversity cost-effectively?

12.1 Introduction

Throughout the ages, the processes of evolution have driven innumerable species to extinc-
tion and, by mutations, have created new species better suited to the evolving environment.
For most species, this process has proceeded gradually. Evidence from fossil records sug-
gests that millions of years may separate one distinct species from another, and new species
are often only slightly genetically different from their predecessors. Periodically, there have
been mass extinctions (Raup, 1988; Stork, 2010) when a iarge number of species have disap-
peared. It is estimated that there have been five mass extinctions in the last 600 million years
when 65-95 per cent of species known in earlier fossils have disappeared (Raup, 1988). The
last of these was due, it is thought, to meteorite strikes that briefly plunged the whole earth
into an extended winter. Many biologists (May et al., 1995; Stork, 2010) believe that we are
currently witnessing a mass extinction, but this time the sudden loss of biodiversity is due
to human actions.
Biodiversity can be defined as: the variability among living organisms from all sources,
including inter alia, terrestrial, marine, and other aquatic ecosystems, and the ecological
complexes of which these are a part. This includes diversity within species, between species,
234 APPLYING THE TOOLS

and of ecosystems (Convention on Biological Diversity, 1992, Article 2). Thus biodiversity
represents the variety of life, whether in terms of different species of birds and mammals,
and the variability across ecosystems.
The exact rate of species extinction is difficult to assess because we do not know how many
species there are, and thus it is difficult to estimate how many have been lost (see Box 12.1).
Estimates range from 5 to 30 per cent of species loss per decade (for a review, see Stork, 2010).
The IUCN (2010) predicts that the rate of extinction that we are currently witnessing is 100
to 1,000 times the average rate measured in the fossil record. The key determinant of the rate
of species loss is the destruction of high-biodiversity habitats, which include rainforest,
Mediterranean vegetation communities, and coral reefs (Myers et al., 2000).
In ecology, a key concept in discussions relating to biodiversity, conservation, and
extinction is the species—area curve. A species—area curve is a relationship between the area
of ahabitat and the number of species found within that area. Larger areas tend to contain
more species than smaller areas, but as the area considered increases the marginal (extra)
increase in species declines (MacArthur and Wilson, 1967). The curve typically follows a
non-linear relationship such as that depicted in Figure 12.1. For concreteness, let us sup-
pose that this curve represents the number of species of higher plants in a given area of
rangeland habitat. The curve predicts that, as habitat area is lost, initially very few species
are lost. For instance, if the area lost is from a, hectares down to a,, then we can see from

There are two surprisingly difficult questions in relation to measuring the state of biodiversity and the loss
of biodiversity due to extinction: First, how many species are there on earth? Second, how many species
have gone extinct due to habitat loss?
So far, 1.7 million species have been described by science (Stork, 1996). These include about 250,000
plants, 44,000 vertebrates, and 751,000 insects. We know with some confidence that the 4,000 mammals
described are an accurate estimate of the total, but the numbers of described insects, spiders, fungi, and
other primitive species are a gross underestimate of the total number of species. Other groups such as
viruses, bacteria, and algae have been relatively unexplored (Colwell, 1996) and their species number
probably exceeds that of all other species combined.
Current estimates are between 5 and 15 million species—this evidence is reviewed in Stork (2010). Of
this large number of species, about 85-95 per cent have not been described by science. There are four
approaches to estimating extinctions: species-area based estimates, estimates from empirical data of
known extinctions, estimates from the use of Red Data lists (IUCN, 2010), and co-extinctions. The
species-area curve shown in Figure 12.1 is used as a statistical modelling approach to estimating
extinctions due to habitat loss. Thus as a rainforest or coral reef habitat is progressively lost, the rate of
species loss is expected to accelerate. Co-extinction occurs when if a species becomes extinct, then all the
species that depend upon it also become extinct. Climate change is also predicted to be an important
driver of species loss (Opdam and Wascher, 2004).
The more comprehensive species-area curve estimates of extinction estimate a rate of extinction per
decade of between 5 per cent and 30 per cent of all species on earth (Stork, 2010: table 1). More recent
estimates tend to predict a lower rate, because it has been found that some species have been able to
‘hang on’ in regrowth forest. For instance, for the critical Amazonian rainforest region, by 2050 there is
expected to be a 12-24 per cent reduction in forest habitat and this is expected to lead to 5-9 per cent of
species becoming extinct (Feeley and Silman, 2009).
BIODIVERSITY 235

S(a) Species
number

So
S}
S2

Area (a) of habitat remaining

Figure 12.1 The species-area curve.

the vertical axes that the corresponding species loss is the difference between s, and s,. A
further habitat loss of the same area results in a greater species loss that is, the difference
between s, and s, is smaller than the difference between s, and s, for an identical reduction
in area. The implications of this fundamental ecological concept for conservation planning
are that if a conservation agency wants to protect the maximum number of species, they
may need to allocate more resources towards relatively rare habitats and less towards com-
mon ones.
Biodiversity loss involves more than the loss of a few high-profile species such as ele-
phants and black rhinoceros; it concerns any regional loss of species or any reduction in the
geographical range of species that reduces their genetic diversity.
Why is this accelerated biodiversity loss a cause for concern? The loss ofa species is a cost
to society for three reasons. First, the species lost may have a direct value as a source of
knowledge that can be used to locate genetic material for food crops, or as a source of active
chemical compounds for new medicines. Second, a species may play a critical role in main-
taining an ecosystem and providing humanity with a range of‘life-support’ ecosystem ser-
vices, such as nutrient cycling and water catchment regulation. Finally, a species may have
aesthetic and non-use benefits; for instance, people like to see ‘charismatic’ species such as
birds and butterflies in the wild, and place a value on knowing that a valued ecosystem is
functioning and protected. The Economics of Ecosystems and Biodiversity (TEEB) project
recently highlighted the potential economic consequences of current rates ofloss in biodi-
versity (TEEB, 2010).
Why are species lost? The main threat to biodiversity is a loss of habitat by conversion
of land to other uses, especially agriculture, and the degradation of habitat, especially
through fragmentation. Particularly at threat are rainforests, old-growth temperate for-
ests, rangelands, heathlands, and coral reefs, which together provide habitats for over
236 APPLYING THE TOOLS

Metrick and Weitzman (1998) employ the following metaphor to explain the problem of choosing which
species to preserve. Consider a species as if it were a library full of books. Each book is a gene. Common
books (genes) are housed in many libraries (species); others in only one. Resources can be used to protect
libraries and reduce the probability of destruction by fire, but ata cost. The problem, then, is to
determine how the fire-fighting resources should be distributed amongst the different libraries.
A library has two attributes, the building (the phenotype of the species) and the information contained
within it (the genes). We value the species for its aesthetic appeal or its use value. We may value the
books (genes) for the diversity that they bring to species in terms of colour and form, or we may place a
utilitarian value upon them for the ideas that they may contain for future medicines and food crops.

50 per cent of known species (Stork, 2010). An additional factor is climate change, which
may drive species to premature extinction because they are not able to evolve rapidly
enough to adapt to a new climate.
By deciding to protect some habitats and depleting others, man decides which species are
destined to become extinct and which survive. Collectively, human society acts in the role
of Noah, deciding which species should be allowed on the ‘Ark ofsurvival’ and which should
be allowed to perish (see Box 12.2).
Human actions have increased the rate of extinction; therefore economics should be
involved in finding solutions. Shogren et al. (1999) highlight the importance of economics
to decisions that society makes concerning biodiversity: economic activity determines the
risk that a species faces of extinction; whilst protecting a species has an opportunity cost in
terms of reduced resources for other valued public and private goods. In other words,
resource scarcity means that we often face choices over what biodiversity to prioritize for
conservation. Finally, economics can offer guidance into the most cost-effective way of con-
serving biodiversity; for example, in the design of agri-environment schemes, or interna-
tional schemes for biodiversity conservation in developing countries (Ferraro and
Pattanayak, 2006; Claassen et al., 2008).

12.2 What to Conserve?

Active conservation of biodiversity has a cost in terms of lost economic opportunities and
the allocation of resources to conservation. Given the large number of threatened species,
we cannot save everything; thus society must make difficult choices about how limited
resources are allocated to biodiversity conservation rather than other public goods. This
chapter explores society values for biodiversity and ecosystem services, and how society
may make these choices.

12.2.1 Economic insights into conservation objectives

The issue of which species to preserve might, in economic terms, be judged on the basis of
whether the benefits of preservation exceed the costs. This calculation is complex, because
both the benefits and the costs are difficult to estimate. Weitzman (1992) developed a theory
BIODIVERSITY page

of which species should be preserved, based on a measure ofdistinctiveness (see also Metrick
and Weitzman, 1998). Distinctiveness can be interpreted as the ‘genetic difference between
a species and its closest relative. However, distinctiveness says nothing about how much
people actually like a species or how effective a species is in providing ecosystem services; it
may say something, in the case of higher plants, about its potential medicinal value or about
its importance in filling a particular niche within an ecosystem.
Policy-makers often have to make choices between species or ecological communities
when allocating resources for conservation. In an ideal world, they might be able to use the
following formula:

w; (a;) = p; (a; {vs ac Vy ale U4;} a c;(a;), Cia)

where the term w,(a;) is the expected net benefit over a long time horizon of conserving a
species or a particular ecosystem i.
Natural species populations are often highly variable and it is impossible to ensure that a
species will survive. The term p;(a,) gives the probability of a species surviving. The probabil-
ity of survival depends upon the decisions of landowners and policy-makers to conserve the
habitat. In this simple model, it is an increasing function of the remaining area of habitat q,.
The value of conserving a species or an ecological community has three components. The
direct value of distinctiveness Vv is due to the expectation that a species will provide ‘infor-
mation’-that is chemical information about potential drugs or genetic material with com-
mercial value; for instance, in agriculture. This is an expected value, because we cannot be
certain that a species or a group of species will provide commercially valuable material. We
consider the nature of this component of value later in this chapter. Species distinctiveness
determines this value as it measures how close a species is to related species with similar
properties. If two species have very similar genotypes, then they may be close substitutes in
terms of their chemical constituents.
The other component of value, the ecosystem service value v; , gives the contribution of the
species to ecosystem services by increasing ecosystem resilience and underpinning the func-
tioning of ecosystems. As we discuss later, this value depends upon how many substitute spe-
cies there are within an ecosystem capable of taking over the niche occupied by a species: a
species with many close substitutes is in general less valuable than a species with few or none.
Finally, the aesthetic and non-use value u; (we outlined this concept in Chapter 3)
depends upon society’s preferences for a species or an ecological community. The species
contributes to the aesthetic value of the conserved area. The total value of a species is there-
fore vj = {vi + vf + uj}.
Costs c,(a;) are in terms of the direct costs of conservation and the opportunity costs of
forgone development opportunities as a function of the area of habitat preserved. This
reflects the fact that as the area of habitat increases, then this increases the opportunity cost
of lost development opportunities. Opportunity costs could, for example, consist of forgone
profits from conversion of rainforest to cattle grazing. Figure 12.2 shows the trade-off
between costs and the probability of survival. In this diagram, the current habitat is given by
dy. Habitat can be increased by restoration and increased conservation management. There
may be an asymmetry of costs, in that if land is lost to save conservation costs, say from ay to
a,, going back from a, to ay by restoration will entail a far greater cost than conservation, that
is c’) > Cy in Figure 12.2.
238 APPLYING THE TOOLS

Cost conservation

Cost restoration

of
Cost
conservation

p(a,) P(d) 1
Probability of survival as a
function of habitat

Figure 12.2 The trade-off between cost and probability of survival.

As more habitat is preserved, or even re-created, then the probability of survival

approaches, but it never reaches one because there is always some possibility that a species
will become extinct whatever conservation actions are undertaken. The cost curve becomes
steeper as society allocates progressively more valuable land, from agriculture, forestry, or
development, to restore the habitat. In fact, due to the irreversibility of destruction for many
habitats, it may be impossible to restore some of the original habitat at any cost. It is also a
general rule that re-creating habitat is many times more expensive than conserving existing
habitat. Prach and Hobbs (2008) make this point in relation to land restoration following
mining.

12.2.2 Allocating resources between species and ecosystems

A problem often faced by government agencies with responsibility for conservation is how
to share a budget allocation between one species or another, or between spending on differ-
ent habitats. Equation (12.1) gives the objective function related to a single species or habi-
tat, but in reality there are often many species and threatened ecosystems, and a choice has
to be made of allocating resources between them. Let us set up a simple version of this
problem, in which the regulator has to pay farmers the opportunity cost for conservation
and limited funds are allocated between just two species. We will name the species after two
Australian endangered species, the chuditch (also known as the western quoll), Dasyurus
geoffroii (a small marsupial predator endangered by introduced predators such as the fox
and cat and loss of habitat, and now confined to the lower south-west of Australia), and the
bilby, Macrotis lagotis (an endangered omnivorous marsupial threatened by loss of habitat
and predation, and now confined to remote arid regions of northern Western Australia).
BIODIVERSITY 239

The Department of Environment and Conservation (DEC) in Western Australia has a

budget to allocate between these species. The problem they face is as follows. They would
like to maximize the expected net benefit from the conservation of both species:

(yi
>Fuad) = W(Ayin,) + W(Aehud )

= Priv (Aoi Von + Penua (Fehud Vihua — Coit (Gray) =Copua (Fena)> (i272)

where the area allocated to bilby habitat is a,,,, and the area allocated to chuditch habitat is
A-yq- The total value of each species is Vj, and v%,,,. The budget constraint, assuming all
funds are allocated, is given by

M = Cyiy (Ayan) + Ccrud (Getua )> (1273)

where M is the total budget available in dollars and c,;», (a,;,,) is the opportunity cost of pro-
viding a given area of bilby habitat and c,,,,4 (@;j,q) the same for chuditch.
If the DEC wishes to maximize the objective function in equation (12.2) subject to the
budget in equation (12.3), what rule might they use? One approach would be to consider the
gains from providing one more hectare of bilby habitat compared to one more hectare of
chuditch habitat. To this end, the marginal net benefit of bilby habitat is as follows:

MNByin, = (AP
sinVoie — ACrin )/AQyiy > (12.4)

and the marginal net benefit of chuditch habitat is

MNP oud = (APchudVenud — ACchua !A@ nua (12.5)

The term A is used to indicate the change in a term with respect to the corresponding area.
If we consider very small changes, then this is interpreted as a derivative.
As proposed by Metrick and Weitzman (1998), these terms can be compared to provide
a criterion to guide investment in habitat. Thus if, for instance, MNB,i», > MNP. ja this
would imply that there should be more resources from the budget allocated to bilby conser-
vation and less to chuditch.
The example presented here is highly simplistic, but since the use of systematic conserva-
tion planning has become more common (Margules and Pressey, 2000), so has the inclusion
of the opportunity cost of conservation in terms of land or marine resources such as fish
catch (see Box 12.3). Examples of this approach include Fernades et al. (2005) for the Great
Barrier Reef and Polasky et al. (2001) for cost-effective terrestrial ecosystem conservation in
Oregon. It is more challenging to put a value on the biodiversity and the probability of sur-
vival. We consider the value of biodiversity in the next section.

12.3 Values of Biodiversity

The total value of all biodiversity is the sum of all human welfare for current and all future
generations. Without any biodiversity, humanity would cease to exist. In this sense, the
240 APPLYING THE TOOLS

In a paper in Nature, two ecologists—Possingham and Wilson (2006)—question biodiversity conservation

priorities and recognize the importance of cost in conservation decision-making:

The variety of life on Earth is in rapid decline and global spending on nature conservation is
inadequate to arrest that decline. Consequently, resources for conservation must be allocated
to secure the ‘biggest bang for our buck’ In recognition of that need, scientists have identified
biodiversity hotspots, where extraordinary concentrations of biodiversity exist.
Biodiversity hotspots (Myers et al., 2000) combine high biodiversity with a high risk of biodiversity loss.
Wilson and Possingham propose that the allocation of funds to biodiversity hotspots around the world
should depend on how much they cost to conserve. Without information on such costs—including the
opportunity cost of land set aside for protection—money spent could be less cost-effective in terms of
conserving a given number of species.

economic value of global biodiversity is infinite. However, this self-evident truth does not
help us with more relevant questions about the value of marginal changes in biodiversity,
which species should be protected and which neglected. Commercial activities that may
push species to extinction, such as agricultural expansion and intensification, deforestation,
and urbanization, have costs and benefits that, at least in part, are measured by market
prices. Valuing a loss of biodiversity is more difficult, because markets for its preservation
are either missing or incomplete. The key economic questions concern the marginal value
of biodiversity itself and the benefits of preserving a particular part of a habitat. As noted
earlier, the benefits of preservation derive from three sources: first, direct values due to
medicine and agriculture; second, ecosystem services such as recycling nitrogen, filtering
sediment, and decomposing waste; and third, aesthetic use and non-use values. We con-
sider each ofthese sources of value in more detail.

12.3.1 Direct values

Currently, we use approximately 40,000 species of plants, animals, fungi, and microbes
(Elridge, 1998; Heal, 2000). Organisms have a vast range of uses, some of which are often
taken for granted. For instance, the bacteria actinobacteria is used to produce a common
form ofantibiotics. From fungi, we obtain the common mushroom, whilst Genea hispidula
provides the structure in legumes for nitrogen fixation. Useful insects include silkworms,
bees (which provide pollination services to agriculture), and ladybirds that predate upon
aphids. Useful animals include sheep for food and fibre, and buffalo, which provide meat,
milk, draft power, transportation, leather, and fertilizer. The higher plants provide crop
plants, timber, insecticides, and fodder, and are also an important source of traditional and
modern medicines. These compounds have provided the basis of many common drugs,
including aspirin, which comes from willow bark, and taxol, a cancer treatment, derived
from the Pacific yew.
Over time, we have become adept at finding and developing species that are useful to us,
and this continues as we search for new crops, and new genetic material for existing crops
and new medicines. Biotechnology has made it possible to take traits such as disease
BIODIVERSITY 24)

resistance from close relatives to crop plants, and incorporate these traits into commercial
varieties. For these genes to be discovered, they must already exist in wild or cultivated spe-
cies. By destroying a habitat, we may inadvertently destroy a close relative to a crop species
that contains valuable genetic material. An example is Zea diploperennis (IItis et al., 1979), a
rare relative of commercial maize (Zea mays), which was unknown and on the verge of
extinction in its last remaining habitat in the Sierra de Monantlan in Mexico. This plant
shows resistance to many of the diseases that affect the commercial crop, and is currently
starting to provide genetic material for new commercial maize varieties.
As diversity amongst wild relatives of crop plants is lost, we are in danger of becoming
dependent on an increasingly small number of ‘supercrops’. But in the future these are at risk
from disease and from pest species that can evolve resistance to pesticides. Without genetic
diversity from older crop varieties and wild relatives, the plant breeder is restricted to the gene
pool available in known varieties. Loss of crop genetic diversity over the last 50 years has been
so rapid that there has been a call to establish a ‘Red List’ of endangered crop species (Hammer
and Khoshbakht, 2005), based on the IUCN (2010) Red List for endangered species.

12.3.2 Bioprospecting for drugs

The genetic codes of organisms contain the instructions to synthesize biologically active
chemicals. Some of these chemicals can be exploited as the blueprint for drugs that are then
produced synthetically. Often, the naturally occurring substance gives the medicinal chem-
ist a ‘lead’ to the likely form of promising molecules, but these need to be modified before
they can be used as medicines. Preservation of biodiversity maintains the number of species
that are available to be used in this way (for a related example in which biodiversity is
exploited commercially, see Box 12.4). However, there is also substitutability between

BOX 12.4. Bioprospecting in Costa

By collecting and cataloguing the endemic plant species of Costa Rica’s rainforests, the Instiuto Nacional
de Biodiversidad (INBio) in Costa Rica (see Aylward et al., 1993) hopes to be able to market genetic
information as a means of financing conservation. In addition to government support, INBio has
commercial contracts with Merck & Co. and the British Technology Group for the exclusive rights over
tested samples.
This type of contract is unlikely to provide significant funds for conservation until the host countries
derive more benefits from the discovery of such material (Swanson, 1994). Despite its financial backing
and worldwide publicity since 1991, INBio has generated a surplus of US$2.5 million from
bioprospecting. This does not suggest that the total value ofthe institute's work in helping conserve the
biodiversity of Costa Rica is equal to this amount. However, it does show some of the difficulties of
converting unique genetic material into substantial profits through commercial contracts.
A case study for the Tapanti National Park puts the value of bioprospecting in perspective. Bernard et al.
(2009) undertook a cost-benefit analysis of the forest to estimate values through stakeholder surveys on
ecosystem services for the 58,000 ha national park. The provision of ecosystem services related to
hydrological regulation and drinking water were estimated at US$2.5 million per year. Bioprospecting was
undertaken in the park by INBio, but was yet to return any funds through the discovery of anew compound.
This illustrates that the value of bioprospecting is likely to be low relative to other ecosystem services.
242 APPLYING THE TOOLS

organisms in an ecosystem. First, organisms may occur across a wide ecological range.
Second, different plants may produce chemicals that have similar medicinal properties. This
implies that the value of the active ingredients cannot be entirely related to the preservation
of a local population of a particular plant.
The value of a new drug is measured by its contribution to improving the treatment of
disease. For instance, a new cancer drug, such as taxol derived from the yew tree, may increase
remission rates, reduce the costs of treatment, and improve the quality of life. Its value is
measured by how much better the new drug is than its closest substitute in treating disease.
Consider the following simple analysis of the value of species (Simpson et al., 1996). The
value of a new product is given by the return R from the new drug and the costs of analysing
a specimen c. The probability of success in finding a new drug is given by p. Therefore, the
expected value of analysing a single plant specimen, v(1), is the expected revenue less the
costs. That is:

vl) = pR=c.

For instance, if R= $10 million, c = $0.5 million, and p = 0.1, so that there is a 10 per cent
chance of finding a substance that will form the basis of a new drug, the expected value of
the specimen is (0.1)10 - 0.5 = $0.5 million. If there is more than one candidate organism to
produce a similar drug, then we identify a search procedure, in which searching stops when
a discovery is made. Therefore, with two organisms, the expected value is given by

v(2) = pR—c+(1— p)(pR—c),

where the second term gives the expected value of the second organism and the term (1 — p)
is the probability ofnocompound being found in the first organism. For instance, ifp=0.1,
then the probability of the second organism needing to be analysed is 0.9, or 90 per cent. The
value of having this extra species to analyse is simply the expected value of the second spe-
cies, (1=p)(pR=c).
The formula can be generalized to give the expected value of any number (n) of species,
as follows:

v(n) = pR-c+(1— p)\(pR—c) + (l= p)?(pR=c)+...+ (1 — p)*"(pR—c).

The value ofthe ‘marginal’ species is

(1 — p)"(pR — ¢),
that is, the expected value of having one more species to test for the medicinal product. The
expected value of the marginal species is inversely related to the probability of success and
the number of species. For instance, if there are 1,000 species and a 10 per cent (p = 0.1)
probability of finding a medicinal product, (1 — p)" = 1.747x10~ (a very small number). In
percentage terms, the success rate is the number of plants, on average, in a hundred that
have a biologically active substance that could form the basis of anew drug. If the probabil-
ity of finding a product is only one-tenth of 1 per cent (p = 0.001), then (1 — p)"= 0.3677; at
BIODIVERSITY 243

the same time, the (pR — c) term declines, but not as quickly as (1— p)” increases. If the
number ofspecies is reduced from 1,000 to 100, and assuming a one-tenth of 1 per cent suc-
cess rate, then (1— p)" = 0.905. In ecosystems, there are often large numbers of species, and
as the number of candidate species increases, then the value of the marginal species falls
close to zero. This implies that the ‘bioprospecting’ values for the marginal species in a rain-
forest, say, may be quite low.
Consider the following example. The higher plants have, historically, produced the larg-
est number of new drugs. So how do we value the conservation ofone plant species in terms
of its potential medicinal value? Simpson et al. (1996) suggest the following approach and
parameters. There are about 250,000 (n) species of plants known to science (Wilson, 1988).
On average between 1981 and 1993, 23.8 new drugs per annum were approved by the US
Food and Drug Administration. This is an approximate measure of world discovery rates,
as drugs first sold in other countries need to gain approval before being sold in the United
States. Of these new drugs, approximately one-third are derived from higher plants
(Chichilnisky, 1993); thus about eight new drugs (k) are discovered from higher plants each
year. The cost of discovering a new product is around US$300 million. The expected rev-
enue (R) is US$450 million, which assumes that the drug companies make a 50 per cent
return on research costs. The cost of evaluating a single sample (c) is US$3,600. Use of the
formula for the value, v, of an individual species, gives an estimate of the value of preserving
an individual plant species on the basis of its medicinal value of US$7,109. Is this value sig-
nificant in preserving an additional hectare of rainforest? The answer depends critically on
how common plant specimens are. If a plant is widespread, a hectare of rainforest may
contribute little to its long-term survival. If, instead, a relatively small area of rainforest is
the host to many rare species, it could be highly valuable. On the basis of providing speci-
mens for drug prospecting alone, as rainforest is cleared and the remaining species are
found in a smaller and smaller area, this value becomes more significant to the conservation
value of a hectare of rainforest. For a discussion of this point relating to Costa Rica, where
the value of bioprospecting accounts for a small proportion of the ecosystem value of one
forest area, see also Box 12.4.

12.3.3 Ecosystem service values

In simple terms, the ecosystems that surround us ‘sustain and fulfil human life’ (Daily, 1997: 3).
As made clear in the Millennium Ecosystem Assessment (2005), ecosystems provide us with
numerous direct and indirect services, including waste assimilation, water purification, and
nutrient cycling (MEA, 2005). Complex ecosystems provide these services, but we are often
unable to assess the value of a particular species within an ecosystem.
Ecosystem cycles relate to energy input, circulation of the elements essential to life—
carbon, nitrogen, hydrogen, and so on (biogeochemical cycles)—and water circulation. All
living organisms require energy to maintain their vital functions and grow. Energy in
terrestrial and most marine ecosystems must be constantly renewed from the sun. Plants
capture the sun’s energy through photosynthesis to produce tissue. Plants are called auto-
trophs: they live on inorganic elements, such as carbon dioxide and water. All animals,
insects and micro-organisms are directly or indirectly fed by plants, and are called consum-
ers or heterotrophs. For example, a rabbit eats grass and produces waste, which feeds the
244 APPLYING THE TOOLS

Primary producers: autotrophs

(e.g. plants, seaweed, algae)

Consumers: heterotrophs
(e.g. grazing animals and fish)

i fi

Energy flows Nutrients, carbon dioxide,

water

Figure 12.3 Ecosystem cycles.

decomposers. A fox that eats the rabbit benefits indirectly from plants. Organic matter cir-
culates in a food chain, is transformed, and decomposed into basic elements (CO,, water,
nitrogen, etc.) that are available for plants to use. The energy captured in the food cycle is
dissipated at each stage as heat.
The number of species in an ecosystem occupying these roles varies. Diverse ecosystems
may have a large number of species in a particular role, and may be able to survive the loss
of anumber of species and continue to function. Damage to one part of the ecosystem may
be compensated for by adjustment in another part. One way of thinking about an ecosystem
is as flows of energy through a web of interacting species. If one of these energy pathways in
the web becomes blocked—for instance, by a species becoming extinct—then it may not
affect the whole system so long as other pathways remain open. Hopkin (quoted in Putman
and Wratten, 1984: 350) uses the analogy of the blood circulation system. If a capillary is
blocked or damaged, the body adapts by developing more capillaries or using existing capil-
laries. If main veins or arteries become blocked, then the body is damaged and finds it
impossible to adapt. Thus diversity leads to resilience, and resilience is important if the
system is to continue functioning when subjected to a shock such as loss ofland to agricul-
ture or an extreme climatic event. In contrast, impoverished ecosystems tend to be more
fragile and less resilient to the loss of key species.
245

Consider some examples of ecosystem services and how they might be affected by spe-
cies loss. Trees in a rainforest act like a sponge, and moderate the local water cycle by
retaining water in the soil and transpiring (evaporating) rainfall. In this way, they reduce
the frequency and intensity of floods. The loss of forest tree species would significantly
reduce this ecosystem service, whereas the local loss of aspecies of monkey may not directly
affect this service. Coral reefs provide protection to coastal areas and protect delicate
coastal wetlands and mangrove swamps from storms. The loss of a significant part of the
coral reef by pollution, sedimentation, or a sea-level rise due to climate change would
result in a significant loss of ecosystem services, whereas the local loss of a species of fish
would not directly.
The general point here is that the fewer close substitutes a species has in terms of its eco-
system functions and productivity, the more damaging is its loss. For instance, mangrove
trees are critical to the functioning of coastal swamps, as they stabilize sediments and filter
the water. The loss of a species of mangrove, the primary producer in that ecosystem, would
have a devastating effect on the local marine environment, since there are no close substi-
tutes for this species in its role as a primary producer capable ofsurviving in salt water and
stabilizing the sediment in coastal swamps (see Box 12.5).
Functioning ecosystems provide society with a range of benefits to people, including
natural resources, flood mitigation, pollination for agriculture, carbon sequestration,
and reduction in soil erosion. These benefits have been termed “ecosystem services’
(Bateman et al., 2011). These services include goods such as fish or fuel wood, and ser-
vices such as recreational opportunities, storm reduction, water purification, and cul-
tural benefits, all of which have economic value (UK National Ecosystems Assessment,
2010). The link between the provision of ecosystem services and biodiversity is that, typi-
cally, biodiverse ecosystems are more resilient to environmental shocks and are often
better able to provide ecosystem services in a range of climatic conditions. Biodiversity
can also play a key role in the functioning of ecosystems, and thus in the quantity of ser-
vices that they deliver.

BOxi2s Mi
The high productivity of the Gulf of Mexico fishery is partially attributed to the extensive areas of coastal
lagoons and estuaries, which are dominated by the mangroves. Mangroves provide ideal breeding
grounds for commercially exploited shrimps and finfish. In this region of Mexico, mangrove is under
threat from mariculture (the development ofshellfish farms) and urban encroachment. In their study of
the Campeche fishery, Barbier and Strand (1998) estimate that, on average, the fishery loses 0.19 per cent
of its annual revenue for each km? of mangrove deforestation. This estimate is based upon a simulation
model that assumes that the loss of mangrove swamp reduces the carrying capacity of the coastal shrimp
fishery, which benefits from the nursery function provided by the mangroves.
An interesting conclusion of the study is that the costs of mangrove deforestation—and, indeed, of the
fishery as a whole—would be greatly increased ifthe fishery was converted from open access to a
regulated fishery. This problem is an example of two missing markets: there is no market for the
ecosystem services provided by the mangrove forests and there is no market for the shrimp stocks. The
lack of amarket for shrimp stocks, due to open access, leads to a reduced value of protecting the
mangrove swamp.
246 APPLYING THE TOOLS

[Link] Case study: marine ecosystems in Alaska

Economies and ecosystems are complex, interrelated systems. This complexity means that,
often, the parts of an ecosystem/economy complex are analysed separately by economists
and ecologists rather than analysing the whole system. An exception is a recent study of the
Alaskan economy and its underpinning ecosystem (Finnoff and Tschirhart, 2008). The out-
line of their model gives an insight into how the ecosystem links to the economy, the deter-
minants of value, and the role of biodiversity.
The Alaskan economy depends heavily on renewable and non-renewable natural
resources to support its economy. A large number of Alaskan firms engage in commercial
fishing and tourism activities that depend directly on the marine and terrestrial ecosystems.
The total value of output of the Alaskan economy is US$32 billion, comprising a fishery
output of US$659 million (2 per cent of total output), an aggregate recreational output of
US$1,661 million (5 per cent of total output), and a composite good output of US$30 billion
(93 per cent of total output).
The model developed by Finnoff and Tschirhart (2008) considers two equilibria, one for
the economy, using a computable general equilibrium (CGE) model that accounts for the
allocation of labour and capital among the three sectors in the economy subject to resource
constraints, and a general equilibrium ecosystem model (GEEM). The models are repre-
sented in Figure 12.4.
The economic model includes production functions that link inputs, labour, capital, and
natural resources to the level of sectoral outputs. The fishery sector is also regulated by a
total allowable catch restriction and a limit on the time for which the fishery is open. The
recreation sector also uses labour and capital, but its output is in terms of the number of
recreational trips sold. In turn, the number of trips sold depends upon the population of
charismatic predators (killer whales, seals, and sea otters). The computable general equilib-
rium model determines an equilibrium allocation of resources between the three sectors by
commodity and factor price adjustments.
The general equilibrium ecosystem model determines the equilibrium species popula-
tions accounting for predation and harvesting. The ecosystem equilibrium is determined by
maximizing behaviour for individual species in the acquisition ofenergy. For instance, indi-
vidual stellar sea lions have the following net energy equation for preying on pollock:

Reread = (Gira oa €) ire * tered Xpvea) ar: Sierg

In this equation, R,,,.q is the net energy flow in kcal per unit of time; Cpred 18 the energy per
biomass unit (kcal/kg); ¢ is the energy expended in locating, attacking, and consuming pol-
lock; x,,,q is the biomass consumed, in kg per unit of time; ved (Xpreq) is the respiration energy
as a function of the level of predation; and Byeq is the resting metabolic rate (Tschirhart,
2004). Ifa predator’s problem is assumed to be net energy maximization, then it will continue
to increase its biomass consumed until the net energy per kg (Cpred ~— €) equals the marginal
increase in respiration. This gives an interpretation ofthe behaviour ofpredators as equating
marginal benefit with marginal cost; however, instead of being measured in money per unit
time, the benefits and costs are measured in terms of energy per unit of time. Other compo-
nents of the model include energy expenditure from the predator avoiding larger predators;
ep
So 10 DIVERS! TY 247
UW93}sKsoq

SMO} Wa}SksOdz
‘saj2YM

10}DE}
SMO|}
UeINe|\yspueys]

J0]|1|

~<G- — — —
/ad\Atas/pooy
‘
‘19]}0

Bas
SUO}|
pue
JE]Ja}s
Bas
dese)
:ss0}epatd

----
eas—
Usa}sey Huljog

‘djay sulysin
‘uo}uejdooz
‘uo}ue|doyAyd

Ayipowiwo
Awiouodq

syodxe
>
AIpowwo05
UexsSe|Vy

(jen1deo

JEW
s}a
pue
inoge})
sjayJEW
10;Ie4

Aypowu07

UBHSE|Y
syodui

spjoyasnoy

"PYSE|\
eunsl4
ZL
SAdIAIaS
Ul
BULLE
Wia}shsode
248 APPLYING THE TOOLS

in this case, stellar sea lions being captured by killer whales. There is an equilibrium that states
that the total biomass demanded by the predator must equal the total amount available from
the prey. Finally, this is linked to a population growth model that determines the population
of sea lions and pollock, the economically important fish stock, in each period.
The value of this interlinked economic and ecological model is that it allows a policy-
maker to understand the joint dependence between the economy and the ecosystem (Finnoff
and Tschirhart, 2008). For instance, a reduction in the pollock catch quota relative to its
1997 levels increases the population of pollock, phytoplankton, sea urchins, sea lions, and
killer whales, but reduces the populations of zooplankton, kelp, and sea otters. Through the
economic model, this increases the price of pollock, whilst the energy ‘price’ marine mam-
mals pay to capture pollock falls. Due to a decline in sea otters, the amount paid by tourists
to view marine mammals falls. In this case study, economic welfare is increased by a reduc-
tion in the pollock catch quota.

[Link] Case study: the decline of Easter Island society

Easter Island is a remote Pacific community some 2,000 km off the Chilean coast. The cur-
rent population is around 2,000 people and this has remained approximately constant since
the first Europeans discovered the island in 1722. Since that time, the island has been viewed
as an archaeological and anthropological puzzle, as it indicated the existence of a sophisti-
cated culture able to produce large stone statue ‘moai’ in the past, and yet explorers visiting
in 1722 noted its small population and found little evidence of cultural richness. Explanations
for the earlier cultural sophistication have included everything from extraterrestrials to
Easter Island being part of the Lost Empire of Atlantis. A more plausible explanation con-
cerns forest resource depletion and mass extinction of trees, and most of the species that
depended upon them, this being an example of co-extinction (Stork, 2010).
Evidence from pollen records indicates that the island was originally covered by exten-
sive palm forests, which provided the population with wood to build fishing boats. The palm
forests also provided wood for quarrying and erecting statues. Initially the population pros-
pered, but the slow-growing palm forests were gradually depleted and without any means of
building boats, the food supply and the population declined. Brander and Taylor (1998)
propose an economic analysis of the decline of the Easter Island community. They term
their model a Ricardo-Malthus model: the reference to Ricardo refers to the structure of
production, and that to Malthus refers to a Malthusian assumption relating population to
the supply of food. The mathematical details of the model are beyond the scope ofthis book,
but some aspects are worthy of note. The authors assume an economy that produces two
goods, a natural resource derived good—fish—and a ‘manufactured’ good, which includes
statues and housing. The initial resource level is fixed, as is the initial supply oflabour. First,
the link between the fishery, woodland, and labour is represented by a renewable resource
model, similar to the one encountered in Chapter 8:

dx/dt = g(x) —h;

that is, the change in the stock x of the natural resource (dx/dt) equals the growth of stock
g(x) less the harvest rate h. In turn, the harvest is given by the Schaefer (1957) model:
BIODIVERSITY 249

h =oxl,,

where @ is a parameter, and L,, is the labour used in harvesting. The palm forest and the
fishery are open access, thus the labour input into the fishery will increase to the point at
which the price of an extra fish equals cost per unit of labour.
The Malthusian component of the model is the link between the population growth and
the food supply:

dL/dt = L(b— m+ ofh/L)).

This equation states that the change in population (dL/dt) (and total labour supply) L is due
to the birth rate b and death rate m plus a fertility adjustment term that gives the harvest
(food) per capita [h/L] multiplied by a parameter @. It is this last Malthusian term that incor-
porates the link between the food supply and population growth. A steady state in this
model is where there is no change in the population or in the natural resource stock; how-
ever, there is no guarantee that it will be a ‘desirable’ steady state, as there is no regulatory
policy in the model to ensure that the forest stock is conserved. In their paper, Brander and
Taylor (1998), using parameters calibrated for Easter Island, predict a population peak in
the early thirteenth century, accompanied by a collapse in the natural resource base fol-
lowed by a gradual population decline, as people overexploited the natural resource basis of
their economy.
It is tempting to draw wider conclusions from the Easter Island example about the dire
consequences for society of overexploiting natural resources today and removing the biodi-
versity that supports a society. To an extent, this may be appropriate: however, there were a
number of features of Easter Island that make it unusual. First, there was the relatively long
time period, between 40 and 60 years, required for the Easter Island palm trees to start pro-
ducing fruit. Second, the island depended heavily on a limited natural resource base. For a
fascinating update on this Easter Island story, and how it relates to current debates over
climate change policy, see Taylor (2009).

12.3.4 Aesthetic use and non-use values

We derive utility from observing species in the wild, or even knowing that a particular spe-
cies is being protected. People value individual species, such as bald eagles, tigers, elephants,
and whales, and whole ecosystems, such as rainforests, prairies, and heather moorland. For
a summary of values that have been estimated for individual species, see Richardson and
Loomis (2009). These values tend to favour the more ‘charismatic’ species and ecosystems.
Thus large mammals, birds, and brightly coloured flowering plants tend to be valued more
highly than insects, fungi, amphibians, and grass species, even though these species may be
essential to the functioning of an ecosystem (Jacobsen et al., 2008). This is not to say, how-
ever, that people do not reveal positive willingness to pay (WTP) for less well-known and
uncharismatic species—as shown, for example, in Christie et al. (2006).
What is likely is that we obtain estimates of the amenity value of individual species and/
or individual habitats, rather than of biodiversity itself (Hanley et al., 1995). These values
may be either use or non-use values (see Chapter 3). Approaches taken to estimating these
250 LYING THE TOOLS

Biodiversity in Poland is largely located in the forests, which cover roughly 30 per cent of the land area.
The Biatowieza Forest is considered the last natural lowland forest in temperate Europe and is especially
regarded for its species richness, and its ecological structures and functions.
For this reason, Czajkowski et al. (2009) selected this forest as the case study region to explore how
Polish respondents valued different attributes of biodiversity conservation. Some ecologists suggest that
the highest priority from a policy perspective should be the protection of all forms of biological variability
within the Biatowieza Forest, including landscape, habitats and their components, and species, as well as
biological and ecological processes. Such a policy would allow for long-term observation offlora
dynamics, succession and regression, fluctuation, degeneration and regeneration, as well as seasonal
changes. Most transformed and actively managed temperate forests in Poland and elsewhere in Europe
do not allow for observations of all these processes, which makes the Biatowieza Forest unique. Almost
40 per cent of currently known species present in Poland (over 11,000) can be found in the Biatowieza
Forest. The forest's habitats are characterized by the presence of a large volume of dead wood, so that
many endangered species that are dependent on this resource are still present. One ofthe flagship
endangered species that exists here is the European bison (Zubr). The Biatowieza Forest has played an
important role in the recovery of this species.
There are two broad approaches to valuing biodiversity: either as a whole, using a contingent valuation
method, in which individuals are asked to provide a WTP for a discrete change in the forest area of
management; or through a choice modelling approach that presents alternative management options as
sets of attributes. The authors applied the second approach and presented four attributes to respondents:
natural ecological processes (conservation that allowed the forest to express dynamic processes of
ecological succession facilitating scientific study); rare species of flora and fauna (conservation targeted at
rare species); ecosystem components, which was concerned with the provision of biotopes and niches
within the forest); and, finally, the cost of protecting the forests, in the form of additional tax payments.
All options were compared with a status quo, which was characterized as what would happen if the forest
management was not changed to enhance conservation. The results suggested €20 per household per
year in Poland for a maximum improvement in all attributes, but also that respondents seemed to be
willing to pay to protect complex ecological aspects of biodiversity in the forest.

values including travel costs, contingent valuation, and choice modelling. Box 12.6 gives an
example of where values on different attributes of biodiversity are assessed using choice
modelling: another example is to be found in Christie et al. (2006).

12.4 Biodiversity and Conservation Policy

Declining levels of biodiversity have led to policy responses at three levels: international,
national, and local. International agreements on biodiversity often provide a guiding frame-
work for national and local policies. In the future, international policies may become more
important in providing funds for conservation in developing countries where biodiversity
is concentrated.
Conserving biodiversity policy concerns securing optimal investment in a global public
good. In common with any public good, there is an incentive for individuals and nations to
free-ride and hope that the good is provided by others. Since the Convention on Biological
Diversity (CBD) brought biodiversity loss into sharp focus, success in stemming the rate of
BIODIVERSITY 25]

biodiversity loss has been limited. The Millennium Ecosystem Assessment (2005) con-
cluded that 60 per cent of ecosystem services worldwide have become degraded, most in the
past 50 years. Past targets introduced as part of the CBD’s efforts to reduce the rate of biodi-
versity loss by 2010 have not been met. In fact, the rate of biodiversity loss has continued to
accelerate. In a commentary on the 2010 COP 10 talks in Nagoya, Moony and Mace (2009)
state that

Even the most conservative estimates suggest that an area of tropical rainforest greater than the
size of California has been destroyed since 1992, mostly for food and fuel. Species extinction
rates are at least 100 times those in pre-human times and are expected to increase.

(op. cit.: 1474)

What is going wrong? The problems appear to be, first, a poor targeting of resources for
conservation; second, a lack of funds, given the scale of the problem; and, third, a problem
of designing policies that give private farmers and fishermen incentives to protect biodiver-
sity (Ferraro and Pattanayak, 2006). It is often overlooked that whilst international treaties
and the national policies inspired by them may have scientifically justifiable objectives, if the
farmers or fishers whose actions directly affect biodiversity has no economic incentive to
modify his or her practices, then biodiversity will continue to be lost.
The challenge in designing conservation policy for private land was put succinctly by
Shogren et al. (1999: 1260) in their economic critique of the US Endangered Species Act
1973:

Just as policy makers cannot ignore the laws of nature, neither can they ignore the laws of human
nature when protecting endangered species. Economic behavior matters in protecting and
recovering endangered species. Effective federal and local policy requires that we adjust our
perspectives and better integrate knowledge about human actions and reactions to species risk
into the mix of influences on endangered species policy.

Significant progress has been made in understanding the nature of conservation science;
however, finding cost-effective policies in developed and developing countries has proved
elusive, despite significant expenditure.

12.4.1 International policy

In Rio de Janeiro on 5 June 1992, the Convention on Biological Diversity came into being at
the United Nations Conference on Environment and Development (UNCED). The conven-
tion was signed by 154 countries, and recognized the need to preserve biodiversity not only
in terms of reducing the rate of species extinction, but also in terms of preserving ecosys-
tems and genetic variability. The broad objectives of the convention are the conservation of
biological diversity, the sustainable use of its components, and the fair and equitable sharing
of benefits from the use of genetic resources.
The objectives of the convention are far-reaching, but what tangible benefits have there
been? There is a danger that a convention with such sweeping objectives, backed up by lim-
ited funding, will be ineffective. There is a need to translate international policy into national
and local policies that provide additional incentives for biodiversity protection.
252 APPLYING THE TOOLS

The CBD includes the General Environmental Facility (GEF), which provides funding
from developed countries for projects in developing countries to protect biodiversity. The
economic rationale for transferring funds to developing countries to preserve biodiversity
is that biodiversity is a global public good, that should be paid for by all countries—not only
those countries that ‘host’ the biodiversity. The principle behind this fund is that it pays for
the incremental costs of projects that provide additional global benefits. Between June 1992
and November 1998, the GEF has provided a total of US$1.74 billion for biodiversity pro-
jects (COP, 1998). The replenishment ofthe fund was expected to be in the region of US$2.75
billion. However, this sum is small relative to the size of the problem and the potential value
of biodiversity. Research indicates that the global funds available will not be sufficient to
compensate the developing countries for the likely costs, including the opportunity cost,
that they face in conserving biodiversity.
The 10th Conference of the Parties meeting, held in Nagoya, Japan in 2010, confronted
the failure of the international community to reduce the rate of biodiversity loss and meet
the targets set in 2002 to reduce the rate of biodiversity loss by 2010. This failure has been
assessed by a range of indicators of biodiversity (Butchart et al., 2010). Indicators are classi-
fied as state indicators, which measure the state of biodiversity—for instance, the area of
natural forest, pressure indicators, which include agricultural nitrogen use, and response
indicators that measure how countries are responding to biodiversity loss. Almost all indi-
cators are heading in the wrong direction for biodiversity; in other words, measures of the
state of biodiversity are declining and pressure is increasing. Some of the response indica-
tors, including aid tied to biodiversity projects, are increasing, and the area protected is
increasing.
In response to this dire state of affairs, the revised CBD includes a new Strategic Plan for
Biodiversity for the Period 201 1-20—‘Living in harmony with nature’ —and a new protocol
related to access and benefit sharing. In his commentary (Harrop, 2011: 119) proposes that
the CBD
... is an example of a ‘hard’ international law that possesses a ‘soft’ nature. [However ...] The
obligations in the text are for the most part textually diluted or heavily qualified leaving exten-
sive discretion to Member States in the manner oftheir implementation.

Thus while 190 countries have signed up to a binding convention, the way in which it is
worded means that it is effectively not binding on countries. However, some significant
progress has been made on sharing the commercial benefits of biodiversity with the coun-
tries that host the biodiversity (see Box 12.7).
International Convention on Trade in Endangered Species (CITES). The International
Convention on Trade in Endangered Species is perhaps the most important international
agreement relating to the protection of biodiversity. The convention was signed in March
1972 and came into force 3 years later. It aims to protect endangered species by restricting
trade in those species, effectively reducing or withdrawing demand for a species through
trade. ‘The mechanism is that endangered species are listed in Appendix I and potentially
threatened species are listed in Appendix II. These lists are revised at biennial meetings. An
Appendix I species requires a permit from the exporting country to ensure that trade will
not be detrimental to the survival of the species, and a permit from the importing country
to guarantee that the species will not be used for commercial purposes. This is effectively a
BIODIVERSITY 222

BOX 12.7 Valuable Viper Venom

It was known that people bitten by a Brazilian pit viper (Bothorps jararaca) suffered an often sharp drop in
blood pressure, and blood thinning that often led to victims bleeding to death. The venom was extracted
by a Brazilian biochemist, Mauricio Rocha de Silva, in the 1960s. Initial work by a British pharmacist,
John
Vane, and then an extended period of commercial and academic development led in 1975 to a synthetic
compound being identified that mimics the effect of the snake venom. The resulting hypertension drug
Captopril, sold by Bristol-Myers Squibb and launched on the US market in 1981, now has asales revenue
that runs into billions of dollars. However, the Brazilian research institute that originally isolated the
venom did not receive a share ofthis revenue; nor did the region that hosts the snakes. The COP 10
agreement on access and benefit sharing aims to rectify similar cases in the future, by putting in place a
protocol for the sharing of benefits. This is important for biodiversity conservation, as it provides
countries with an incentive to protect biodiversity as a source of information for future drugs. For a more
detailed account, see Patlak (2004).

ban on commercial trade. An Appendix II listing is less restrictive: it only requires a permit
from the exporting country. The CITES Secretariat is also an important source of informa-
tion on the state of endangered species, as it receives annual reports from member states on
the level of trade in listed species (Secretariat of the Convention on Biological Diversity,
2010). The listing is an official declaration that a species is endangered and this may be ben-
eficial in reducing demand or, in the case of an Appendix II listing, may actually increase
speculative demand. For a discussion on the role of trade policy in environmental protec-
tion, see Chapter 8.
Restrictions on trade provide, at best, a sub-optimal approach to wildlife conservation.
By reducing demand, they provide no incentive for landowners or governments to devote
resources to protecting habitats for the species. For instance, banning trade in ivory gives no
incentive to rural populations to protect elephant habitat. However, in some instances, an
Appendix II listing may be more effective in terms of conservation than an Appendix I list-
ing, as there is some incentive in regulated trade to actually protect the habitat (Bulte and
van Kooten, 1999).
Some of these unanticipated perverse incentive effects have started to be addressed by
delegates from developing countries. The 1981 New Delhi meeting ‘downlisted’ some spe-
cies from Appendix I to Appendix II for the purposes of managed exploitation. For instance,
the Zimbabwean population of Nile crocodile was downlisted, so that its commercial util-
ization could be expanded to provide funds for habitat protection. In 1983, this approach
was extended to include trade quotas; for instance, the African leopard was allocated a
quota. This approach was further extended in 1985, when downlisting was allowed on
a range of species so long as a quota on trade was accepted. In 2010, Tanzania and Zambia
requested that trade in ivory from their elephants be allowed (Wasser et al., 2010). For a
discussion of the effectiveness of the ban on the ivory trade, see Box 12.8.

12.4.2 National policies are more challenging

National biodiversity conservation policies are concerned with investing in an asset that
provides a range of private and public goods. One approach to securing this good is through
254

In December 1989, CITES placed the African elephant on the Appendix | list of endangered species. This
decision made trade in ivory with the main consuming nations—the USA, Japan, and the European
Union-illegal. This decision was prompted by a 50 per cent decline in the elephant population between
1979 and 1989, reducing numbers to 609,000. The decline was due to illegal hunting for ivory and a loss
of elephant habitat to agriculture.
Has the ban succeeded in increasing the probability of long-term elephant survival? The population
estimates give a mixed picture. The Zimbabwe population has increased, and this led in 1997 toa
downlisting by the parties of CITES to Appendix II. This allows trade with Japan subject to quotas. The
situation in the other countries, apart from Kenya, is one of further population decline.
How can this be the case when the trade in ivory is banned? The answer is that as with any criminal
activity, the level of poaching effort depends upon the potential cost and benefits. So long as an illegal
world market remains in ivory and some countries are relatively lax in enforcing anti-smuggling and
anti-poaching legislation, then an incentive remains for elephant hunting (Wasser et al., 2010).
Consider the following simple model. The expected profit from poaching is given by

profit = (1—prob)pq— costs — prob fine;

that is, the profit from a poaching trip is equal to the revenue from ivory sales, where p is the price of
ivory and q is the quantity of ivory. The costs are the costs of labour and transport on a poaching trip:
‘prob’ is the probability of getting caught and ‘fine’ is the fine charged for poaching when the poachers
are caught (Burton, 1999).
To reduce the incentive for poaching to zero, the profit must be zero. How can this be achieved? First,
by reducing the effective ivory price by making smuggling more expensive; second, by increasing the
resources devoted to catching poachers, which increases the probability of being caught, ‘prob’; and
third, by increasing the size of the fines charged when a poacher is caught.
Ifa country is not committed to either catching poachers or imposing large fines or custodial sentences
upon them when caught, then it is likely that poaching will remain profitable. One role of international
agencies might be to provide resources to catch poachers. However, a more constructive approach is to
provide local people with benefits for protecting elephants and elephant habitats. If this system works,
then perhaps local people will sort out the poachers themselves!

land purchase and direct management through nature reserves and national parks, but this
is a relatively high-cost means of achieving conservation targets and thus in most countries
a significant area of habitat is on private land. Therefore, governments have taken to con-
tracting with farmers to provide biodiversity conservation on private land.
Investing in biodiversity is complicated by two aspects of what is called asymmetric infor-
mation: first, adverse selection—that is, it is not possible for the regulator to observe the farm-
ers’ costs of undertaking conservation actions so that the landowner can earn so-called
‘information rents’ by the regulator paying the farmer more than it costs; and second, moral
hazard, where farmers may not actually undertake agreed-on conservation actions because the
regulator is unable to observe directly what farmers do. For instance, if biodiversity conserva-
tion requires reduced use of fertilizers to protect water quality, a farmer can state that this has
occurred and receive a compensation payment, but the regulator cannot validate this claim.
The approaches that regulators can take to solve this policy design puzzle are becoming
increasingly sophisticated (Ferraro, 2008). Contracts for conservation actions tend to be
BIODIVERSITY 255

based on inputs or management actions. Thus, if a farmer reduces grazing and fences an
area of native vegetation, he or she receives a payment for changing management actions in
the hope that an ecological improvement (here, an increase in the area ofnative vegetation
conserved) will result. Alternatively, contracts can be based on ecological outcomes. It turns
out that because ecological outcomes such as habitat condition are difficult to define, and
also highly random through time due to climatic and environmental conditions, few con-
tracts are outcome based. This is an aspect of policy that should be developed in the future,
as it gives incentives for landowners to develop ways to make conservation more efficient
(Ferraro, 2008; White and Sadler, 2012).
Contracts can be fixed-price; that is, actions receive a fixed amount. For instance, in the
United Kingdom, the Country Stewardship Scheme (Natural England, 2011) provides
fixed payments to farmers for conservation actions. The scheme is offered in two levels, an
Entry Level and a Higher Level. The advantage of this scheme is that it is relatively straight-
forward to administer, but it will tend to attract farmers with low compliance costs rather
than necessarily the highest marginal gain in biodiversity protection. An alternative is to
use a price-discriminating reverse auction (an auction in which the bidders offer to sell
something to a buyer). An example is the Bushtender auction in Victoria, Australia where
farmers submitted tenders for biodiversity conservation projects on bushland that were
assessed on the basis of an expected increase in a biodiversity metric (an approximate
measure of biodiversity) per dollar of tendered amount (Stoneham et al., 2003). The
advantage of this approach is that by getting farmers to compete for conservation projects,
they can reduce the cost of conservation and make a limited conservation budget go
further. However, the transaction costs will be higher for auctions than for fixed-price
schemes.
US Endangered Species Act. In 1973, the Endangered Species Act (ESA) acknowledged
that the ‘ecological, educational, historical, recreational and scientific value’ of species
diversity was inadequately accounted for in the process of “economic growth and develop-
ment’. The stated purpose of the Endangered Species Act is ‘to provide a means whereby the
ecosystems upon which endangered and threatened species depend may be conserved’. The
Act is administered by the Fish and Wildlife Service and the National Marine Fisheries
Service. Administration involves: first, listing a species as endangered or threatened—
endangered being at greatest risk; second, designating habitats critical to a species’ survival;
third, banning activities that threaten the species; fourth, developing and implementing a
recovery plan; and fifth, removing a species from the list when its population has recovered
sufficiently for it not to be in danger of extinction.
The stated intention ofthe Act is to save all species. Nothing in the original wording ofthe
statute requires that benefits or costs of species extinction should be taken into account.
Although the Act treats all species as equal, budgetary restrictions force the regulatory
authorities to identify priorities and thereby run the risk of some species becoming extinct.
Of the 1,104 species in the USA listed as threatened and endangered, by July 1997 just over
40 per cent had approved recovery plans. The existence of a recovery plan does not guaran-
tee that funds will be available to implement its recommendations. Evidence from Carroll
et al. (1996) and Tear et al. (1993) suggests that survival prospects for just less that 60 per
cent of listed species are actually deteriorating. Further, some approved recovery plans
entail significant risks of extinction.
256 APPLYING THE TOOLS

Schwartz (2008) reviews the more recent evidence on the success of the ESA and is mod-
erately optimistic. On the basis of the US Fisheries and Wildflife Service (USFWS) report-
ing, fourteen species have been recovered and been delisted, while seven have gone extinct
(USFWS, 2008a). A total of twenty species have changed listing status, indicating recovery
(from endangered to threatened) compared to seven shifting towards decline (from threat-
ened to endangered) (USFWS, 2008b). To an economist (Shogren, 2004: ch. 1), the success
of the ESA relates to efficiency: Is the maximum species protection being achieved from the
substantial budget available, US$1.64 billion (as of 2010) (USFWS, 2012)?
The status of US wildlife is further complicated by a lack of information on a list of 3,600
‘indefinite’ species that may be threatened or endangered. The listing process itself is partly
the product ofthe preferences ofthe specialists in the Office of Endangered Species, and this
may account for the high proportion of mammals, birds, and flowering plants listed and the
low proportion of spiders and amphibians. This situation was partially rectified by the 1982
Congress amendment, which required scientific ‘objectivity’. This led to the development of
an eighteen-point scale that included measures of the ‘degree of threat’, ‘recovery potential’,
‘taxonomy ,and ‘conflict with development’. Recovery expenditure (Metrickand Weitzman,
1998) is correlated with the level of conflict between a habitat and development and whether
or not the species is megafauna, such as bears, wolves, and eagles. This tendency is con-
firmed by Ferraro et al. (2007), but their results indicate that over time the scientific basis of
species listing tends to outweigh the charismatic and political motivations for endangered
species listing. These authors find, as their main result, that listing alone reduces the chance
of a species recovering, but listing accompanied by funding increases the probability of
survival. The negative effects of listing alone on species recovery may reflect a perverse
incentive effect. Once listing has occurred or is expected to occur, a landowner may have an
incentive to eradicate an endangered species from his or her land to avoid restrictions on
land use or development.
Economic considerations are accounted for implicitly in species listing, since the role for
economics only emerges at the stage at which a critical habitat is designated. Under a 1978
amendment, the Secretary of the Interior may exclude a critical habitat on cost-benefit
grounds so long as that exclusion does not lead to extinction.
The power ofthe Act rests with its ability to restrict the activities of private parties and
public agencies. Private parties cannot harass, harm, or wound a listed species, where ‘harny’
includes damaging the ecosystem. This is backed by fines of between US$1,000 and
US$50,000, and jail sentences of between 10 and 1,170 days (GAO, 1995). For a fuller
account of the US Endangered Species Act and its performance, see Brown and Shogren
(1998) and Schwartz (2008).
Local policies. International and national conservation policies can be well intentioned,
but if they have little support from local communities they are unlikely to be effective. This
is critical, as the burden ofthe costs of conservation measures tends to fall upon the rural
populations who previously benefited by exploiting the protected resources. In Africa and
other developing countries, these are often amongst the poorest people. By establishing
conservation areas through national parks and the like, we are excluding people from the
resources that had previously sustained them by providing food or income. There is thus a
strong incentive for these people to turn to poaching, or to allow their cattle to encroach on
protected areas.
BIODIVERSITY 257,

Is this outcome inevitable? Drawing on examples from Africa, McNeely (1993) proposes
that the solution to the problem ofproviding biodiversity is to design conservation schemes
which:

e provide incentives for local communities to protect resources;

e avoid perverse incentives that actually encourage people to exploit the resources more
than they would otherwise;
e and establish disincentives that penalize damaging actions.

Let us consider some examples for conservation schemes in Africa.

Kasungu National Park in Malawi provides an example of positive incentives for conser-
vation, where local people have been given the right to collect tree caterpillars and establish
beehives in return for curbing damaging actions such as excessive firewood removal. The
income from these activities, US$198/ha for tree caterpillars and US$230/ha for beekeeping,
provides cash for agricultural inputs, which have allowed some farmers to produce more
cash crops (Mkanda, 1992).
Ecotourism can also provide a strong incentive for conservation. Heal (2000) analyses
ecotourism as a way of bundling a public good with a private good. In this scheme, the tour-
ism revenues are taxed to provide funds to support the public good. Thus in Kenya, tourism
revenues are used to support the Kenyan Wildlife Service.
Unfortunately, examples of perverse incentives are also common. Often, these represent
government failure, where natural resource or agricultural policies give an incentive to
degrade the environment. For instance, in Botswana, prices for export beef supported by the
European Union combined with subsidized inputs including fencing, veterinary services,
and borehole development have led to widespread overgrazing, which benefits a small num-
ber of ranchers (Perrings and Stern, 2000).
Incentives need to be reinforced by disincentives that discourage damaging actions.
However sympathetic a conservation scheme is to the economic well-being of the local
community, there is almost always an incentive for people to misappropriate open-access or
common-property resources, to catch more fish than is their right, to poach, or to allow
their cattle to overgraze. Hannah (1992) found that the stronger the enforcement proce-
dures in place, then the more effective were conservation projects in relation to their objec-
tives. Enforcement may be in terms of conventional monitoring backed up by legal action.
Misappropriation of resources may also be reduced by community opprobrium. If a com-
munity has a sense of shared ownership over a resource, then the community itself is more
likely to protect it.

Summary
The earth is currently undergoing a phase of mass species extinction. Unlike previous mass
extinctions, this current episode is due to habitat loss and disruption. By appropriating rich
and diverse habitats for agricultural production and other disruptive forms of land use,
humankind determines which species survive, which thrive, and which are pushed to
extinction. In taking these decisions, we are determining which species survive to provide
258 APPLYING THE TOOLS

future generations with new drugs, crops, genetic material, aesthetic values, and ecosystem
services. These decisions are often taken without any sense of their gravity; instead, we often
blunder on in profound ignorance of which species are being destroyed in terms of their
genetic make-up, their form, their behaviour, and their role in the ecosystem in which they
are embedded.
The economics of biodiversity should attempt to value biodiversity in all its complexity,
including the possibility that a species contains information that is irreversibly lost by
extinction. Unfortunately, we are a long way from being able to do this. We can assess the
value of biodiversity in terms of its potential medicinal value. We can value some ecosystem
services, especially where those are associated with market values. However, we have a
problem in valuing a species or a group of species in terms of their role within a complex
ecosystem. The other challenge for economists is to design policies that present landowners
and land users with an incentive to protect valued species and habitats cost-effectively.

Tutorial Questions
12.1 In your own words, define biodiversity.
12.2 Why should society conserve biodiversity?
12.3, How do we decide which species to save and which to leave unprotected?
12.4 ‘Sometimes biodiversity conservation policies can make things worse.’ Discuss this
statement in relation to a specific policy.
12.5 Define what might represent a cost-effective conservation policy.

References and Additional Readings

Aylward, B.A., Echeverria, J., Fendt, L., and Park, Costa Rica’ Forest Policy and Economics 11:
Barbier, E.B. (1993). The Economic Value of 174-83.
Species Information and its Role in Biodiversity
Brander, J.A., and Taylor, M.S. (1998). “The simple
Conservation: Case Studies of Costa Rica’s
economics of Easter Island: a Ricardo—Malthus
National Biodiversity Institute ofPharmaceutical
model of renewable resource use, American
Prospecting. Report to the Swedish International
Economic Review 88: 119-38.
Development Authority (London: Environmental
Economics Centre), Brown, G.M., and Shogren, J.F. (1998). ‘Economics
of the Endangered Species Act) Journal of
Barbier, E.B., and Strand, I. (1998). ‘Valuing
Economic Perspectives 12: 3-20.
mangrove-fishery linkages—a case study of
Campeche, Mexico, Environmental and Resource Burton, M. (1999). ‘An assessment of alternative
Economics 12(2): 151-66. methods of estimating the effect of the ivory trade
ics
Bateman, I.J. Mace, G.M., Fezzi, : ban on poaching effort’ Ecological Ec ics 30:
C., Atkinson, G., 93 me caching effort, Ecological Economics 30:
and Turner, K. (2011). ‘Economic analysis for sat a
ecosystem service assessments, Environmental Butchart, S.H.M. et al. (2010). ‘Global biodiversity:
and Resource Economics 48: 177-218. indicators of recent declines’ Science 328:
Bernard, F., de Groot, R.S.,' and Campos, J.J. 1164-8.
—
(2009). “Valuation of tropical forest services and Bulte, E., and van Kooten, C. (1999). ‘Marginal
mechanisms to finance their conservation and valuation of charismatic species, Environmental
sustainable use: a case study of Tapanti National and Resource Economics 14; 119-30.
BIODIVERSITY 259

Carroll, R., Augspurger, C., Dobson, A., Franklin, J., methods, Journal of Environmental Economics
Orians, G., Reid, W., Tracy, R., Willcove, D., and and Management 54: 245-61.
Wilson, J. (1996). ‘Strengthening the use of science
—— and Pattanayak, S.K. (2006). ‘Money for
in achieving the goals of the Endangered Species
nothing? A call for empirical evaluation of
Act: an assessment by the Ecological Society of
biodiversity conservation investments, PLoS
America, Ecological Applications 6: 1-11.
Biology 4: 482-8.
Chichilnisky, G. (1993). ‘Property rights and
Finnoff, D., and Tschirhart, J. (2008). ‘Linking
biodiversity and the pharmuceutical industry,
dynamic economic and ecological general
Working Paper, Columbia University Graduate
equilibrium models, Resource and Energy
Business School, New York.
Economics 30: 91-114.
Christie, M., Hanley, N., Warren, J., Murphy, K., Gallai, N., Salles, J-M., Settele, J., and Vaissiére,
Wright, R., and Hyde, T. (2006). ‘Valuing the
B.E. (2009). “Economic valuation ofthe
diversity of biodiversity, Ecological Economics vulnerability of world agriculture confronted
58(2): 304-17.
with pollinator decline’, Ecological Economics
Claassen, R., Cattaneo, A., and Johansson, R. (2008). 68(3): 810-21.
‘Cost-effective design of agri-environmental GAO (General Accounting Office) (1995).
payment programs: U.S. experience in theory and Endangered Species Act: Information on Species
practice, Ecological Economics 65: 737-52. Protection on Non-federal Lands, GAO/
Colwell, R.R. (1996). ‘Microbial biodiversity and RCED-95-16 (Washington, DC: USGAO).
biotechnology; in M.J. Reaka-Kudla, D.E. Hammer, K., and Khoshbakht, K. (2005). “Towards
Wilson, and E.O. Wilson (eds.), Biodiversity II
a “Red List” for crop plant species, Genetic
(Washington, DC: James Henry Press). Resources and Crop Evolution 52: 249-65.
Convention on Biological Diversity (1992). http:// Hanley, N., Spash, C., and Walker, L. (1995).
[Link]/ (accessed 31 July 2011). ‘Problems in valuing the benefits of biodiversity
COP (1998). COP4 Report, [Link] protection, Environmental and Resource
(accessed 15 November 2012). Economics 5(3): 249-72.
Czajkowski, M., Buszko-Briggs, M., and Hanley, N. Hannah, L. (1992). African People, African Parks
(2009). “Valuing changes in forest biodiversity, (Washington, DC: Conservation International).
Ecological Economics 68: 2910-17. Harrop, S.R. (2011). ‘ “Living in harmony with
Daily, G. (ed.) (1997). Nature’ Services: Societal nature”? Outcomes of the 2010 Nagoya
Dependence on Natural Ecosystems (Washington, Conference of the Convention on Biological
DC: Island Press). Diversity, Journal of Environmental Law 23(1):
117-28.
Elridge, N. (1998). Life in the Balance: Humanity and
the Biodiversity Crisis (Princeton, NJ: Princeton Heal, G.M. (2000). Nature and the Marketplace:
University Press). Capturing the Value of Ecosystem Services
Feeley, K.J., and Silman, M.R. (2009). ‘Extinction (Washington, DC; Island Press).
risks of Amazonian plant species, Proceedings of Iltis, H.H., Doebley, J.E, Guzman, R., and Pazy, B.
the National Academy of Sciences of the United (1979). Zea diploperennis (Gramineae)—new
States ofAmerica 106(30): 12382-7. teosinte from Mexico, Science 203: 186-8.
Fernandes, L., Day, J., Lewis, A., Slegers, S., Kerrigan, IUCN (International Union for Conservation of
B., Breen, D., Cameron, D.F., Jago, B., Hall, J., Nature) (2010). IUCN Red List of threatened
Lowe, D., Innes, J., Tanzer, J., Chadwick, V., species, Version 2010, IUCN, Gland, Switzerland,
Thompson, L., Gorman, K., and Possingham, [Link] (accessed 10 June
H. (2005). ‘Establishing representative no-take 2010).
areas in the Great Barrier Reef: large scale Jacobsen, J.B., Boiesen, J.H., Thorsen, B.J., and
implementation of theory on marine protected Strange, N. (2008). “What's in a name? The use of
areas, Conservation Biology 19(6): 1733-44. quantitative measures versus “iconised” species
Ferraro, PJ. (2008). ‘Asymmetric information and when valuing biodiversity, Environmental and
contract design for payments for environmental Resource Economics 39(3): 247-63.
services, Ecological Economics 654: 810-21. MacArthur, R.H., and Wilson, E.O. (1967). The
— McIntosh, C., and Ospina, M. (2007). “The Theory of Island Biogeography (Princeton, N):
effectiveness of the US Endangered Species Princeton University Press).
Act: an econometric analysis using matching McNeely, J.A. (1993). ‘Economic incentives for
260 APPLYING THE TOOLS

conserving biodiversity—lessons for Africa, Putman, R.J., and Wratten, S.D. (1984). Principles of
Ambio 22(2-3): 144-50. Ecology (London: Croom Helm).
Margules, C.R., and Pressey, R.L. (2000). ‘Systematic Raup, D. (1988). ‘Diversity crises in the geological
conservation planning’ Nature 405: 243-53. past, in E.O. Wilson (ed.), Biodiversity
May, R.M., Lawton, J.H., and Stork, N.E. (1995).
(Washington, DC: National Academy Press).
‘Assessing extinction rates’ in J. H. Lawton and R. Richardson, L., and Loomis, J. (2009). “The total
M. May (eds.), Extinction Rates (Oxford: Oxford economic value of threatened, endangered and
‘University Press). rare species: an updated meta-analysis, Ecological
Economics 68(5): 1535-48.
MEA (Millennium Ecosystem Assessment) (2005).
Ecosystems and Human Well-being: General Schaefer, M.B. (1957). ‘Some consideration of
Synthesis (Washington, DC: Island Press). population dynamics and economics in relation to
Metrick, A., and Weitzman, MLL. (1998). ‘Conflicts
the management of marine fisheries, Journal of the
and choices in biodiversity preservation, Journal Fisheries Research Board of Canada 14: 669-81.
of Economic Perspectives 12: 21-34. Schwartz, M.W. (2008). “The performance of the
Mkanda, C.X. (1992). “The potential of Kasungu Endangered Species Act, Annual Review of
National Park in Malawi to increase income and Ecology, Evolution, and Systematics 39:
food security in neighbouring communities, 279-99.
Paper presented at IV World Congress on Secretariat of the Convention on Biological Diversity
National Parks and Protected Areas, Caracas, (2010). Global Biodiversity Outlook 3 (Montreal).
Venezuela. Shogren, J.F. (ed.) (2004). Species at Risk: Using
Moony, H., and Mace, G. (2009). ‘Biodiversity policy Economic Incentives to Shelter Endangered Species
challenges: editorial, Science 325: 1474. on Private Property (Austin, TX: University of
Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Texas Press).
Fonesca, G.A.B., and Kent, J. (2000). “Biodiversity — Tschirhart, J., Anderson, T., Whritenour Ando,
hotspots for conservation priorities, Nature 403: A., Beissinger, $.R., Brookshire, D., Brown,
853-58. G.M. Jr, Coursey, D., Innes. R., Meyer, $.M., and
Natural England (2011). Environmental stewardship, Polasky, S. (1999). ‘Why economics matters for
[Link] endangered species protection’, Conservation
farming/funding/es/[Link] (accessed 3 June Biology 13(6): 1257-61.
2011). Simpson, R.D., Sedjo, R.A., and Reid, J.W. (1996).
Opdam, P., and Wascher, D. (2004). ‘Climate ‘Valuing biodiversity for use in pharmaceutical
change meets habitat fragmentation: linking research, Journal ofPolitical Economy 104:
landscape and biogeographical scale levels 163-85.

in research and conservation, Biological Spash, C., and Hanley, N. (1995). ‘Preferences,
Conservation 117: 285-97. information and biodiversity preservation,
Patlak, M. (2004). ‘From viper’s venom to drug Ecological Economics 12: 91-208.
design: treating hypertension, The FASEB Journal —— and Simpson, I. (1994). “Utilitarian and
8: 421-34.
rights-based alternatives for protecting sites of
Perrings, C., and Stern, D.1. (2000). ‘Modelling special scientific interest, Journal ofAgricultural
loss of resilience in agroecosystems: rangelands Economics 45(1): 15-26.
in Botswana, Environmental and Resource Stoneham, G., Chaudhri, V., Ha, A., and
Economics 16; 185-210. Strappazzon, L. (2003). ‘Auctions for
Polasky, S., Camm, J.D., and Garber-Yonts, B. conservation contracts: an empirical
(2001). ‘An application to terrestrial vertebrate examination of Victoria’s BushTender trial’
conservation in Oregon, Land Economics 77: 68-78. Australian Journal ofAgricultural and Resource
Possingham, H.P., and Wilson, K.A. (2006). Economics 47: 477-500.
‘Biodiversity: turning up the heat on hotspots, Stork, N.E. (1996). ‘Measuring global biodiversity
Nature 436: 919-20. and its decline} in M.J. Reaka-Kudla, D.E.
Prach, K., and Hobbs, R.J. (2008). ‘Spontaneous Wilson, and E.O, Wilson (eds.), Biodiversity II
succession versus technical reclamation in the (Washington, DC: James Henry Press).
restoration of disturbed sites, Restoration Ecology — (2010). ‘Re-assessing current extinction rates,
16: 363-6. Conservation Biology 19: 357-71.
BIODIVERSITY 261

Swanson, T.M. (1994). The International Regulation UNEP (United Nations Environment Program)
ofExtinction (Basingstoke: Macmillan). (1993). Guidelines
forCountry Studies on
Taylor, S.M. (2009). ‘Environmental crises: past, Biological Diversity (Nairobi: UNEP).
present and future, Canadian Journal of — (1998). Conference ofthe Parties to the
Economics 42: 1240-75. Convention on Biological Diversity, Fourth
Tear, T., Scott, J., Hayward, P., and Griffith, B. Meeting, Bratislava.
(1993). ‘Status and prospects for success of the Wasser, S. et al. (2010). “Elephants, ivory and trade,
Endangered Species Act: a look at recovery plans, Science 327: 1331-2.
Science 262: 976-7. USFWS (US Fishery and Wildlife Service) (2008a).
TEEB (The Economics of Ecosystems and Conservation Plans and Agreements Database
Biodiversity) (2010). The Economics of Ecosystems (Washington, DC: USFWS).
and Biodiversity: Mainstreaming the Economics of — (2008b). Summary Reports to Congress on the
Nature—A Synthesis of the Approach, Conclusions Recovery Program for Threatened and Endangered
and Recommendations of TEEB, [Link] Species (Washington, DC: USFWS).
[Link]/ecological-and-economic-
— (2012). ‘Budget at a glance’ [Link]
foundations-report/ (accessed 15 November
gov/budget/2012/PDF%20Files%20FY2012%20
2012).
Greenbook/04.%20Budget%20at%20a%20
Tschirhart, J., Anderson, T., Whritenour Ando, A., Glance%[Link] (accessed 12 May 2012).
Beissinger, S.R., Brookshire, D., Brown, G.M.
Weitzman, M.L. (1992). ‘On diversity, Quarterly
Jr, Coursey, D., Innes. R., Meyer, S.M., and
Journal of Economics 108: 157-83.
Polasky, S. (1999). ‘Why economics matters for
endangered species protection, Conservation White, B., and Sadler R.J. (2012). ‘Optimal
Biology 13(6): 1257-61. conservation investment for a biodiversity rich
agricultural landscape’ Australian Journal of
Tschirhart, J. (2004). ‘Integrated ecological-,
Agricultural Economics 56: 1-21.
economic models, Annual Review of Resource
Economics 1: 381-407. Wilson, E.O. (1988). “The current state of biological
diversity, in E.O. Wilson and EM. Peter (eds.),
UK National Ecosystems Assessment (2010). http://
Biodiversity (Washington, DC: National
[Link]/ (accessed 3 August
Academy Press).
2011).
Non-renewable Natural
Resources and Energy

Economists have long been concerned with the efficient use of scarce natural resources.
Adam Smith examined the nature of capital for land, mines, and fisheries; Ricardo explored
how land quality matters for economic rent; Malthus worried about population, poverty,
and the limits of agricultural resources; and Jevons feared the social consequences of the
depletion of coal quantity and quality. These classical economists treated natural resources
as a factor of production provided freely by nature, which made it distinct from costly capi-
tal and labour. The general mindset framed the problem as one in which a resource owner
made choices to maximize the net present value of the natural resource.
At the start of the twentieth century, economics started to treat natural resources as
something more distinct than just as a free factor of production. Theorists such as Gray and
Hotelling argued that an additional intertemporal cost for extracting or harvesting natural
resources existed. They argued that a resource owner should account for an additional cost
above and beyond the cost of extraction and processing—the opportunity cost of depletion
or harvesting sooner rather than later. After the Second World War, fishery economists
explained how weakly defined property rights can lead people to overexploit resources that
inhabit the commons. In the late 1970s and early 1980s, economics began to examine the
social inefficiencies associated with stock pollutants such as carbon emissions and climate
change, the loss of services from reductions in the stock of global biodiversity, and the risks
to life support and aesthetic services provided by natural resources left unpriced by the
market.
Today, natural resource economics continues to turn these early and new insights into
mature theories that can help explain how people and societies choose to manage and use
their limited resources, both non-renewable resources such as minerals and fossil fuels, and
renewable resources such as fisheries (Chapter 7) and forests (Chapter 10). The field consid-
ers how societies makes choices to (mis)manage their stocks of biological diversity cost-
effectively, to reduce risks from climate change efficiently, and to value natural resource
services that are not bought and sold in the marketplace.
This chapter examines how people develop and conserve their scarce non-renewable
resources and renewable energy resources; and how these resources are transformed into
energy. Many questions arise when thinking about natural resources and our usage: Are we
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 263

running out of resources? Do resources get depleted too fast as countries try to maximize
their profits too quickly? What are the predictions for the future of world energy demand
and supply? In this chapter, we will examine these questions by:
e Discussing what is meant by ‘resources’.
e Explaining an economic model of how resource extraction should be managed.
e Evaluating alternative measures of resource scarcity.
e Examining the special role of energy in the economy.
e Discussing what a nation’s energy policy has to address.

13.1 Natural Resource Types

What do we mean by ‘natural resources’? A distinction is often made between ‘material’ and
‘energy resources. This distinction relates to the conventional end-uses of these resources,
in that material (or mineral)! resources are used as part of the physical constituency of com-
modities (iron ore, converted into steel, in car bodies, or copper in pipes); whilst energy
resources are converted into heat and other forms of energy. The chemical energy in natural
gas is converted into heat energy when gas is burnt in domestic central-heating boilers.
Some resources are used both as materials and as energy sources: oil is the prime example
here, being used for propulsion ‘in internal combustion engines, and to make plastics.
Conversion of material resources into usable forms also requires inputs of energy resources
(for smelting and mining). There are some eighty-eight minerals occurring on earth. Of
these, only twelve make up 99 per cent of the earth’s crust: the most common elements in
these minerals are silicon (27%), aluminium (8%), and iron (6%).
One obvious distinction between resource types is by their potential for natural growth.
A fishery is different from a deposit of iron ore; fish have a natural rate of growth, iron ore
does not. Most economists distinguish between ‘renewable’ and ‘non-renewable’ resources,
with the renewable classification reserved for those resources exhibiting a positive natural
rate of growth. This is a cleaner distinction than the classification into ‘exhaustible’ and
‘non-exhaustible’ resources, since even a renewable resource can be exhausted (by continu-
ing to harvest in excess of the natural rate of growth, for example), and a non-renewable
resource may not be exhausted if it becomes uneconomic to extract.

13.2 The Extraction of a Non-renewable Resource

In this section, we focus on how economists think about using non-renewable resources
such as oil and coal. We focus on the classic Hotelling model of resource extraction as our
benchmark case. We next consider whether the owner of the resource has a monopoly on the
resource, and finally we examine how economic behaviour changes once we add extraction
costs.

1 A mineral is defined as a solid crystalline chemical element or compound in fixed composition. Rocks are
aggregates of one or more minerals. A mineral deposit is an accumulation of aspecific mineral.
264 APPLYING THE TOOLS

13.2.1 The baseline case—an introduction to Hotelling’s rule

Consider an oil market made up of a large number of firms, each one owning its own oil
well. All firms want to maximize the present value of their profits, where profit is the price of
oil times the quantity extracted. We assume, for the time being, that oil can be produced at
zero cost. The problem that the firms face is how much oil to supply in each time period. The
use of a present value of profit implies that delaying oil extraction has an opportunity cost
in terms of the return (r) that money tied up in oil reserves could earn from alternative
investments (such as money in interest earning deposits in a bank or invested in stocks and
shares). So why do these firms not empty their oil wells in the first year and live off their
investments?
A similar question led Hotelling (1931) to develop a model of how non-renewable natural
resources are extracted through time. The key result is now known as Hotelling’s rule. The
general version of the rule is complex: see Hanley et al. (2007: ch. 7) for a fuller introduction,
but the basic idea can be explained in a simple two-period example. We define an equilib-
rium in which the producer is indifferent between selling the last unit of oil in the current
period or in the next period. For this to be the case, the present value of a barrel of oil would
have to be the same in both periods. If the price in period 1 is given as p, and the price in
period 2 is given by p,, then the equilibrium condition is given by

pP, = p.d/d+r)), (13.1)

where (1/(1 + r)) is the discount factor. What does this condition imply? Suppose the dis-
count rate is 10 per cent (r= 0.1); this implies (by rearranging this formula) that

(l+r)p, = py

where the price in period 2 must be 10 per cent greater than the price in period 1 for the
firm to be in equilibrium. Another way of representing this formula is by rearranging to
give:

poset
varie i (13.2)

which states that the proportional price rise for oil equals the discount rate. Hotelling’s rule
predicts that the oil price will rise through time. The next question is what actually drives the
price increase? The answer is market demand combined with all firms reducing their output
through time. In aggregate, firms extract at a rate in each period that ensures that the equi-
librium shown in equation (13.2) holds.
We now have a basic condition to determine the price and the quantity of oil produced in
each period. This is not enough, however, to determine the life cycle ofoil reserves, since we
also need to know the initial stock of oil and the price at which demand falls to zero. The
price at which demand falls to zero is known as the ‘backstop price’ and it can be interpreted
as the price of asubstitute for the non-renewable resource. For instance, in the case ofoil it
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 265

might be the price at which demand switches to a renewable alternative such as fuel made
from ethanol.
Once we know the backstop price and the initial stock, on the basis of Hotelling’s rule,
we can calculate the price and quantity through time. This can be represented in a four-
quadrant diagram (Figure 13.1). The north-west quadrant of Figure 13.1 gives the demand
curve, and the north-east quadrant the extraction path. Note that through time the quan-
tity extracted falls and reaches zero at the end time T when the resource is exhausted. The
area under the extraction curve is the total initial stock, which is the sum of all extraction
through time. As the rate of extraction falls then the price, in the south-west quadrant, rises
at a rate that satisfies Hotelling’s rule, until at time T it reaches the backstop price p>, when
demand falls to zero. The south-east quadrant contains a 45-degree line that simply trans-
fers time from one axis to the other.
Is this resource extraction plan socially optimal? The discount rate reflects the current
generation's preferences for resource use through time. If this is accepted, so long as the
discount rate is the social time preference rate, then the resource extraction path is at least
efficient. But if firms face imperfect capital markets and interest rates, or returns on other
assets are distorted, this may lead firms to extract the non-renewable resource either too
rapidly or too slowly. In many countries, the interest rate tends to be higher than the social
discount rate due to the inclusion of a premium to cover risk. This implies that firms may be
inclined to extract a non-renewable resource too rapidly. This is predicted by Hotelling’s
rule: high discount rates lead to more rapid rises in the price and a more rapid decline in the
rate of extraction and a shorter time to economic exhaustion of the resource.

Demand curve Extraction path

Price path

Figure 13.1 The Hotelling model of resource extraction.

266 APPLYING THE TOOLS

13.2.2 Resource extraction and the monopolist

In contrast to the competitive market example in the previous section, the structure of many
non-renewable resource industries is highly concentrated in one or two countries. For
instance, as an oil cartel, the Organization of Petroleum Exporting Countries (OPEC) rep-
resents a significant proportion of world oil production, and the same holds for South Africa
and diamonds. An interesting question is: How would a monopoly decide to extract a non-
renewable resource? The answer is that the profit-maximizing zero-cost monopoly follows
a slightly different version of Hotelling’s rule. Instead of equating the present value of price
through time, it equates the present value of the marginal revenue. Remember that marginal
revenue is the change in total revenue when we increase output by one unit:

MR, — MR, =a (13.3)

MR,
The implication of this rule is that if the marginal revenue increases at the discount rate,
then this implies that the monopoly price starts higher than the competitive price, but rises
more gradually through time. Remember that a higher price implies a lower level of demand.
In practice, the monopoly acts as a price discriminator where the market is separated
through time. The implication of this rule is that a monopoly extracts relatively slowly com-
pared with firms in greater competition with each other. In fact, we may view the activities
of members of OPEC as being ‘conservationist’ in the sense that by acting as a cartel they
reduce the supply of oil and extend the life of the resource.

13.2.3 Model including extraction costs

Hotelling’s rule has been given for a zero-cost firm, which is unrealistic. The model can be
easily modified to include costs. Note that in the zero-cost case, the price can be interpreted
as the marginal profit—the addition to profits for each extra unit extracted. If we introduce
costs, then the marginal profit becomes the price minus the marginal cost, given as Mz =
p— MC. Hotelling’s rule with costs is then given as follows:

Mz, — Mz,
Mn,

Instead of the price increasing by the discount rate, the marginal profit increases by the
discount rate. This means that the rate of increase in prices now depends partly on how costs
change over time. Costs for non-renewable resource firms are typically given as an increas-
ing function ofthe rate of extraction (q) and a declining function of the remaining reserves
(x), c= c(x, q). This form ofcost function explains the physical reality of many mineral and
petroleum reserves, that extraction tends to move to progressively more costly deposits as
reserves are depleted. For instance, mines move further underground and this tends to
increase costs. Oil extraction tends to reduce the pressure in oil reservoirs and this increases
the cost of production (see Box 13.1). The marginal profit (Mz =p — MC) also gives a meas-
ure of resource scarcity, in that it is the value of the marginal unit ofthe resource stock. We
NON-RENEWABLE NATURAL RESOURCES AND E 267

BOX 13.1 Testing Hotellinge Rule

Hotelling’s rule is remarkable as an example of adynamic model that is able to trace out the life cycle ofa
non-renewable resource industry from the current time to the time when the last drop of oil worth
extracting is extracted. As a model, it has two possible applications. It can be used prescriptively to inform
non-renewable firms how they should schedule resource extraction through time, or it can be used to
predict how firms actually behave. In the first of these roles, the model might be useful as a guide to
extraction rates, so long as the firm can predict how the oil price changes through time. However, even a
monopoly firm faces a demand curve that changes through time due to changing tastes, technology, and
general economic conditions. Many authors have tested how good Hotelling’s rule is at predicting the
behaviour of firms (for a review, see Hanley et al., 2007: ch. 8). The overall conclusion is that the standard
form of Hotelling’s is not good at explaining how firms actually behave. Why should this be?
In a review paper, Livernois (2009) concludes that scarcity by itself is not sufficient to drive prices
upwards through time: ‘It appears that other factors, notably technological change, revisions to
expectations regarding the resource base, and market structure, have had a more significant influence on
the evolution of price.’ A recent analysis of trends in non-renewable resource prices (Lee et al., 2006) for
eleven major commodities concludes that for most series there is no clear upward trend; in fact, the trend
from 1870 to 1990 includes periods of rising, falling, and constant real prices, with a number of structural
breaks. The results confirm the findings of Krautkraemer’s review (1998). It seems that if we want to
understand the workings of non-renewable resource markets, we need to understand demand and
supply responses through time and the critical role of price expectations (Pesaran, 1990).

can see that scarcity increases as the price increases, but falls as the marginal cost increases.
Measures of scarcity are discussed in the next section.

13.3 Measuring Resource Scarcity

One of the most common questions in debates over the use of natural resources is ‘are we
running out?’ For any non-renewable resource, a positive rate of extraction means that the
physical stock of the resource is reduced in size. However: (i) there are major problems in
defining what this physical stock represents; (ii) the economic measure of the size of reserves
is not the same as the physical size ofthe reserves; (iii) the value of the economic reserve will
change over time; and (iv) there are alternative measures for the scarcity of this economic
reserve, which may well give different answers to the above question.
Consider Figure 13.2. The figure shows a more realistic representation of anon-renewable
resource through time in terms of marginal costs and prices. This diagram can be considered
as a modified version of the south-west quadrant of Figure 13.1. At any one time, there is a
range of marginal costs of extraction for a resource such as oil. The International Energy
Agency (2008) analysed the global oil supply curve on the basis of production costs per barrel:
the lowest-cost reserves are found in the Middle Eastern and North African oilfields, at $6 to
$28 per barrel; the most expensive are oil from oil shales, at between $58 and $113 per barrel.
In Figure 13.2, the upper and lower limits of the distribution of marginal costs change
through time. The lower marginal cost rises due to extraction, but the rate of rise may be
reduced by new technology that makes extraction cheaper. The higher marginal cost falls
268 APPLYING THE TOOLS

Marginal
cost Marginal cost highest
price per
unit output $
ie Backstop price

Py 5 ee A= ak eek ae ee er

Marginal cost lowest

Time if

Figure 13.2 Resource extraction, technological innovation, and exploration through time.
Note: The ‘marginal cost lowest’ is the expected lowest marginal cost of extracting from the known proven resource stocks;
the ‘marginal cost highest’ is the expected marginal cost from extracting from possible resource stocks; the price is the
expected market price and the backstop price gives the price of alternate technology.

due to new technology that makes, for instance, extraction of oil from remote areas less
expensive. Note that most expensive reserves (highest marginal cost) are not extracted until
the end of the resources’ life.
The backstop price generally falls through time due to new technology. Non-renewable
resource firms may anticipate this trend and extract more rapidly.
The term “economic reserves’ is often used to describe that portion of a deposit (or collec-
tion of deposits) that it is profitable to extract, given current prices and costs. Costs depend
partly on the state of technology, and on cumulative extraction: these costs change over
time. Prices will also change in response to the decisions of extractors over extraction rates
(which might depend, for example, on the agreement reached by a cartel of producers, such
as OPEC), demand for the material, and government intervention on prices. In Figure 13.2,
the economic reserves are those reserves with a marginal cost less than the price and above
the lowest marginal cost.
The diagram shows that reserves in the economic reserve change through time. For
example, the minimum concentration of copper in a copper deposit required for profitable
extraction fell from 3 per cent in the 1800s to 0.5 per cent in the mid-1960s with technologi-
cal progress, which at constant real prices would result in the size of the economic reserve
increasing over time.
Uncertainty also exists over the actual amount of a resource in a given geographical area.
For example, it is not known with certainty how much oil lies under the North Sea. Some oil
deposits have been found and are in production; others have been found and are not in
production. Other deposits are thought to exist given the nature of the surrounding geol-
ogy. But the total size of deposits may be greater than this. But even with respect to defining
the physical size of adeposit, or of all deposits for a particular material, difficulties arise. For
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 269

example, should all amounts of copper be counted, irrespective of their concentration, or of

the form in which they are present?
Several writers argue for a crucial concept called the ‘mineralogical threshold’ (e.g. Harris
and Skinner, 1982). Below this threshold, minerals occur as silicates, in that they are chemi-
cally bonded to silica. The total amount of a mineral that exists on earth is known as its
‘crustal abundance’, also referred to as the resource base. However, only a small fraction
(roughly 3% on average) of the crustal abundance of most minerals exists in non-silicate
form, as oxides, sulphides, or carbonates. For some minerals, few deposits exist that are in
non-silicate form. Skinner (1976) calls these minerals ‘geochemically scarce’.
To take the example oflead, its average concentration in the earth’s crust is 0.001 per cent,
but extraction currently takes place in ore deposits where lead is found in concentrations of
between 2 and 20 per cent. Once these ore deposits have been extracted, then vastly more
energy would be required to extract lead from its silicate form, where it is trapped by atomic
substitution. The extraction of the deposits would also produce large quantities of geochem-
ically abundant minerals as a by-product. Geochemically scarce minerals include copper,
lead, mercury, and gold (Anderson, 1985). For geochemically abundant minerals, such as
iron, the energy required to extract the mineral increases smoothly as the purity of the ore
declines, as Figure 13.3 shows; for geologically scarce minerals, energy use jumps at the
mineralogical threshold. A prediction from the mineralogical threshold model is that geo-
chemically abundant minerals will be substituted for geochemically scarce minerals as the
threshold is approached.
Table 13.1 gives recent estimates of world total reserves, the countries with the largest
deposits, the number of years of consumption remaining (at current consumption levels)
(Gordon et al., 2006; Cohen, 2007), and also the percentage of consumption met from

used
recovered
mineral
of
Energy
unit
per

Mineralogical Ore grade (%)

threshold

A. Geochemically abundant minerals

B. Geochemically scarce minerals

Figure 13.3 The mineralogical threshold.

270 APPLYING THE TOOLS

Table 13.1 Abundance, regional concentration, and recycling for metallic minerals
eo ieAbySei ete aa ee eee AS ass i ee ee eS SS
Uses Stock Units Country with Years reserves Percentage
largest reserves remaining atcurrent supply from
(%) consumption recycling

Aluminium Cans, white goods 32,350 mt Guinea (27) 1,027 49

Antimony Drugs 3.86 mt China(62) — 30 0
Chromium Chrome plating VS mt Kazakhstan (60) 143 2S
Copper Wire, coins, 937 mt Zambia (38) 61 31
plumbing
Gold Dentistry, jewellery 89,700 t South Africa (40) 45 43
Indium LCDs 6,000 t Canada (33) 18 0
Lead Pipes, batteries 144 mt China (25) 42 We
Nickel Batteries, turbine 143 mt Australia (19) 90 35
blades
Phosphorous — Fertilizer 49,750 mt Morocco 345 0
and Western
Sahara (42)
Platinum/ Jewellery, catalytic 79,840 t South Africa (88) 360 0
rhodium converters
Silver Jewellery, catalytic 569,000 t Poland (25) 29 16
converters
Tantalum Cell phones, camera 153,000 t Brazil (48) 116 20
lenses
Tin Tins, solder We mt China(31) 40 26
Uranium Power generation 33 mt Australia (23) 59 0
Zinc Galvanizing 460 mt USA (20) and 46 26
China (20)

Source: Cohen (2007).

recycling. The minerals contain some contrasting examples; for instance, the world is
unlikely to ever exhaust its aluminium reserves. We have over 1,000 years’ reserves ofalu-
minium, and 49 per cent of consumption comes from recycling. Other metals are more of
a concern. For instance, one high-tech metal—indium, used in LCDs—has limited known
reserves and is expected to only last another 13 years. Similarly, copper has 61 years of
reserves remaining. The question is: Will impending resource depletion seriously affect
economic growth and welfare, or will Hotelling’s rule lead to an increase in prices and a
drive to find substitutes? Economists tend to be more optimistic on this issue than
geologists.

13.3.1 Resource lifetime

A frequently cited measure of resource scarcity is the lifetime ofaresource. This is usually
expressed as the economic reserve of aresource divided by its current annual consumption
rate, with perhaps an allowance for a predicted growth in this rate over time. Fisher (1981)
quotes (but does not endorse) a measure of 45 years for copper in 1974: a prediction that by
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 271

the year 2019, the world will run out of copper. The most immediate problem here is that if
we instead divided the total resource base by annual consumption, we would arrive at a
much larger figure, one that allows for higher-cost deposits being brought on line as prices
rise—but which measure is correct?
The answer is neither. As a resource gets scarcer, its price will, other things being equal,
tend to rise. This will reduce consumption (by substitution, for example), and increase pro-
duction. These changes will change the lifetime measure. What is more, as prices rise pro-
ducers will be encouraged to engage in more exploration, which will increase the resource
base if finds are made. Lifetime measures for many resources have been found to be approx-
imately constant over time, and have been argued by Fisher to say more about firms’ atti-
tudes to holding inventories of minerals than about scarcity.

13.3.2 Unit cost measures

The earliest arguments in natural resources economics about scarcity centred around the
costs of extraction. As a mine is depleted, the miners have to travel further and further
underground to recover ores or coal, causing labour costs per unit of output to rise. As a
country mines its copper, it has to move on to less and less pure grades of ore. Cumulative
production increases average costs, which are an indicator of scarcity.
In the 1960s, Barnett and Morse (1963) studied trends in average costs over the time
period 1870-1957 for a variety of primary products. With one exception (forestry), they
found that an index of real unit (capital plus labour) costs had declined over the period,
indicating decreasing scarcity: real capital-plus-labour inputs had declined by 78 per cent
for the minerals sector and by 55 per cent for the total extractive sector. Barnett and Morse’s
work was repeated by Johnson et al. (1980), who found that, if anything, the rate of decline
in unit costs had increased over the period from 1958 to 1970.
Are these results proof that these materials were becoming less scarce over this time
period? Unfortunately, this unit cost measure has problems. First, technological progress
has undoubtedly reduced unit costs over this time period (for empirical evidence from the
oil sector, see Norgaard, 1975). This will also have the effect of increasing the size of eco-
nomic reserves. Second, the unit cost hypothesis relies on the assumption that firms will
always deplete the lowest-cost deposit first; yet to know which deposit is the lowest-cost one
implies a perfect knowledge of the characteristics of all deposits, some of which are yet to be
discovered. Norgaard (1990) has termed this the “Mayflower problem’:
[I]f the pilgrims knew where the best places for an agricultural colony were, they would not have
gone to Plymouth Rock. Many generations passed before American agriculture shifted from the
relatively poor soils of the east coast to the more productive mid-west.

Third, while unit capital and labour costs may have been falling, this might be due to
substitution of some other input for capital and labour. The obvious missing input here is
energy. Hall et al. (1986) recomputed Barnett and Morse’s figures for the coal and petro-
leum sectors, including energy use with capital and labour use: they found that whilst the
Barnett and Morse data showed a 35 per cent decline in unit costs for the petroleum sector,
the inclusion of energy use changed this to a 10 per cent increase. Fourth, unit costs are a
poor predictor of future scarcity, since they are based entirely on past experience, and are
272 APPLYING THE TOOLS

not ‘forward-looking’: technological advances could increase future economic reserves

even if, historically, unit costs have risen.

13.3.3 Real prices

Prices are well established in conventional microeconomics as indicators of scarcity. For natu-
ral resources, a rising real price has been argued by many to be a potentially good measure of
increasing scarcity (e.g. Fisher, 1981). This will be so when prices signal all future and current
opportunity costs of using up a unit of a non-renewable resource today. In basic versions of
the Hotelling model, the price of a resource rises at the rate of interest along an optimal deple-
tion time path, until it is equal to the price of the ‘backstop resource’—its closest substitute.
Several empirical studies have looked at real price data. The earliest comprehensive study
was by Barnett and Morse (1963), who found that for most primary products, real prices had
remained approximately constant from 1870 to 1957. Slade (1982) suggested that the time path
of prices might follow a U-shape, as an initial decline in prices due to technological progress was
eventually overcome by the tendency for increasing cumulative production to increase costs,
and by the desire of resource extractors to see user rents rising at the real rate of interest. Slade
found that a U-shape fitted the price series of twelve materials better than a linear form, indicat-
ing that for aluminium, for example, real prices started to rise in the 1960s. Finally, Moazzami
and Anderson (1994) repeated Slade’s analysis, using a somewhat different statistical technique.
They found strong evidence of increasing scarcity for some materials (such as coal and copper),
but only weak evidence of increasing scarcity for others (such as aluminium and iron).
But many criticisms can be levelled at the use of real prices as scarcity measures. First, the
influence of producer cartels on prices of primary products can be significant, and yet not
reflect scarcity changes. For example, the large oil price increases produced by OPEC in 1974
and 1979 were more to do with a voluntary reduction in supply to increase oil revenues than
an increase in scarcity. Other commodities (such as tin) have been similarly affected. Second,
governments intervene in resource markets, imposing price controls that distort price sig-
nals. An example here is the action taken by the United Kingdom government in the 1970s
and 1980s to keep gas prices high, to reduce a loss in sales by the nationalized electricity
companies (gas is a substitute for electricity in domestic heating and cooking). Tietenberg
(1992) documents distortions caused by the imposition of maximum prices (price ceilings)
by the US government for natural gas.
Third, natural resource prices do not measure social opportunity costs, partly because
producers are not forced to pay for the environmental damage caused by the extraction and
processing of these resources. For example, oil prices could be argued to be too low since not
all ofthe external costs associated with oil-drilling and refining are imposed on producers;
whilst a similar statement could be made for aluminium extraction (via bauxite-mining)
and processing. Natural resource prices do not measure one element of social opportunity
costs; namely, the environmental benefits forgone in their production.

13.3.4 Economic rent

Economic rent (or user rent or resource rent) is defined as the difference between price and
marginal cost. One result from the Hotelling model is that an efficient depletion path
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 273

involves resource rents rising at the rate of interest. The intuition behind this is clear: if
resource rents represent the rate of return on ‘holding’ a non-renewable resource deposit,
then this should be equal at the margin to the return on holding any other kind ofasset, such
as a savings bond. Rising rents are an indicator of scarcity.
But several problems exist with this measure. First, empirical data are scarce. Economic
rents are the difference between price and marginal extraction costs, but are not the same as
accounting profits. Neither firms nor governments are in the habit of recording these data.
Empirical economists have relied on proxy measures, such as exploration costs. The argu-
ment here is that rational firms will spend no more on exploration than the expected net
benefits (i.e. the expected future rents) to be gained. Devarjan and Fisher (1980) measured
average exploration costs for oil in the United States over the period from 1946 to 1971, and
found them to be rising, an indicator of increasing scarcity despite the fact that no such trend
exists in oil prices over that period. Yet as expected prices are a component of expected rents,
the criticisms of the real price measure given in the previous section also apply to the rent
measure.
Second, the use of rent as a scarcity measure assumes that firms are following optimal
depletion plans (Faber and Proops, 1993). Yet there is little evidence that this is so in reality
(see Box 13.2). What is more, to be able to follow the optimal depletion plan, firms need to
be fully informed about future prices and extraction costs: a rather more extreme version of
the Mayflower problem (although it is certainly possible to define a best-depletion pro-
gramme under conditions of uncertainty). Interest rate movements will also affect optimal
depletion programmes, such that changes in rent will pick up these macroeconomic effects
too. Rent is perhaps the best scarcity indicator from a theoretical point of view. After all, it
shows that gap between what society is willing to pay for one more unit of the resource and
the cost of extracting that unit. But it suffers from empirical drawbacks. It is possible for the
rent on a resource to decrease even though its physical abundance is falling.

The value of a unit of resource stock in the ground is the marginal profit that it earns:

Mr = p—MC.
We can approximately measure resource scarcity by either costs or prices, but neither is ideal, as the
discussion in Section 13.3 suggests. Empirical studies tend to arrive at conflicting conclusions: Why might
this be?
Farzin (1992) explores the impact that new technology has on costs. The cost curve for a renewable
resource firm depends on the rate of extraction, q, the remaining reserves, x, and the state of cost-
reducing technology, z. The cost function is c(q, x, Z), where costs are increasing in q, decreasing with the
size of the stock, and decreasing in z. The reason why costs are decreasing with the size ofthe stock is that
as extraction proceeds, the firm shifts to reserves that are more costly to exploit. What happens to costs
through time? New technology reduces costs, and increases the value of the resource stock. This might
be characterized as a ‘race’ between depletion and technological change to keep costs down.
What is the evidence? Lasserre and Ouellette (1991) show that, even with technical change, the decline
in the grade of asbestos ore extracted still increased marginal costs.
274 APPLYING THE TOOLS

13.3.5 Hotelling and the oil market

If we consider a graph ofthe oil price (Figure 13.4) from the 1970s to 2010, we see that prices
are not monotonically increasing, and that there is no straightforward evidence
(Krautkraemer, 1998; Livernois, 2009) to support Hotelling’s rule. The question then arises:
What drives price movements in non-renewable resource markets? This section considers
the case of oil and follows the recent analysis of Smith (2009).
Oil is the leading commodity in world trade, accounting for 13 per cent of commodity
trade by value (United Nations, 2008) and the oil price has a profound effect on the econo-
mies of most countries. The market structure is imperfectly competitive with OPEC, which
includes nine countries—Saudi Arabia, Nigeria, the United Arab Emirates, Nigeria,
Venezuela, Iran, Kuwait, Iraq, and Libya—operating as a cartel. These countries account for
a significant proportion of supply, 55 per cent in 2007, and a larger proportion of reserves,
87 per cent in reserves (Petroleum Intelligence Weekly, 2008; Smith, 2009). It is also notable
that a large proportion of the world production and supply is from state-owned companies;
for instance, Saudi Aramco in Saudi Arabia (see Box 13.3).
It is notable from Figure 13.4 that since 1972 the oil market has shown extreme price
volatility (variability) that cannot be explained by a market price driven by resource
scarcity.
Smith (2009) explains this volatility in terms of demand and supply shifts. The demand
elasticity for oil is relatively low. For instance, the US demand elasticity has been estimated
to range from —0.05 for the short run to -0.30 in the long run (OECD, 2004; US Energy
Information Administration, 2007). These elasticity values mean the demand for oil is unre-
sponsive to price changes. The US supply elasticities are estimated to range between 0.04
(short run) and 0.35 (long run) (OECD, 2004; US Energy Information Administration,
2007). The peak in oil prices in 2008 was due to a 33 per cent increase in demand from 2003
to 2008, set against a fall in supply from non-OPEC countries and a small rise in OPEC sup-
plies (as measured, OECD, 2004, by Smith 2009).

120.00
— Nominal price ——Real price 2010=100
100.00

80.00

60.00

40.00
barrel
dollars
US
per
20.00

0.00
1976 1981 1986 1991 1996 2001 2006
Figure 13.4 Dubai oil prices, 1972-2010.
Source: Platts, as reported in [Link]
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 275

Box 13.3 Energy Facts

According to the US Energy Information Administration 2010 database:

® Global oil production has increased to 86.8 million barrels per day (bbI/day), with Saudi Arabia leading
production at 10.5 million bbl/day; Russia at 10 million bbl/day, and the USA at 9.7 million bbl/day.
The USA consumes over 19 million day bbl/day of petroleum, China 9 million bbl/day, Japan 4.4
million bbl/day, and India 3.1 bbl/day.
@ The USA is the leading consumer of natural gas, and Russia follows, and together they account for
43 per cent of world demand.
® World coal production is just under 8 billion short tons. China is the largest consumer of coal, using
1.31 billion short tons, followed by the USA, which consumes 1.04 billion short tons, India, Russia, and
Germany.
® Canada, the USA, Brazil, China, and Russia are the five largest producers of hydroelectric power. Their
combined hydroelectric power generation accounts for 51 per cent of the world total.
® The USA leads the world in nuclear electric power generation, France is second, and Japan ranks third.
® The USA leads the world in biomass, geothermal, solar, and wind electric-power generation. Japan is
second, followed by Germany, Brazil, and Finland. These five countries account for 65 per cent of the
world’s biomass, geothermal, solar, and wind electric-power generation.

In 2008, Hubbert’s (1956) hypothesis of“peak oil’ came to the fore as an argument to explain
the price peak. The geology-based model is aptly described by Smith:
Hubbert’s idea was that, in the case of oil production, prices and other incentives are superflu-
ous; it is all a matter of time. His resulting model of production behaves like a ballistic missile,
first rising and then falling of its own accord.
(2009: 161)

The problem with this model is that it does not address the economic effects that high prices
have on supply and demand. Smith points out that estimates of remaining reserves are cur-
rently at an all-time high. In addition, if we consider other sources of oil, such as shale oil,
we may have up to another 160 years of oil supply.
The International Energy Agency (2008) presents a field-by-field analysis of expected oil
production up to 2030. It expects that demand for oil will increase from 82.3 million barrels
per day (bbl/day) in 2007 to 103.8 million bbl/day in 2030. The share of the OPEC countries
in the global output increases from 44 per cent in 2007 to 51 per cent in 2030. The new sup-
ply will come from natural gas liquids and new methods of oil recovery.
The same report analyses the long-term oil supply: approximately 2.4 trillion (10") barrels
of conventional oil were available before any production commenced. To date, we have
extracted 1 trillion barrels, and another 2-3 trillion barrels remain as currently accessible
reserves. That leaves another 5-6 trillion barrels as currently uneconomic reserves in the form
of ‘oil shales natural gas to liquid’ and ‘coal to liquid’. The point is that as conventional oil sup-
plies are depleted and if prices rise, these backstop technologies will become economic.
In the next section, the analysis is broadened to show how the set of non-renewable
energy resources contribute to the global energy market. At one time, oil was the key
energy resource, it is still the most important, but the energy market is evolving to reduce
276 APPLYING THE TOOLS

the dependence on oil and to include a wide range of non-renewable and renewable
resources. .

13.4 Global Energy Demand and Supply

People focus on natural resource scarcity in part because these resources are the source of
the energy and material needed to drive modern economies. From an economics perspec-
tive, energy plays three key roles in our lives:

e Energy is a consumer good. The energy derived from renewable and non-renewable
resources such as petroleum, natural gas, coal, hydro, nuclear, biomass, geothermal,
solar, and wind helps us grow and cook our food, warm and light our homes, and
power our cars.
e Energy is a factor ofproduction. Energy, combined with capital, labour, and land, is an
essential input in the production of nearly all goods and services around the globe.
e Energy is a strategic resource. Energy also has enormous strategic value for a nation, and
the threat of its loss has led to war. People and governments follow energy prices with
intense interest, because it is so vital to our daily lives.

Today, the world produces and consumes nearly 500 quadrillion (10!°) Btu (British ther-
mal units) of power. Global energy use is expected to increase by another 50 per cent over
the next two decades. As a comparative benchmark, energy use four decades ago in 1970
was about 200 quadrillion Btu. China, Russia, and the USA are the biggest producers and
consumers of world energy. Five nations—Canada, China, Russia, Saudi Arabia, and the
USA—currently produce about half of the world’s energy. The USA produces over 70 quad-
rillion Btu of energy; Russia and China produce over 40 and 33 quadrillion Btu. And five
nations—China, Japan, Germany, Russia, and the USA—consume nearly half of the world’s
energy. The USA consumes nearly 95 quadrillion Btu, three and four times that demanded
by China and Russia, at 34 and 26 quadrillion Btu, respectively. The next big consumers are
Brazil, Canada, France, India, and the UK.
As the world economy grows, so does the demand for energy. Emerging economies
demand more energy. Consider now how energy supply and demand has changed by region
over the past three decades, from 1980 to 2008. The largest regional increase in energy pro-
duction occurred in the Asia and Oceania region—production increased by nearly 100
quadrillion Btu in three decades (Table 13.2). The energy consumption in this region tripled
with an increase of 114 quadrillion Btu, again the largest change during this period (Table
13.3). This increase in energy use matches the rapid growth in this region’s economies over
the past three decades.
In North America, overall energy production and consumption have not grown as fast,
but have still increased by about 19 and 30 quadrillion Btu. The Middle East has increased
its production and consumption by 26 and 20 quadrillion Btu, respectively. Africa has
increased its production and consumption by 20 and 10 quadrillion Btu, respectively. In
Central and South America, we see production and consumption increased by 17 and 14
quadrillion Btu. Europe has increased supply and demand by about 6 and 14 quadrillion
NON-RENEWABLE NATURAL RESOURCES AND ENERGY Dale

Table 13.2 World primary energy production (quadrillion Btu)

Region 1980 1995 2008

North America 83.3 90.3 101.7

Central and South America ui 225) 29.6
Europe 40.2 49.0 46.5
Eurasia 5615 ile) UN olf
Middle East 423 48.3 68.2
Africa 174 24.1 BAS
Asia and Oceania SEES) WES 1343
World total 287.5 357.6 489.5

Source: US Energy Information Agency, [Link]

ipdbproject/[Link]?tid=44&pid=44 &aid=1 &cid=ww, 11 ,r2,r3,r4,r5,
r6,r7 ,&syid=2004 &eyid=2008&unit=QBTU.

Btu. In Eurasia, energy production has declined and then increased by about 15 quadrillion
Btu; energy consumption, however, has declined slightly, the only region to witness a
decline in energy demand.
Energy is not a homogeneous input. Different primary energy sources will be used based
on the relative energy prices. Cheaper energy sources will dominate the mix, and continue to
do so until there is a change in relative prices—either through increased scarcity or a change
in technology. People made changes in their energy mix and use when the OPEC cartel
restricted oil supply, causing energy prices to spike in the 1970s. Technological progress and
innovation also induce people to switch towards less expensive energy sources over time, as
old capital is retired and new technologies prove themselves to be effective and reliable.
What is the current mix of primary energy supply around the world? Table 13.4 shows that
the non-renewable resources of petroleum, coal, and natural gas are the big three energy
sources today. Petroleum remains the most important source of energy (168 quadrillion Btu
in 2006): Saudi Arabia, the USA, and Russia are the three largest suppliers. Coal is second,

Table 13.3 World primary energy consumption (quadrillion Btu)

Region 1980 1995 2008

North America 91.6 109.3 WS)

Central and South America Ws lH 25.8
Europe 718 how) 85.8
Eurasia 46.7 42.2 45.8
Middle East 5.8 1338 25.5
Africa 6.8 10.7 16.1
Asia and Oceania 49.0 Oeil 168352
World total 283.2 365.4 493.0

2 Data for 2007

Source: US Energy Information Agency, [Link]
[Link]?tid=44 &pid=44 &aid=1 &cid=ww,r1 12,13,r4,r5,r6,r7,&syid=2004&
eyid=2008&unit=QBTU.
278 APPLYING THE TOOLS

Table 13.4 World production of primary energy by energy type (quadrillion Btu), 1980-2006

Energy type/country group 1980 1990 2000 2006

Petroleum 1338 136 156 168

Dry natural gas 5p 76 91 107
Coal ral vil 90 129
Net hydroelectric power \ 18 22 26 30
Net nuclear electric power 8 20 26 28
Net geothermal, solar, wind, and wood and waste electric power 0.5 2 3 5
US production of biomass, geothermal and solar energy not used 2 2 2 3
for electricity generation
Total energy 288 350 396 469
Source: Energy Information Administration, International Energy Annual 2006, [Link]
[Link].

with 129 quadrillion of production: China and the USA are the leading producers. Natural
gas ranks third, supplying about 107 quadrillion Btu, and has increased its share over the past
decade: Russia is the leading producer. The remaining energy sources—hydro, nuclear, bio-
mass, geothermal, solar, and wind—make up the balance. Together, they accounted for a
combined total of 66 quadrillion Btu. While currently a small share, these other sources of
energy have begun to increase over the past decade.
An understanding of how prices drive the mix of energy has implications for the future of
global energy use. As we showed earlier in the chapter, the most relevant economic model
is Hotelling’s rule for the pricing of non-renewable energy resources. So how well has
Hotelling’s rule explained price movements? Economists have had mixed success in empiri-
cally validating the rule. The needed data are difficult to come by, and have forced most
studies to use proxies for user rents and expectations of future interest rates. Since user rents
are usually confidential and we do not observe expectations in practice, whether the
Hotelling rule is a reasonable guide for the future remains an open question.

Table 13.5 World estimated recoverable coal, 2008

Region Recoverable reserves Share of recoverable reserves (%)

JSA and Canada 243.9 28.3

Russian Federation (SA 18.2
China 114.5 13.3
Europe (incl. Turkey) 110.4 12.8
Australia 76.4 89
ndia 60.6 7.0
Kazakhstan S216 aus)
South Africa 30.2 Sin
All other 34.3 4.0
World total 860.9 100
SE ee ee eee
Source: World Energy Council estimates of proved recoverable reserves (hard coal and lignite) at end-
2008 (billion metric tons).
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 219

What does the future of energy demand look like? The US Energy Information Agency’s
International Energy Outlook 2012 ([EO2012) predicts that world energy demand will
increase by about 44 per cent over the next two decades. The outlook forecasts that world-
wide energy use could increase to over 770 quadrillion Btu in 2035, from 505 quadrillion Btu
in 2008. While the developed nations have long been the dominant users of energy, emerg-
ing economies such as China and Russia are predicted to drive much of the future growth in
demand. Energy use in Asia and Central and South America is projected to more than
double.
What does the mix of energy sources look like into the future? Oil currently supplies the
largest share of world energy consumption, about 86 million barrels per day in 2010. The
IEO2012 predicts that worldwide oil demand will hit about 113 million barrels per day by
2035. Growth in oil use in the developed nations should come from the transportation sector,
since the few alternative sources are expensive. Oil demand in the developing countries could
come from both transportation and heating demand. Table 13.6 gives reserves of fossil fuels.
Natural gas will remain the fastest-growing component of world energy demand. The
IEO2012 predicts that natural gas use could increase to 153 trillion cubic feet (tcf) in 2035,
from 113 tcf in 2010. In 2012, many experts see the emergence of shale gas as a ‘game
changer’ in US natural gas energy markets. Shale gas can now be recovered cost-effectively
from low-permeability geological areas, given new advances in hydraulic fracturing.
According to one estimate supplied by the US Energy Information Agency, shale gas
reserves could be about 750 trillion cubic feet in the USA alone.
Coal is predicted to grow at about 0.9 per cent between 2008 and 2035, increasing from
139 to 175 quadrillion Btus. The main reason why coal will continue to grow is to meet the
expanding energy demands in Asian nations. China and India are predicted to account for
most of the increased world demand for coal. The future of nuclear power is less clear. The
IEO2012 shows an increase in nuclear power from 2,600 to 4,900 billion kilowatt hours
from 2008 to 2035. Nuclear power remains one of the key substitute energy sources given
climate change policy, but fears of nuclear meltdown remain in many people’s minds, espe-
cially after the Fukushima Daiichi nuclear accident in 2011.
Renewable resource development will be slow if the expected price of fossil fuels remains
relatively low in the near future. Renewables cannot compete with low fossil fuel prices, but
this is changing as relative costs are declining. The IEO2012 projects a doubling of energy
through hydroelectric projects—world consumption is predicted to increase to 102 quadril-
lion Btu in 2035, from 55 in 2010. The report also suggests that the installed wind-powered
generating capacity could more than double; increasing to over 520 gigawatts in 2035, from
180 in 2010. Similarly, the solar generating capacity is estimated to increase to 117 gigawatts
in 2035, from 25 in 2010.

13.5 Global Issues in Energy Policy

We care about the future energy trends because energy plays such a crucial role in the
modern world. Governments remain tempted to intervene in energy markets. Policy-
makers have long wanted to control energy for national security reasons. Many reasonable
arguments exist to justify government intervention on national security grounds; that is,
280 APPLYING THE TOOLS

Table 13.6 World crude oil and natural gas reserves (March 2009)

Region/country Crude oil (billion Crude oil (billion | Natural gas (trillion | Natural gas
barrels) Oil and barrels) World cubic feet) Oiland _ (trillion cubic
Gas Journal Oil Gas Journal feet) World Oil

North America 209 Sy 308 314

Central and South America 122 04 267 245
Europe 14 14 169 169
Eurasia 98 126 ess 2,104
Middle East 746 Ven, Zoe. 2570
Africa Wid 15 494 504
Asia and Oceania 34 40 430 528
World total 1,342 1,184 6,254 6,436

Source: US Energy Information Agency, [Link]

to reduce the odds of another ‘energy crisis, to promote the public good of research and
development into new technologies. Now policy-makers have added environmental con-
cerns as a justification for intervening in energy markets. As we have read throughout this
book, private markets can fail to provide the socially desired level of agood or service, and
energy markets are no exception.
Global climate change, regional acid rain, and local air pollution problems are all linked
to fossil fuel use. Petroleum and coal, no matter how efficiently burned, all produce the
carbon dioxide feared to be warming the planet to unacceptable levels; they also produce the
sulphur oxides and nitrogen oxides that lead to the acidification of soils and surface water;
they also produce smog and particulate matter that affect human health. People have also
worried about the health and safety issues from the actual mining of energy sources—for
example, black lung disease—and the impacts ofstrip mining.
The question is should governments intervene in energy markets for environmental pro-
tection, and if so, how? Justifiable intervention for environmental protection depends on
whether one can reasonably argue that the gains of new rules and regulations outweigh the
risks of slowing down the productivity that drives the development of an economy. If the
gains dominate the costs, there are three general ways in which the government can inter-
vene in energy markets: by changing economic incentives through taxation of fossil fuels
and the subsidization of renewable fuels; by expanding technological options through the
promotion and subsidization of research and development (R&D); and by providing infor-
mation about options that promote energy efficiency.
Changing economic incentives. Relative prices drive the mix of energy demand. Energy
prices encourage people to make the least costly adjustments based on their own personal
information. Governments that wish to alter this mix can change the relative prices through
economic incentives such as taxes and subsidies. Energy policy could provide economic
incentives in a variety of ways. A typical economic incentive policy is to increase the private
cost of fossil fuels, forcing users to address the social costs of emissions. This policy raises
the price of emissions, by taxing emissions on the basis of their potential to cause environ-
mental harm, removing existing subsidies that increase fossil fuel use (e.g. parking subsi-
dies), or adding new subsidies that promote renewable fuels and lower-emissions fuels or
28)

The UK Renewable Energy Advisory Group has defined renewable energy as ‘those energy flows that
occur naturally and repeatedly in the environment and can be harnessed for human benefit’ (REAG,
1992). Europe has a great and largely untapped potential for the generation of power and heat from
renewable sources of energy. Renewables currently supply 5.3 per cent of the European Union’s energy
consumption, but only 1 per cent in the UK. In a 1996 green paper, the European Commision suggested a
target to increase renewables’ share of gross energy consumption to 12 per cent by 2010 (ENDS, 1996).
Achievement of this target would reduce the EU's annual carbon dioxide emissions by around 250 million
tonnes (ENDS, 1997). Although renewable energy technologies are being continually developed, many
are still significantly underutilized due to financial, technical, environmental, and social factors.
Private-sector development of renewable energy is promoted in the UK by guaranteeing a certain
portion ofelectricity demand from electricity supply companies for renewable sources, and offering a
higher payment to renewable operators per kilowatt supplied, compared to the price of fossil fuel
electricity. This difference is funded by a fossil fuel levy on consumer's electricity bills. The two schemes that
implement this policy (in England and Wales, the Non-Fossil Fuel Obligation, and in Scotland the Scottish
Renewables Orders) have been successful in increasing private-sector investment in renewables. In
Scotland, this has mainly taken the form of wind farms, biogas plants, and small-scale hydropower schemes.
But while the environmental benefits of renewable energy are well known (mainly avoided emissions
from fossil fuel power stations), the environmental costs of renewables receive less attention (European
Commission, 1995). In Scotland, these include landscape impacts from wind farms, local dis-amenities
from waste-to-power schemes, and ecosystem impacts from hydro plants. Hanley and Nevin (1999)
report some results from a study of these environmental costs for a range of alternative renewable energy
investments in the crofting community of Assynt, in the north-west Highlands. This is an interesting case,
since the investments would be made on behalf of the local community. However, there is a concern that
tourists could be put off by such changes. An energy audit and environmental impact analysis were
carried out, which resulted in three main viable options being identified: a small wind farm, a small-head
hydro, and a biomass planting scheme. Contingent valuation was used to study impacts on local
residents: this revealed community views that differed across the three options. A tourist survey was also
undertaken to estimate the differing impacts of each option on tourist spending. The results were then
presented as a ranking of options.
The following table shows the energy options and potential tourism effects:

Renewable energy option Possible effect on total Possible effect on total Implied
season's income from season's income from ranking
tourism (E£) if trips were tourism (£) if trips were
shortened by 1 day shortened by 0.5 days

Wind farm at Raffin -2,590 -1,295 1

Small-scale hydro on Loch Poll -17,208 =8,604 2
Biomass schemes at Culkein/ ~26,829 -13,415 3
Stoer
a ————-..”0000 OOO —————————————LLLLQLQGQ—Lv——

The next table shows the energy options and the willingness to pay (WTP) of residents.
The relationship between support for renewable energy options and number of years in residence on
the estate was interesting. With respect to both the wind and hydro options, there was less support
amongst those who had lived on the estate for 10 years or less. In contrast, a low level of support for
biomass development was evident among residents who had lived in the area for many years. The taking
of land from traditional crofting areas that this option would require may have been a factor in this result.
282 APPLYING THE TOOLS

Renewable energy option Percentage Proportion of those Implied community Implied

in favour opposed who would WTP(perannum)? ranking
ofscheme acceptcompensation? (£)

Wind farm at Raffin 78 3/10 13,585

Biomass schemes at Culkein/Stoer 42 7/26 6,642 a
Small-scale hydro on Loch Poll 87 0/6 14,282 1

2 The sum ofthose willing to accept personal income compensation and local jobs compensation.
> Based on population of 260 adults living on the estate.

technologies (see Box 13.4). Another policy to change relative prices is to limit emissions or
energy use, and let people trade permits to pollute.
Changing the relative price of fossil fuels would give people an incentive to reduce their
energy consumption—they would drive less, and they would turn down the thermostat.
People would also have incentives to buy more energy-efficient equipment; for example,
more insulation or a car with better fuel consumption. Sellers would also adjust if the
demand for energy-inefficient goods declined. Everyone has incentives to develop more
energy-efficient technologies, and find low-emission energy sources.
However, economists argue that these energy savings will not be as high as predicted by
engineering models. People who buy energy-saving products experience a ‘rebound effect’;
people who drive cars with better fuel consumption drive more miles, since they get a better
mileage per gallon. This rebound effect works based on both a substitution and an income
effect. Increased fuel efficiency reduces the costs of fuel consumption, which—due to the
substitution effect—increases consumption of fuel. In addition, since fuel is cheaper, the
consumer has more real income, which allows him or her to consume other goods and ser-
vices that also use energy.
Expanding technological options. Governments can intervene in the energy market by
promoting the R&D of new technologies that address the environmental problems associ-
ated with fossil fuel use. The private sector generally underinvests in this type of R&D,
because it cannot capture all the benefits. These public programmes include government-
funded programmes and subsidies for private R&D. Options include R&D into fossil fuel
technologies with greater conversion efficiencies and non-fossil fuels such as biomass, wind,
solar, geothermal, and nuclear energy.
Information programmes. Governments can also try to alter the energy market by pro-
viding information to people about different energy-efficient options. Policies that pro-
mote the market penetration of new technologies include information and outreach
programmes, green programmes (e.g. offering electricity supplies from renewable sources
at higher tariffs), market identification, and targeting. Governments can form partnerships
with industry and others to promote less environmentally damaging fuels. For example,
governments can try to overcome the information problems associated with landowners
and renters. Neither has any incentive to invest in energy efficiency, because the landowner
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 283

does not pay the utility bill, while the renter might not get reimbursed for investments in
energy efficiency.
The ability to cause much change through information is an open question.
Unexploited opportunities to use cost-reducing technologies might exist, but other fac-
tors impede the realization of these opportunities. People typically respond to changes
in energy prices and clearer information on the potential to remove the hidden costs of
technology switching not addressed in engineering cost estimates. The choice of energy
technology offers no free lunch. Even if new technologies are available, many people are
unwilling to experiment with new devices at current prices. Factors other than energy
efficiency also matter to consumers, such as quality, and the time and effort required to
learn about a new technology and how it works. People behave as if their time horizons
are short, perhaps reflecting their uncertainty about future energy prices and the relia-
bility of the technology. Some people use less fossil fuel because it is, in their opinion, the
right thing to do; others will need more convincing, based on climate change risks or a
change in energy prices.

Summary
Adam Smith said in The Wealth of Nations that ‘[e]very man endeavours to supply by his
own industry his own occasional wants as they occur. When he is hungry, he goes to the for-
est to hunt; when his coat is worn out, he clothes himself with the skin of the first large ani-
mal he kills: and when his hut begins to go to ruin, he repairs it, as well as he can, with the
trees and the turf that are nearest it? What was true then is true now. People use natural
resources for their own benefit, sometimes at their own expense, sometimes at the expense
of others. The goal of natural resource economics is to address these human needs and natu-
ral resources limits—to define the constraints, to confront them, and to design rules to
increase the efficient use of land, forests, and fisheries. Natural resource economics has
shown why physical and biological constraints matter to economic reasoning about the eff-
cient nature of capital stocks; how natural resource economic reasoning can alter manage-
ment strategies when one accounts for the additional opportunity costs of depletion or not
harvesting a resource; and what challenges remain, including defining durable property
rights systems, designing effective regulatory tools through incentives, and valuing changes
in the non-market services provided by natural resources such as climate change and
biodiversity.
The energy that we derive from natural resources help drive the global economy. In this
chapter, we have taken an economic perspective to explore how resource extractors should
manage their resource deposits—given certain simplifying assumptions. We have also
reviewed different measures of resource scarcity. We have presented the case as to why
‘resource lifetimes’ is a flawed measure of scarcity. We have examined the role of energy in
the world economy, and potential changes in global supply and demand. Finally, we have
looked at alternative tools with which governments can implement an energy policy that
tries to reduce the environmental costs of energy. An economic perspective provides a useful
mindset to help understand the factors that either limit or promote the switch towards a
more sustainable energy future.
284 APPLYING THE TOOLS

Tutorial Questions
13.1 Economists believe that energy plays three key roles in our lives. Explain.
13.2 Explain the economic model of how society should extract its natural resources.
13.3. Why do the predictions of the simple version of Hotelling’s rule not always hold?
13.4 Is the market price of oil an accurate measure of resource scarcity?
13.5 Define the idea of lifetime of a resource as a measure of resource scarcity. Why don’t
most economists believe it is a helpful measure?
13.6 Explain the idea behind the rebound effect in energy efficiency questions.
13.7 How do high oil prices affect how one should think about the idea of Hubbert’s
peak oil?

References and Additional Readings

Anderson, F. (1985). Natural Resources in Canada Hall, D., Cleveland, C., and Kaufman, R. (1986).
(Toronto: Methuen). Energy and Resource Quality: The Ecology of the
Barnett, H.J., and Morse, C. (1963). Scarcity and Economic Process (New York: Wiley).
Growth: The Economics of Natural Resource Hanley, N., and Nevin, C. (1999). ‘Appraising
Scarcity (Baltimore, MD: Johns Hopkins renewable energy developments in remote
University Press). communities: the case of the North Assynt Estate,
Cohen, D. (2007), ‘Earth audit, New Scientist, 194 Scotland’ Energy Policy 27: 527-47.
Issue 2605: 34-7. —,, Shogren, J.F., and White, B. (2007).
Devarjan, S., and Fisher, A. (1980). “Exploration Environmental Economics: In Theory and Practice
and scarcity, Working Papers in Economic (Basingstoke: Palgrave).
Theory IP-290, University of California, Harris, D., and Skinner, B. (1982). “The assessment of
Berkeley. long-term supplies of minerals, in V.K. Smith and
ENDS (1996). The Ends Report 263: 17-18 J. Krutilla (eds.), Explorations in Natural Resource
(Environmental Data Services Ltd). Economics (Baltimore, MD: Johns Hopkins
University Press).
— (1997). The Ends Report 275: 43-4
(Environmental Data Services Ltd). Hotelling, H. (1931). ‘The economics of exhaustible
resources, Journal ofPolitical Economy 39:
European Commission (1995). Eastern Externalities
137-75.
of Energy, vi: Wind and Hydro, Report 16525
(Brussels: EC). Hubbert, M.K. (1956). Nuclear Energy and the
Fossil Fuels, Publication 95, Shell Development
Faber, M., and Proops, J. (1993). ‘Natural resource
Company Exploration and Research Division.
rents, economic dynamics and structural change’
Ecological Economics 8(1): 17-44. International Energy Agency (2008). World Energy
Outlook (OECD/IEA: Paris).
Farzin, Y.H. (1992). “The time path of scarcity rent
is the theory of exhaustible resources, Economic Johnson, M., Bell, E, and Bennett, J. (1980). ‘Natural
Journal 102: 813-31. resource scarcity: empirical evidence and public
policy, Journal of Environmental Economics and
Fisher, A.C. (1981). Resource and Environmental
Management 7: 256-71.
Economics (Cambridge: Cambridge University
Press). Krautkraemer, J. (1998), ‘Nonrenewable resource
scarcity, Journal of Economic Literature 36:
Gordon, R.B., Bertram, M., and Graedel, T.E.
2065-107.
(2006). ‘Metal stocks and sustainability’
Proceedings ofthe National Academy ofSciences Lasserre, P., and Ouellette, P. (1991). ‘The
103; 1209-21. measurement of productivity and scarcity
NON-RENEWABLE NATURAL RESOURCES AND ENERGY 285

rents: the case of asbestos in Canada, Journal of U.K. Continental Shelf, Economic Journal 100:
Econometrics 48: 287-312. 367-90.

Lee, J., List, J.A., and Strazicich, J. (2006). Petroleum Intelligence Weekly (2008). Vol. XLVII, No.
‘Nonrenewable resource prices: deterministic 50, 15 December, [Link]
or stochastic trends?’ Journal of Environmental REAG (Renewable Energy Group) (1992). Energy
Economics and Management 51: 354-70. Paper 60 (London: HMSO).
Livernois, J. (2009). ‘On the empirical significance Skinner, B. (1976). ‘A second Iron Age?’ American
of the Hotelling rule) Review of Environmental Scientist 64: 258-69.
Economics and Policy 3: 22-41.
Slade, M. (1982). “Trends in natural resource
Moazzami, B., and Anderson, F. (1994). ‘Modelling commodity prices: an analysis of the time
natural resource scarcity using the “error- domain; Journal of Environmental Economics and
correction” approach, Canadian Journal of Management 9: 122-37.
Economics 27: 801-12. Smith, J.L. (2009). “World oil: market or mayhem;
Norgaard, R. (1975). ‘Resource scarcity and new Journal of Economic Perspectives 23: 145-64.
technology in US petroleum development, Natural Tietenberg, T. (1992). Environmental and Natural
Resources Journal 15: 265-82. Resource Economics (London: HarperCollins).
— (1990). ‘Economic indicators of resource United Nations (2008). Commodity Trade Statistics
scarcity: a critical essay, Journal ofEnvironmental Database, 2006, [Link]
Economics and Management 19: 19-25. (accessed 1 June 2012).
OECD (Organisation for Economic Co- US Energy Information Administration (2007).
operation and Development) (2004). ‘Oil price NEMS International Energy Module (IEM),
developments: drivers, economic consequences Model Documentation Report, DOE/EIA-M071.
and policy responses, in OECD Economic Outlook —— (2012) Annual Energy Outlook 2012
No. 76 (Paris: OECD). ‘ (Washington, DC: USEIA), [Link]
Pesaran, M.H. (1990). ‘An econometric analysis forecasts/aeo/pdf/0383(2012).pdf (accessed 15
of the exploration and extraction of oil in the November 2012).
Glossary

Adaptation An investment to reduce the severity of Clean development mechanism A system in which
realized damages. developed nations buy the carbon reductions in
developing nations.
Adverse selection A problem related to the hidden
type of aperson or firm such that a buyer cannot Climate change A change in the climate (global
distinguish the quality of agood or service. temperatures and precipitation) over time due to changes
in natural process or external forcings.
Ambiguity Uncertainty about the probabilities of an
outcome. Coase theorem [Iftransaction costs are low, two
disputing parties can bargain to a solution provided that
Ancillary benefit An additional benefit that arises
a third party assigns property rights to one of the
from reducing fossil fuel consumption; for example, lower
parties.
levels of carbon monoxide, sulphur and nitrogen oxides,
and toxic trace pollutants in exhaust gases. Common property A good defined by rival
consumption and open or controlled access to a resource.
Autarky Where a country does not engage in
international trade.
Comparative advantage A measure of the difference
Behavioural economics The application of in the opportunity cost of producing a good between two
psychological insights into economic theory and countries. The existence of acomparative advantage is a
observation. prerequisite for there being a gain from trade.

Benefits transfer The practice of extrapolating existing Co-operative game theory This concerns the returns
information on the non-market value of goods or services. to forming coalitions and sharing the pay-offs.

Bid curve ‘he use of regression analysis to determine Core In co-operative game theory: the set of pay-ofts
the relationship between willingness-to-pay (WTP) that can be agreed that will keep a coalition of players
measures and the factors that are thought to influence together. They must receive at least as much as part of the
these measures, coalition as they do either alone or in another coalition.

Bid vehicle The means by which individuals pay in a Cost-benefit analysis A decision-making rule used to
hypothetical market. evaluate a number ofdifferent policy options or projects in
order to determine which contributes the most to net
Biodiversity hotpots Regions identified by social well-being or, equivalently, which is the most
Conservation International as combining high levels of
efficient in terms of its use of resources.
biodiversity with high levels of threat to habitat.
Critical natural capital That part of natural capital
Bioprospecting Analysing compounds derived from
that is either essential for human survival and/or cannot
flora and fauna for compounds that could form the basis
be substituted for by other forms of capital.
for new drugs.

Carbon footprint A measure that accounts for all Deforestation Kuznets curve The hypothesis that
carbon emissions generated by a nation’s citizens,
the rate of deforestation will decline when a country
regardless of where the good was produced. reaches higher levels of income.

Carbon sink A process that destroys or absorbs Dominant strategy A strategy that a player can choose
greenhouse gases, such as the absorption of atmospheric that is best regardless of what the other players in the game
carbon dioxide by trees, soils, and other types of choose.
vegetation,
Dose-response model A measure of the economic
Carbon tax A fee added to the price of fossil fuels value of the environment as an input to production by
according to their relative carbon content. considering the impacts of pollution on market-valued
outputs.
Characteristics Theory of Value The theory that the
value of something is best explained in terms of its Economic efficiency Getting the most net benefits
characteristics or attributes, one can with the available resources.
GLOSSARY 287

Economic growth The change in gross national services, and increases demand for environmental goods
product (GNP) over a given time period—say, a year. and services.

Economic rent The difference between resource price Genuine savings (GS) An indicator of sustainable
and the marginal cost of extraction (also called ‘user rent’ development that measures the value of the net change in
or ‘resource rent’). all forms of capital—produced, natural, human, and
social—in an economy over a given period of time; a
Ecosystem services The valuable resources and negative genuine savings value indicates that development
services provided by natural ecosystems; for example, is unsustainable.
carbon sequestration, nutrient cycling, and climate
regulation. Ecosystem services provide inputs into the Geochemically scarce mineral A metal held in ores
production of some goods and services, and may also be a (typically silicates) that requires a relatively high level of
direct source of utility. energy to extract. Such ores can be too costly to extract.

Environmental ‘bads’ Those environmental inputs for Green net national product (green NNP) A
which the individual prefers less to more. sustainable development indicator that involves correcting
the national accounts to allow for environmental inputs to
Environmental ‘goods’ Those environmental inputs
the economy that are unpriced by the market; for example,
for which the individual prefers more to less.
depreciation of natural capital and the value of changes in
Environmental Kuznets curve (EKC) The idea that environmental quality that impact directly on people's
the relationship between income and environmental utility; rising green net national product over time
quality is an inverted U-shape; as income per capita indicates that development is sustainable.
increases, environmental impact also increases initially,
Gross national product (GNP) The monetary value
until it reaches a maximum and then begins to decline.
of the aggregate output, expenditure, or incomes of a
Existence or non-use value The utility people derive country in a given time period. Gross domestic product
from just knowing that a particular natural asset exists. (GDP) is another way of measuring this aggregate output,
income, or expenditure.
Expected utility The combination of probabilities and
the utility associated with outcomes of good and bad Hartwick rule The idea that an economy can maintain
events. constant consumption levels over time ifall rents from
non-renewable resource depletion are reinvested in
Experimental auction A method that values risk produced capital.
reduction using a bidding auction.
Hidden information When one person knows more
Externalities When one party imposes costs or benefits about his or her actions (e.g. protection level) or type (e.g.
on another party without paying or receiving low-cost) than the environmental regulator.
compensation.
Hotelling’s rule An economic theory proposed by
Fat-tail distribution A fat tail reflects the idea that the Hotelling (1931), which predicts that, in equilibrium, the
probability of an extreme-impact event is not as rare as marginal profit of anon-renewable resource firm increases
believed under the ‘business as usual’ climate scenarios. at the interest rate through time. Simpler versions of the
model predict that the price (when costs are zero) of the
Faustmann formula This formula gives the optimal non-renewable resource will increase at the interest rate.
rotation for a forest that is replanted and clear-felled at
regular rotation lengths. Human development indicator A measure
developed by the United Nations that uses three
Flow pollution The annual rate of carbon emission. indicators—GDP, education levels, and life expectancy—
to represent the development of a given country.
Framing effect When people are affected by how
information is presented.
Hypothetical market bias The tendency for
Free-rider A country that does not contribute to respondents in a stated-preference survey to overstate or
climate policy, but that still captures all the benefits of understate their actual willingness to pay (WTP) fora
climate protection and pays none of the costs. change in environmental quality.

Frontier model A theory of deforestation that is driven Immizerization model A theory of deforestation that
by opening up new regions to logging through investment: is driven by incremental expansion of shifting agriculture
this contrasts with the immizerization model. driven by population growth and impoverished families
striving to increase their incomes through expanding
Gains from trade The hypothesis that trade raises agricultural production: this contrasts with the frontier
income: it allows countries to attain more goods and model.
288 GLOSSARY

Indirect benefits Environmental benefits that are Non-point pollution Emissions that enter water-
measured indirectly via their role in the production bodies in a diffuse manner, such as through run-off from
process; for example, the role that wetlands play in the agriculture or forestry.
production of fish caught by commercial fishing.
Non-renewable resource Any resource—for
Insurance An investment that transfers wealth from instance oil, gold and iron ore—that has a fixed stock. This
good to bad states of nature given that a bad event has contrasts with a renewable resource (such as a fishery) that
been realized. grows through time by reproduction and growth.

Kyoto Protecol The international environmental Non-timber forest products (NTFP) This includes
agreement that binds the participating countries to carbon products taken from forest areas, such as bushmeat, nuts,
emissions reductions below 1990 levels by 2008-12. berries, and rubber, that depend upon a standing forest.

Loss aversion A concept from behavioural Non-uniformly mixed pollutants Pollutants whose
psychology claiming that people value losses more impact on water quality varies spatially.
highly than equivalent gains, because they value that
which they already have more than that which they Opportunity approach to sustainable
could acquire. development This requires that future generations
have at least as much capital as the current generation, so
Marginal abatement cost (MAC) The cost of that they have the opportunity to achieve the same levels
decreasing emissions by an additional unit. of well-being as the current generation.

Market An exchange institution to create economic Opportunity cost The best alternative use forgone.
value through trade.
Option price The willingness to pay (WTP) to reduce
Market failure When the allocation of scarce resources the probability of abad outcome.
achieved through markets is socially inefficient.
Outcome approach to sustainable
Maximum sustainable yield In relation to fisheries development This requires that utility or consumption
economics, this is the maximum catch that a fishery can do not decline over time.
sustain without depleting the stock.
Pigovian/green tax A price set equal to the marginal
Measure of resource scarcity The value of a unit of damages caused by pollution.
resource in the ground is measured as the difference
between the resource price and the marginal cost of Plantation forest Typically, a monoculture of trees
extraction. If the measure ofscarcity is increasing, the planted either for timber or for other forms of production.
resource is becoming scarcer in economic terms. It is distinct from a natural forest, as it is often comprised
Economic scarcity is distinct from geological scarcity. of a non-native species, whereas natural forests are
communities of species that are native or endemic to a
Mitigation An investment to reduce the probability of region.
damages due to carbon emissions
Point source pollution Emissions that enter water-
Moral hazard The hidden actions ofaperson or firm bodies from a single identifiable source, such as a pipe.
that allow them to capture benefits or reduce costs to the
detriment of others. Point-non-point trading An emissions trading system
in which point sources of pollution and non-point sources
Nash equilibrium [na gaming situation, a set of
of pollution trade emissions permits with each other.
strategies, one for each player, where no player can
improve his or her pay-off by switching to another Pollution havens hypothesis This predicts that
strategy. In other words, this occurs when each player is pollution-intensive industries will relocate to developing
playing the best response to all the other players’ countries as environmental regulation in developed
strategies. countries is made more stringent.

Natural forest This is difficult to define, because most Porter hypothesis A claim that environmental
forests have been modified to some extent. It can be taken regulation stimulates innovation and productivity.
to refer to a forest largely made up of the endemic native
species found in a region, In many cases, these forests have Preference heterogeneity ‘The idea that people's
been modified by human intervention. preferences vary and thus that individuals place different
values on a given good.
Net present value The sum of discounted benefits
associated with a particular project minus the sum of Preference reversal When a person is inconsistent in
discounted costs. his or her preference rankings and the monetary
GLOSSARY 289

evaluation of lotteries; for example, you rank lottery A Self-insurance/adaptation An investment to change
higher than lottery B, but you put a higher dollar value on the severity of the realized outcomes of events.
B than on A.
Self-protection/mitigation An investment to change
Production-function approaches Valuing non- the probability of events.
market changes in environmental quality by estimating
the implications of such changes for the output and price Shadow project This requires any action that reduces
of amarket-valued good or service. the stock of natural capital to be offset by a physical project
that replaces the natural capital loss.
Production possibility frontier (PPF) This gives the
combinations of goods that it is possible to produce in an Species-area curve A relationship between the area of
economy given the resources (land, labour, and capital) of land considered and the number of unique species found.
that country. For instance, if there are only two goods As more land is considered in a region, then the number
produced—food and cloth—then the PPF tells us how of extra species found tends to decline.
much food can be produced if an amount of cloth is
produced. The slope of the PPF gives us the opportunity
Stated-preference methods Direct questioning of
people using surveys in order to determine their
cost of producing one more unit ofagood.
willingness to pay (WTP) or willingness to accept
Property rights The ownership of resources, and the compensation (WTA) for a hypothetical change in
obligations and responsibilities that go along with environmental quality; such methods include contingent
ownership. valuation and choice experiments.

Public goods Goods defined by non-rival consumption Stern Review A report commissioned by the UK
and non-excludability to access the resource. government to study the economics of climate change. The
main conclusion was that the benefits (i.e. avoided
Race-to-the-bottom hypothesis The idea that damages) of extreme and immediate climate protection
countries open to trade adopt looser environmental outweigh the costs.
standards as a means of maintaining international
competitiveness. Stock pollution The total level of accumulated carbon
in the atmosphere.
Rational choice When people make consistent
choices that maximize their well-being given budget Strong sustainability This requires a non-declining
constraints. stock of natural capital and so assumes that natural capital
cannot be directly substituted for by human, social, or
Repeated game A situation in which a one-shot game produced capital.
is repeated. This may lead to a different solution to the one-
shot game, especially as players gather information about Sustainable development Economic development
that allows the current generation to meet its needs
the strategies of the other players.
without compromising the ability of future generations
Revealed-preference method Determining the also to its own needs.
value that individuals place on non-market environmental
Total economic value The sum of direct benefits
goods by studying their behaviour in markets for related
(use + existence values) and indirect benefits.
goods; such methods include hedonic pricing and travel-
cost models. Tradable pollution permit A market created to buy
and sell the rights to emit a unit of pollution into the
Risk The combination of probabilities and consequences
environment (also called marketable permits or cap-and-
of good and bad states of nature.
trade).
Risk assessment The quantitative estimation of risks to
Trade liberalization A process by which a group
human and environmental health.
of countries reduce or eliminate import quotas and
Risk perception A belief that people hold about risk tariffs.
and risky events.
Tragedy of the commons In the extreme version, due
Safe minimum standard (SMS) This identifies the to Garrett Hardin, a situation in which the overuse of a
minimum viable level for a natural resource, and only allows common-property resource leads to resource degradation
this safe minimum standard to be breached ifthe social and environmental collapse. The term is also used to
opportunity cost of maintaining the standard is too high. describe a situation in which there is an incentive for an
individual firm to overexploit a resource (e.g. a fishery),
Safe operating space A series of threshold levels for leading to a Nash equilibrium in which all firms
important ecosystem processes that should not be crossed. overexploit the resource.
290 GLOSSARY

Uniformly mixed pollutants Pollutants that have the Willingness to accept compensation (WTA) The
same impact on water quality regardless of where they minimum monetary compensation that an individual
originate. would be willing to accept to forgo something good, such
as an improvement in environmental quality, or to put up
Use value The utility that people derive from directly with something bad, such as a decrease in environmental
making use of a natural asset. quality.

Value of statistical life The monetary value of Willingness to pay (WTP) The maximum income
reducing the risks of death. that an individual would be willing to give up to gain
something good, such as an improvement in
Weak sustainability This requires a non-declining
environmental quality, or to avoid something bad, such as
total capital stock and so assumes that all forms of capital
a decrease in environmental quality.
(natural, human, social, and produced) are substitutes for
each other.
Index

A policy 249-57
abatement costs 164-5, 181 resilience 5, 241, 243-5
marginal (MAC) 26-33, 36, 215-16, 220-2 value 239-50
absolute advantage 12-13, 154 water quality 223, 225
access, to risk reduction 97-9 biological oxygen demand (BOD) 27, 32-3
acid rain 6, 36, 132, 146-9 bioprospecting 241
adaptation, to climate change 177-8, 188-9 biotechnology 240-1
adult literacy rates 110 birth rate 110-11
adverse selection, markets 22, 254 Bolivia 208
aesthetics, loss of 17 Brazil 196, 199, 207, 253
Agenda 21 132, 206 Bretton Woods Conference 165-6
agriculture 200-3 BSE 91
and climate change 111, 185 business as usual (BAU) 185, 189, 228
shifting cultivation 196, 200
water quality 217-18
air pollution 16 Cc
Alaska 246-8 California, pollution 56-7, 162
Allais paradox 90 California effect 162
Amazon rainforest 196, 199, 207 CAP (Common Agricultural Policy) 8, 23
amenities 5 cap-and-trade 31-7, 167, 182-3, 215-18, 220-2
aquatic ecosystems 6 capital. see also human capital; natural capital; produced
asymmetric information 254-5 capital; social capital
Australia 238-9 substitution 119-22, 126-8
autarky 154-5 types of 117-18
averting behaviour 97 carbon dioxide 164-5, 175-7
carbon footprint 188
carbon offsetting 208. see also cap-and-trade
B carbon sinks 4, 189-91
basic trade model 153-61 forests 194-6
BAU (business as usual) 185, 189, 228 Kyoto agreement 179
behaviour moorlands 79
averting 97 carbon tax 182-4. see also green taxes
climate change response 189 Carson, R. 60
context-dependent choices 82-3 catastrophe, climate change 187-9
hyperbolic discounting 94 CBA. see cost-benefit analysis
rational choice in practice 89-93 CBD (Convention on Biological Diversity) 250-2
rational choice theory 7, 84-8 CE. see choice experiment method
behavioural economics 83, 102-3 Centre Wildlife Care (CWC) 68
benefit transfer 59, 78-80 CFCs (chlorofluorocarbons) 147-9, 169, 179
Bernoulli, Daniel 85-6 characteristics theory of value 63-4
Bernoulli, Nicholas 84-5 children, exposure to pollution 100-1
bias, hypothetical 62, 67-8 China
bias, in surveys 62-3 climate change 178-9
bid vehicle 60-1 oil 275
bids, protest 61 chlorofluorocarbons (CFCs) 147-9, 169, 179
biodiversity 132, 233-6, 257-8 choice experiment (CE) method 63-8, 223-6, 240-50
climate change 189-90 cholera 185-6
conservation choices 236-9 CITES (Convention on International Trade in Endangered
forests 196-7 Species of Wild Fauna) 153, 168-9, 252-4

hotspots 240 classical economics 110-12

loss of 23, 252 Clean Air Act 36, 101-2, 162
292 INDEX

climate change 175-7 Country Stewardship Scheme 255

adaptation 177-8, 188-9 critical natural capital 119
biodiversity 235-6 crustal abundance 269
cost/benefit 77-8, 185-90 cultivation frontiers 111
impact on agriculture 111 cultural ecosystem services 47
international response 178-85 CVM (contingent valuation method) 60-3, 223-6, 250
trade-offs 190-2 CWC (Centre Wildlife Care) 68
Coase theorem 24-6
co-evolution 5-8
co-extinction 224 D
colonialism 111 Darwin, Charles 111
command-and-control intervention 23 death rates 110
Common Agricultural Policy (CAP) 8, 23 decentralized market 19
common property 15-18, 21, 132-6, 262. see also tragedy decision-making 131-2. see also behaviour
of the commons definitions
pollution 145-9 biodiversity 233-4
social capital 118 economic value 39-40
companies, reputation 102-3 economy 4
comparative advantage 12, 154-7, 161 environment 4
comparative risk analysis 100 glossary 286-90
compensating surplus (CS) 66 growth and development 105-8
compensation. see also willingness to accept compensation markets 11
(WTA) natural resources 263
Kaldor—Hicks compensation test 50 opportunity cost 7
Kyoto agreement 179 scarce 7
complete markets 13 sustainable development 116-19
composition effect 159 water pollution 212-13
conditional logit model 65-7 deforestation 194-203
conservation Easter Island 248-9
choices 63-4, 236-9 Reducing Emissions from Deforestation and Forest
forests 205-10 Degradation 190, 207-8
leading to poverty 203 demand and supply 274, 276-9
policy 250-8 democratic methods 49-50
consumer good 276 economic voting 55
consumerism, green 113-14, 124 development
context-dependent choices 82-3, 93 definition 105-8
contingent valuation method (CVM) 60-3, 223-6, 250 employment sectors 114
Convention on Biological Diversity (CBD) 250-2 Human Development Index 109-10
Convention on International Trade in Endangered Species diminishing returns. see trade-offs
of Wild Fauna (CITES) 153, 168-9, 252-4 direct benefits 45-6, 48
co-operative, game theory 143-5 discount rate 53, 200
corruption discounting 53, 94, 142-3
monetary rewards 102-3 disease spread 98, 185
political 114 disincentives 257
cost distribution of capital 107, 110, 112
extraction of resources 271-2 diversity. see biodiversity; genetic diversity
intertemporal 262 division of labour 154
labour 271-2 dolphins 167
ofliving 26 dose-response models 76-7
opportunity 39-40, 154-6 double dividend 26
of pollution 16 drugs 240-3
Costa Rica 241-3 Dubai 274
cost-benefit analysis (CBA) 49-57, 206-10
case study 52
climate change 77-8, 185-90 E
conservation 236-9 E. coli 186
risk reduction 100-1 Earth Summit, Rio 122-3, 132, 178
water quality 214-17, 223-4, 227-30 earth system thresholds 122-3
cost-effective measures 25-35, 182 Easter Island 248-9
count models 73-6 economic disarmament 187
INDEX 293

economic rent 272-3 policy 279-83

economic reserves 268 sources 111-12
economic value supply and demand 274-9
biodiversity 239-50, 257-8 trade-offs 100
definition 39-40 English Heritage 63
estimating 59 environmental Kuznets curve 112-16, 160-1, 164, 201-2
economic voting 55 Environmental Protection Agency (EPA) 27-35, 97-8
economics perspectives environmental quality. see also pollution
behavioural 83, 102-3 effect of growth 112-16
classical 110-12 international trading policy 165-70
neoclassical 112 relationship to social capital 118
economies of scale 114 trading policy 156-65
ecosystem externality 17 EPA (Environmental Protection Agency) 27-35, 97-8
ecosystem services 3,5 equilibrium, Lindahl 20
biodiversity 239-50 equilibrium, of climate 185-90
decline in value 250-1 equilibrium, of markets 14, 153-7
forests 194-7 housing 69-70
types 47-9 pollution 163-4
in UK 46-7 equilibrium, steady state 135-6
value 77-9 equity/inequity 107, 110, 116
ecosystems Escherichia coli 186
aquatic 6 European regulation 168
biodiversity 233-4 European Union (EU)
cycles 243-4 cap-and-trade 36
forests 194-6 carbon trading 182-3
fragmentation 194 Common Agricultural Policy 8, 23
mangroves 245 fishing quotas 143
marine 246-8 Kyoto agreement 179
eco-taxes. see green taxes water quality 229-30
ecotourism 257 eutrophication 217-18
Ecuador 16 evolution 233, 235-6
education 3-4, 9-10, 282-3 existence values 48, 62-3, 76, 205, 237, 249-50
efficiency/inefficiency 21-2 exogenous factors 6
decision-making 82-3 expected utility (EU) theory 86
energy 190 experimental auctions 97
of markets 12-14 exploitation. see overexploitation
net present value 54 externality
resource use 262 between countries 145-6
Egypt 6 in markets 15-17, 132
elasticity 274 extinction, of species 233-6, 257-8
elephants 253-4 extraction, non-renewable resources 263-7
emerging economies 178-80, 190-1 extreme events 189
emissions. see also greenhouse gases
modelling 76-7, 184
F
regulation 25-35
relationship to economic growth 112-16 FAO (Food and Agriculture Organization) 207-8
employment sectors 114 fat tails argument 189
endangered species 167-9, 233-6 Faustmann, Martin 203-4
choosing which to conserve 236-9 Fisher, A.C. 270-1
climate change 186 fishing industry
International Convention on Trade in Endangered Alaska 246-8
Species 153, 168-9, 252-4 game theory 135-45
USAMI2 impact on dolphins 167
Endangered Species Act 251, 255 mangroves 245
endogenous risk 88 overfishing 132
energy. see also non-renewable resources; renewable quotas 143
resources water quality 227
efficiency/inefficiency 190 flexibility/stringency trade-off 190-1
nuclear 91 flow pollution 177
paradox 83 Pood and Agriculture Organization (FAO) 207-8
294 INDEX

food production 6 green consumerism 113-14, 124

food-borne illness 98 green growth 3
forests 194-6. see also timber green net national product 123-8
Amazon rainforest 199, 207 green taxes 25-6, 83. see also marginal abatement costs
conservation policy 206-10 behavioural economics 102-3
distribution and losses 197-202 carbon tax 182-4
Easter Island 248-9 international agreements 167-70
management 203-6 questionnaire 60-2
Poland 250 UK 55-6
tropical 195-6, 202-3 greenhouse gases 164-5, 175-7, 184, 189-90
value 64 forest stores 195-6
fossil fuels 175-7, 182-4, 186 Kyoto agreement 179
fragmentation, of ecosystems 194 gross domestic product (GDP) 106-8
free market 7 climate change 185-6
as myth 119 trade 152, 161-2
free-riding 19-20 gross national product (GNP) 105-9, 124-6, 199-201
climate change 178-81 growth
contingent valuation method 62 benefits of 8
free trade 151-8 definition 105-8
income 161 environmental effects 112-16
frontier model 197, 199-202 green 3
future scenarios population 110-11
climate change 185-90 sustainable development 105-10
energy 279-83 trade 151-3
resources 267-76 Gulf of Mexico 245

G H
gambling. see risk hantavirus 186
game theory 8, 131-9, 145-51 Hardin, G. 15-17, 19-20, 139-43
climate change 178-81 Hartwick rule 120-1, 127
co-operative 143-5 health hazards 88-9
repeated fishing 140-3 climate change 177-8, 185-6
GDP. see gross domestic product hedonic pricing method (HPM) 67-71, 96
General Agreement on Tariffs and Trade (GATT) 166-7 herbal medicine 240-3
general equilibrium model 153, 157, 246-8 heritage sites 63
genetic diversity 235, 240-1. see also biodiversity heuristic methods, risk 82, 89
Genuine Savings 123, 125-8 Hicks, Sir John 124
geochemical scarcity/abundance 269 hidden information markets 22
Global Environmental Facility (GEF) 207-8 historical perspectives 110-12
global warming. see climate change HIV (human immunodeficiency virus) 186
globalization 4, 8 Hotelling’s rule 263-7, 270-6, 278
environmental 146-9 hotspots, of biodiversity 240
pollution 145-9 housing market 69-70
trade 15] HPM (hedonic pricing method) 67-71, 96
GNP (gross national product) 105-9, 124-6, 199-201 Hubbert, M. K. 275
government. see also European Union; international human capital 96, 117-18, 127-8
agreements Human Development Index (HDI) 109-10
economic voting 55 hunting 253-4
national accounting 57 hydropower 56
self-governance 139-40 hyperbolic discounting 94
government intervention 8. see also policy; international hypothetical market bias 62, 67-8
agreements
biodiversity 250-3
climate change 178-85 I
energy 272, 279-83 ice sheet, West Antarctic 187
failure 23, 209-10, 257 Iceland 132
forests 206-10 IEA (international environmental agreements) 146-9
trade 165-70 IFF (Intergovernmental Forum on Forests) 206-7
grandfathering, permits 34 IMF (International Monetary Fund) 165-6
INDEX 295

immiserization model 197, 200-2 Kh. see human capital

incentives, environmental protection 16, 257 Kn. see natural capital
climate change 179 Kp. see produced capital
Common Agricultural Policy 8, 23 Krutilla—Fisher model 205
corruption 102-3 Ks, see social capital
flexibility 189-91 Kuznets curve 112-16, 160-1, 164, 201-2
international environmental agreements 146 Kyoto agreement 178-83, 190-1
lack of 250-1 carbon footprint 188
monetary 7, 220, 253, 280-2 costs of meeting targets 36-7
perverse 257
water quality 215
income 161-2 L
inequality 107, 110, 112 labour costs 271-2
India 178 laissez-faire policy 162
indifference curves 42-4 land use conversion 194-6, 235-6
indirect benefits 45 limits to growth school 113
Industrial Revolution 6 Lindahl equilibrium 20
industrialization 113 literacy rates 110
inefficiency/efficiency. see efficiency/inefhiciency living standards 107
inequity/equity 107, 110, 116
infant mortality 110
information rents 254 M
insurance, climate change 177-9, 186 MAC. see marginal abatement costs
interdependence 3-8, 44-5, 103, 110-11, 246-8 mad cow disease 91
Intergovernmental Forum on Forests (IFF) 206-7 malaria 185-6
Intergovernmental Panel on Climate Change (IPCC) Malawi, Kasungu National Park 257
175, 178 Malthus, T. 110-11, 200, 262
interlinked systems 3-8, 44-5, 103, 110-11, 246-8 Malthusian perspectives 197, 200-2, 248
international agreements management. see also regulation
biodiversity 250-3 biodiversity 249-57
climate change 178-85, 188 of risk 99-103
energy 279-83 water quality 223-9
forests 206-10 mangroves 245
trade 165-9 marginal abatement costs (MAC) 26-33, 36, 215-16,
International Convention on Trade in Endangered Species 220-2. see also abatement costs
(CITES) 153, 168-9, 252-4 marginal damage (MD) 160
International Energy Agency 267 marine ecosystems 246-8
international environmental agreements (IEA) 146-9 markets 8-9, 11-15, 36-7
International Monetary Fund (IMF) 165-6 adverse selection 22, 254
International Trade Organization (ITO) 165-6 definition 11
intertemporal choices 94 and environment 23-6
intertemporal costs 262 failure 15-23, 214-15
intervention. see government intervention free market 7
invasive species 6 green taxes 25-35
invisible hand 7-8, 12, 20, 112 Hotelling rule 263-7, 270-6, 278
IPCC (Intergovernmental Panel on Climate Change) housing 69-70
175, 178 missing 15-16, 239-40
irrigation 6 pollution regulation 12
ITO (International Trade Organization) 165-6 property rights 13
Ivory Coast 200-1 Marrakech Agreement 151-2
ivory trade 253-4 mass extinction 233, 257-8
maximum sustainable yield 135-6
Mayflower problem 271, 273
J MD (marginal damage) 160
Jevons, W.S. 111-12, 262 medicinal uses of plants 240-3
Mexico 167
Gulf of 245
K microeconomics 272
Kaldor—Hicks compensation test 50 Mill, J.S. 111-12
Keynes, John Maynard 111 Millennium Development Goals 202-3
296 INDEX

Millennium Ecosystem Assessment 3-5, 46-9, 243, 251 non-renewable resources 8, 177-8, 262-3, 283-4
mineralogical threshold 269-70 demand and supply 276-9
minerals 263-4, 268-70 extraction 263-7
mining 271-2 forests 205
missing markets 15-16, 239-40 measuring scarcity 267-76
Mitchell, R. 60 policy 279-83
mitigation, of climate change 177-8, 186, 188-9 types of 263
models non-rival consumption 18-19
basic trade 153-61 non-use values 48, 62-3, 76, 205, 237, 249-50
climate change 77-8, 185-8 normative perspectives 131
Conditional Logit 65-7 North America, cost-benefit analysis (CBA) 50. see also
deforestation 197-202 United States
dose-response 76-7 North Sea oil 268-9
emissions 76-7, 184 NPV (net present value) 54
environmental Kuznets curve 112-16 nuclear power 91
fish stocks 135-9
frontier 197, 199-202
Hotelling rule 263-7, 270-6, 278 Oo
immiserization 197, 200-2 OECD (Organisation for Economic Co-operation and
Krutilla-Fisher 205 Development) 114-15, 184, 188, 274
marginal abatement cost curve 26-33 offsetting 120, 184, 208
natural capital 124-8 oil 263-70, 272, 274-6
random utility 73-6 pollution 56-7
rational choice 84-8 opportunity approach 117
species preservation 236-9 opportunity cost 39-40, 154-6
species value 241-3 optimal level, within markets 20
travel-cost 70-6, 226 option price 95-6
monetary incentives 16, 102-3 Organisation for Economic Co-operation and
monopolies 263-7 Development (OECD) 114-15, 184, 188, 274
Montreal Protocol 146-7, 168-9 Ostrom, E. 140
moorlands 79 outcome approach 116-17
moral hazard 22, 254 outdoor recreation 70-4
morbidity 110 overexploitation 6, 119-20
Morgenstern, O. 131 common property 21, 132-6, 262. see also tragedy of
mortality 110 the commons
most-favoured nation (MEN) clause 166 overfishing 132. see also fishing industry
multilateral agreements. see international agreements ozone depletion 146-9

N P
Nash equilibrium 136-40 Pareto optimal 139
national accounting 57 perception
national parks 70 of risk 88-91, 189
National Trust 63 environmental value 228, 249-50
natural capital 8-9, 117-23 perception gap 91
forests 194-6, 205 perverse incentives 257
modelling 124-8 Pigou, Alfred 25
natural resources. see resources Pigovian taxes 25. see also green taxes
neoclassical economics 112 planetary boundaries 122-3
neoclassical model of growth 116-17 point-source pollution 212, 214, 219-20, 221-3
Neolithic society 6 Poland 250
net present value (NPV) 54 policy 6-8. see also government intervention;
Noah's library 236 international agreements; regulation
noise pollution measures 71 biodiversity 250-8
non-discrimination, tariffs 166 Clean Air Act 36, 101-2, 162
non-excludability 18-19 climate change 178-85, 188, 190-1
non-market services 4 deforestation 199
non-point source pollution 214, 217-22 energy 279-83
definition 212-13 environmental quality 156-65
trading policy 221-2 forests 206-10
INDEX 297

laissez-faire 162
mad cow disease 91
race-to-the-bottom hypothesis 162
trade 152-6
random utility models 73-6
policy appraisal 55
rational choice in practice 89-93
policy-as-lottery 82-3
rational choice theory 7, 40, 82-8, 94
policy-makers
rebound effect 282
conservation choices 236-9
reciprocity, tariffs 166
estimating environment value 79
recreation 70,79
political corruption 114
Reducing Emissions from Deforestation and Forest
polity 162
Degradation (REDD) 190, 207-8
pollution 5, 15-17, 22. see also greenhouse gases; waste;
regulating ecosystem services 47
water pollution
forests 195-6
effect on children 100-1
regulation, effect on environment 156-65
fossil fuels 186
regulation, to protect environment 82-3. see also green
international agreements 8, 165-70
taxes; policy
markets 12, 163-4
climate change 178-85
noise 71
of energy 279-83
oil 56-7
European 168
optimal level 157-61
international agreements 165-70
regulation 25-35
of markets 12, 24-5, 119
relationship to economic growth 112-16 of risk 99-103
relationship to social capital 118 sustainable development 119-22
stock/flow 177 trade 156-65
trading policy 152-3, 156-65 water quality 215
transboundary/international 145-9 regulatory chill 162
pollution havens hypothesis 152, 162-5, 167-70, 188 renewable resources 8, 177-8, 280-2
population increase 110-11 forests 205
Porter hypothesis 162-3 hydropower 56
positive perspectives 131 repeated fishing, game theory 140-3
poverty 202-3 reputation, of firms 102-3
PPF (production possibility frontier) 157-9 resilience, increased by biodiversity 5, 241, 243-5
prediction. see future scenarios resource-dependent countries 121
preference heterogeneity 46 resources 4-5, 7-8, 109, 117. see also non-renewable
present value (PV) 51, 53-4 resources; renewable resources
primordial poverty 202-3 allocation 51, 236-9
prisoner's dilemma 133-4, 178-81 demand and supply 276-9
probability. see risk economic value 39-40
produced capital 117-18 efficiency/inefficiency 262
production 109 extraction 263-7, 271-2
factor of 276 forests 205
specialization 12-13 lifetime 270-1
production possibility frontier (PPF) 157-9 limited base 248-9
production-function approaches 59-60, 76-8 measuring scarcity 267-76
project appraisal 55 policy 279-83
property rights 13, 24-5, 149, 262 relationship to economic growth 112-16
protection of environment 9, 82-3, 103-4 threat from climate change 177-8
cost effective 25-35 use thresholds 121-3
rational behaviour in practice 89-93 revealed-preference approaches 59, 67, 96, 225-7, 230
rational behaviour in theory 7, 40, 82-8, 94 hedonic pricing method 67-71, 227
reducing risk 93-9 travel-cost models 70-6, 226
regulating risk 99-103 Ricardo, D. 111-12, 151, 248, 262
protest bids 61 rights. see property rights
provisioning ecosystem services 47 Rio Earth Summit 122-3, 132, 178
public goods 18-20 risk
PV (present value) 51, 53-4 ambiguity 92
assessment 87-8
climate change 177-8, 189, 192
Q economic cost 16
quotas 151-3 environmental policy 82-3
fishing industry 143 heuristic methods 82, 89
298 INDEX

risk (continued) substitute sites 73-6

perception 88-91 sulphur dioxide 162-3
rational choice theory 84-8 supercrops 241
reduction 93-9 supply and demand 274, 276-9
regulation 99-103 support ecosystem services 47
rivers 223-5. see also water quality surplus dose-response models 76-7
as study sites 78-9 sustainable development 3, 128, 206-7
Thames 213-14 definition 116-19
Royal Society for the Protection of Birds (RSPB) 20 fishing industry 135-9
RP approaches. see revealed-preference approaches growth 105-10
Kuznets curve 112-16
maximizing yields 8
s measures 122-8
safe minimum standard (SMS) 120-2 opportunity approach 117
safe operating space 122 outcome approach 116-17
scale effect 113, 158-9 predictions from the past 110-12
scarcity 7-8, 267-76 regulations 119-22
Scotland Sweden 109
sustainable development 125 System of National Accounts 124-6
water quality 214, 218 systems, interdependence 3-8, 44—5, 103, 110-11,
SD. see sustainable development 246-8
sedentary agriculture 202-3
self-interested marketism. see markets
services. see ecosystem services i
shadow projects 120 tariffs 151-3, 166
shifting cultivation 196, 200, 202 taxation. see also green taxes
silicates 269 carbon 182-4
slash-and-burn agriculture 196 energy 280-3
Smith, A. 151, 154, 262, 283 of environmental degradation 25-35
invisible hand 7-8, 12, 20, 112 perception 102-3
SMS (safe minimum standard) 120-2 water quality 215-21
social capital 117-18 willingness to pay 40
social preferences 102-3 technique effect 114, 159, 164
soil erosion 197 technological advance
Spain 70 climate friendly 190
specialization, market 12-13 energy 282
species technology-based standards
endangered 12 of risk 100
invasive 6 temperature scenarios 185-7
preservation 236-9 Thames, River 213-14
species richness. see biodiversity thresholds 115-16
species—area curve 234-5 climate change 187-9
standards ofliving 107 earth system 122-3
stated-preference approaches 59-60, 96-7 mineralogical 269-70
choice experiment method 63-8, 223-6, 240-50 resource use 121-3
climate change 187 timber 194-204, see also forests
contingent valuation method 60-3, 223-6, 250 tipping points 187-9
steady state equilibrium 135-6 tourism
Stern Review on the Economics of Climate Change 186 ecotourism 257
stock pollution 177 UK 281-2
Stonehenge 63 tradable permit systems 31-7, 167, 182-3, 215-18
strategic interactions 131-2, 149-50 European Union 36
co-operative games 143-5 trade 11, 151-3, 169-70. see also cap-and-trade
game theory 133-9, 145-9 and climate change policy 188
repeated fishing games 140-3 environmental effects 157-65
self-governance 139-40 international gain 153-7
stringency/flexibility trade-off 190-1 policy 165-9
strong sustainability 119-20 risk 84
subsistence agriculture 201 trade intensity 164
INDEX 299

trade liberalization 151-3, 158, 164 perception 228, 249-50

trade-offs. see also willingness to pay production-function approaches 76-8
climate change 190-2 revealed-preference approaches 67-76
risk 83-6, 96, 100 stated-preference approaches 59-67
tragedy of thecommons 15-17, 19-20, 139-43 water quality 223-9
transboundary agreements. see international agreements willingness to accept compensation vs. willingness to
Transboundary Movements of Hazardous Waste 153, 169 pay 40-4
transitional economies 178-80, 190-1 value of statistical life (VSL) 95, 97-8
transparency, tariffs 166 values
travel-cost models 70-6, 226 existence/non-use 48, 62-3, 76, 205, 237, 249-50
tropical forests 195-6, 202-3 leisure time 73-6
Turkey 141 present value 51, 53-4
risk reduction 93-9
use 48, 76
U voluntary institutions, managing common property
United Kingdom (UK) 140-1
biodiversity 23 von Neumann, J. 86, 131
Country Stewardship Scheme 255 VSL (value of statistical life) 95, 97-8
ecosystem services 46-7
effects of Common Agricultural Policy 23
fishing industry 132 Ww
green taxes 182 waste 45. see also pollution
renewable resources 281-2 disposal 39-40, 169
rivers as study sites 78-9 sink 4-5
water quality 213-14, 228-30 Transboundary Movements of Hazardous Waste 153,
United Nations Conference on Environment and 169
Development (UNCED) 206 water pollution 212-14
United Nations Framework Convention on Climate benefits of improvement 223-7
Change (UNFCC) 179 cost-benefit analysis 227-30
United States of America (USA) costs of control 214-17
cap-and-trade 36 non-point source 217-22
choice experiment 224 water quality 57, 215, 223-9
climate change 178-9 agriculture 217-18
endangered species 12 cost-benefit analysis 227-30
Endangered Species Act 251 European Union 229-30
fishing industry 167 moorlands 79
national parks 70 weak sustainability 119-20, 126-8
oil 274-5 Weitzman, M.L. 124, 236-9
stated-preference evidence 187 welfare economics 93-9, 228. see cost-benefit analysis
water quality 226 well-being 50-1, 107-8, 112
wildlife 255-6 West Antarctic ice sheet 187
United States Marine Mammal Protection Act 167 wetlands 48-9, 76-7
unpriced services 4 WHO (World Health Organization) 16, 98
use values 48, 76 willingness to accept compensation (WTA) 14-15, 40-4,
utility theory 86, 89-93 46, 60-2
willingness to pay (WTP) 14-15, 39-44, 124-5, 160-1
benefits transfer 78-80
Vv biodiversity 249-50
valuation methods 59, 79, 96-7, 230. see also production- case study 46
function approaches; stated-preference approaches climate change 187
water quality 223-9 hypothetical bias 68
value, environmental 5, 39-49, 57-9 questionnaire 60-2, 65-6
benefits transfer 78-80 renewable energy 280-2
biodiversity 239-50 risk reduction 96-9
conservation 236-9 water quality 226-9
cost-benefit analysis 49-57 World Bank
decline 250-1 genuine savings 125-8
forests 64, 194-6 project appraisal 55
heritage 63, 70 Where is the Wealth ofNations? 121
300 INDEX

World Health Organization (WHO) Y

16, 98 yield maximization 8, 135-6
world rankings
by energy consumption 274-9
by GDP 106-10 7
World Trade Organization (WTO) 151-3, 166-7
zero net investment rule 120
WTA. see willingness to accept compensation
WTP. see willingness to pay zero risk 99
& The writing style is of a high quality and is pitched at the appropriate level for an introduction
to Environmental Economics. Economic concepts are introduced using easy-to-understand
language and examples. I Dr Mike Christie, Aberystwyth =

examines a range of [Link] studies and fit is based heavily on empirical evidence
Flare also provides an extensive number
assesses a variety of economic theories and their
approach to environmental issues. Clearly written, global of examples, which explain well the
Tae: |) °)ces-(e1amur-lalom (al-xe)
ali(er-li\"me)cer-tomianliave|-ve MMtallom (-)4a[s) definitions mentioned in the text. J
M-la le l-r-1Mlajicore|U(o((olam com ale )Wm-Yero)atoltal(er-Mor-1aM of-MUy-\0 co) Leon Vinokur, Queen Mary, University of London
fe)coho[Manto] c-M=lanVi[celalnat=vae=1Mo)co)<-tel ((o)aie-\ al(=\\-Morel—)

An understanding of economics is vital to. any Ca online resource centre

understanding of why environmental problems occur. — [Link]/orc/hanley2e/
Elato M'\/al-1mof=1:]mu Come (olr-lolel0] miai-\gahim
MalicMefole)alla) icere[Ue(-1<)
For students:
the reader to the key ideas underlying environmental
economics without assuming any in-depth prior Links to news stories and
knowledge of economics. Part One explains the fo)cef=lal raelte)ats
fundamental economic concepts. Part Two uses these Exercises
fofo]atoi-Jo]cmCoM) alo(-i¢-ir-lale-lale Me(-)V/-)(6)ofefe)| (es t-\-) oe)aty= (0)
Datasets
the major environmental issues of our time. Examples are
drawn from all over the world and include such vital issues For registered adopters:
as Climate change, pollution, and loss of biodiversity. Suggested solutions to exercises
in the text
Clr-Te) al-Mico)gam tal-M eele)4
Building on the success 3 Of the firstTeli (olan the second
Telid(eyame)i ore | WiiKey}
features the following pada vrualatiel!
Nick Hanley, Professor of
Substantial revision of all existing chapters Environmental Economics, University
Greater focus on environmental policy and of Stirling, Jason Shogren, Stroock

management links, and links to policy issues that have Professor of Natural Resource
recently emerged Conservation and Management,
University of Wyoming, and Ben
Updating of box sections with recent findings fro
the literature Intro to Environmental Economics ©
Schwan/
Martin
(top)
(bottom
Skubisz
Piotr
fae:

New pedagogical features Including tutorial ques

and definitions

lt
978 -0-19 -956873-4

OXFORD
UNIVERSITY PRESS
2" 966199"568733
[Link]
00000 RSRV

Common questions