Phil. Trans. R. Soc.

B (2010) 365, 2769–2777

doi:10.1098/rstb.2010.0180

Introduction

The future of the global food system

H. Charles J. Godfray1, *, Ian R. Crute2, Lawrence Haddad3,
David Lawrence4, James F. Muir5, Nicholas Nisbett6, Jules Pretty7,
Sherman Robinson3, Camilla Toulmin8 and Rosalind Whiteley6
1
Department of Zoology and Institute of Biodiversity at the James Martin 21st Century School,
University of Oxford, South Parks Road, Oxford OX1 3PS, UK
2
Agriculture and Horticulture Development Board, Stoneleigh Park, Kenilworth,
Warwickshire CV8 2TL, UK
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

3
Institute of Development Studies, Falmer, Brighton BN1 9RE, UK
4
Syngenta AG, PO Box CH-4002, Basel, Switzerland
5
Institute of Aquaculture, University of Stirling, Stirling, Stirlingshire FK9 4LA, UK
6
Foresight, UK Government Office for Science, 1 Victoria Street, London SW1H OET, UK
7
Department of Biological Sciences, University of Essex, Wivenhoe Park, Colchester, Essex CO4 3SQ, UK
8
International Institute for Environment and Development, 3 Endsleigh Street, London WC1H 0DD, UK
Although food prices in major world markets are at or near a historical low, there is increasing con-
cern about food security—the ability of the world to provide healthy and environmentally
sustainable diets for all its peoples. This article is an introduction to a collection of reviews
whose authors were asked to explore the major drivers affecting the food system between now
and 2050. A first set of papers explores the main factors affecting the demand for food (population
growth, changes in consumption patterns, the effects on the food system of urbanization and the
importance of understanding income distributions) with a second examining trends in future
food supply (crops, livestock, fisheries and aquaculture, and ‘wild food’). A third set explores
exogenous factors affecting the food system (climate change, competition for water, energy and
land, and how agriculture depends on and provides ecosystem services), while the final set explores
cross-cutting themes (food system economics, food wastage and links with health). Two of the clear-
est conclusions that emerge from the collected papers are that major advances in sustainable food
production and availability can be achieved with the concerted application of current technologies
(given sufficient political will), and the importance of investing in research sooner rather than later
to enable the food system to cope with both known and unknown challenges in the coming decades.
Keywords: food security; food system; population growth; consumption growth;
agriculture; climate change

1. INTRODUCTION equal number are over-fed. Looking ahead, we can

The supply and availability of food has been a crucial identify known threats to the food system and factors
factor shaping the emergence, development and per- that will increase the risks of a rise in hunger. Population
sistence of human civilizations throughout the ages. and consumption growth will lead to the demand for
For the last few decades food has been cheaper in food increasing for most of the current century, while
real terms, and more readily available, than probably increasing competition for land, water and other
at any time in history, which partly explains why resources threaten the supply of food. Overarching this
food policy has received less prominence in national is the threat of global change, and the needs to make
and international decision-making than in earlier the food system resilient to shocks that cannot be pre-
times. Yet, we cannot be said to have a functioning dicted in advance. As the reactions to the 2008 spike
global food system when one in seven people today in commodity prices presaged, food policy is likely to
still do not have access to sufficient food, and an increase in importance in the coming decades.
The authors of the reviews in this issue were asked
to explore the main drivers of change affecting the
* Author for correspondence ([email protected]). global food system between now and 2050. The
While the Government Office for Science commissioned this review,
reviews were commissioned as part of the UK’s Gov-
the views are those of the author(s), are independent of Government, ernment Office for Science Foresight project on
and do not constitute Government policy. Global Food and Farming Futures, which will report in
One contribution of 23 to a Theme Issue ‘Food security: feeding the late 2010.1 The project is based around the organizing
world in 2050’. question of how a future global population of 9 billion

2769 This journal is q 2010 The Royal Society

 2770 H. C. J. Godfray et al. Introduction. The future of the global food system

people can all be fed healthily and sustainably, and the Nations Population Division projections with the
reviews contribute towards an evidence base under- ‘medium fertility’ assumption (United Nations
lying an exploration of the policy options to address 2004). However, they argue that the UN projection
these challenges. Science has a major role to play in of approximately 1.8 children per adult female overes-
feeding the world, as the green revolution amply timates fertility in China which is more likely to be 1.5
demonstrated. But questions of food security require (though not as low as the official estimates of 1.2):
a multi-disciplinary approach involving the social China is big enough that this assumption makes a
sciences and economics; hence, the authors of the difference. With this adjustment, global population
reviews are drawn from a broad range of disciplines. growth is predicted to decelerate and reach just over
We have organized the reviews into four sections, 9 billion in 2050. There are marked regional vari-
though there is some overlap. First we explore factors ations: Europe’s population will decline, Africa’s will
affecting the demand for food with four reviews on double, while China will peak in about 2030 and be
population growth, changes in consumption patterns, overtaken by India around 2020. Populations will age
the effects of urbanization on the food system and almost everywhere, but as the old will be healthier,
the importance of understanding income distributions. rethinking age in terms of time to expected death
The second section examines the supply side of the (rather than time since birth) may give a different
equation with reviews of likely changes in crop and and more positive perspective on increased longevity.
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

livestock production, and the different components Increasing demand for food is caused not only by a
of aquatic food—marine capture, freshwater capture rise in population size but, as Kearney (2010)
and aquaculture. The third section explores exogenous explores, a rise in per capita consumption. As people
factors affecting the food system: the possible effects of who are initially undernourished obtain access to
climate change, competition for water, energy and more food calories, they first go through an expansion
land, and agriculture’s dependency on, and provision phase where diets contain more food—typically,
of, ecosystem services. The final section explores grains, roots, tubers and pulses—and then a substi-
cross-cutting themes, many involving economics: tution phase, where the latter are replaced by more
what is the relationship between productivity and energy-rich foods such as meat and those with a high
investment in research and extension? How do we concentration of vegetable oils and sugar. The result
model the food system? What are the consequences is the nutritional transition (Popkin 1998), which has
of globalization and of volatility? This section also con- major implications for food supply as typically the pro-
siders the issue of reducing waste in the food system duction of high-energy food requires more resources
and the interrelationship between diet and health. (for example, instead of grain being directly consumed
by humans, it is used as animal feed for livestock pro-
duction which is then consumed by humans, overall a
2. DRIVERS OF DEMAND more inefficient process). Increased consumption of
Understanding trends in population size are critical to high-energy foods can increase the risk of obesity
estimating the future demand for food. Lutz & Samir and the chronic diseases associated with being over-
(2010) review how reliable population projections are weight: indeed, some countries that are still coping
now constructed. Populations in different countries with under-nutrition in parts of their population are
are assumed to be composed of different age cohorts now suffering the additional burden of over-nutrition.
of the two sexes that vary in demographic rates such The overall pattern of the nutritional transition hides
as mortality and fertility. Models can be extended to many interesting local variations. For example, while
include differences between rural and urban popu- China has seen a very strong increase in the consump-
lations (connected by migration) and, most tion of high-energy foods, in India for cultural and
importantly, educational status. There is very convin- religious reasons, the rise has been much less
cing evidence of the critical importance of female marked, for equivalent levels of income.
education and access to contraception in causally Global dietary patterns are also being influenced by
affecting fertility, and these are probably the chief a complex web of socio-economic trends and drivers.
mechanisms behind the decline in fertility as countries On the demand side, more and more people live in
develop economically and go through the demographic cities where they have relatively sedentary occupations
transition. Of particular relevance here is evidence that and often have relatively high disposable incomes.
education rates are also negatively correlated with On the supply side, economic growth, regulatory
malnutrition and food insecurity (Lutz et al. 2004). liberalization, the encouragement of foreign direct
How to deal with uncertainty is a perennial problem investment and globalization in general has allowed
in population and other forecasting and of the four strat- a burgeoning fast-food and supermarket sector to
egies listed: ignoring it, constructing scenarios; develop. As Kearney describes, in the 10 years
exploring a plausible range of variation; and making between 1990 and 2000, the service and retail sectors
fully probabilistic projections; Lutz & Samir strongly in Latin America made changes that had previously
argue for the last. Though there is a risk that probabilis- taken 50 years in North America, and much of Asia
tic projections appear spuriously precise, all assumptions and Eastern Europe are only a few years behind.
are made rigorously specific. Interestingly, imprecision This increased economic activity in the food sector
about the state of current populations can be as big a brings advantages such as employment and investment
source of error as uncertainty about the future. opportunities, and often increases the availability and
In studying future food demands, Lutz & Samir safety of the food on offer to its consumers. But by
recommend adopting, with some caveats, the United making cheap foods rich in fats and sugars easily
Phil. Trans. R. Soc. B (2010)
 Introduction. The future of the global food system H. C. J. Godfray et al. 2771

available (many processed foods are 30% fat), there Cirera & Masset (2010) explore how these pro-
are important health implications. Significantly, blems have been tackled in the economics literature
South Korea, which has vigorously promoted local and how they have been incorporated into the major
foods rather than a western diet, has lower rates of general and partial equilibrium models that are used
obesity than similar countries (Kennedy et al. 2004). to model future food supply. They show how income
The implications of urbanization are explored in disparity can affect estimates of future food demand,
more detail by Satterthwaite et al. (2010). They the nub of the aggregation problem. Most major
point out that around 1940, the global economic models do not explicitly allow for this variation. To
value of industry and services for the first time illustrate their arguments, the authors explore how
exceeded that of the primary sector (food production, changes in income distribution affect food demand
forestry and mining), and that by about 1980 more in an Indian state, and then cautiously extrapolate
people were employed in the former than the latter. their results to global food supply. Not taking into
Industry and services are concentrated in cities and by account income distribution may affect estimates of
about 2008 more people on the Earth lived in cities food demand, though perhaps by only approximately
than in rural areas. This trend towards urbanization is +10 per cent.
certain to continue and the last few decades have seen
the rise of megacities in developing countries, with
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

Mumbai, Sao Paulo and Mexico City all having more 3. SUPPLY-SIDE DRIVERS
than 16 million inhabitants. Urbanization is strongly Crop yields have improved dramatically over the last
associated with increasing wealth, and sufficiently 50 years, but there is evidence that rates of increase
advanced logistics and infrastructure are essential to have declined in recent times. Jaggard et al. (2010)
feed the inhabitants in very large conurbations. ask if yields in industrialized countries have reached
Increasing urbanization and urban spread has a a ceiling and conclude that this is not the case; they
direct, though sometimes, exaggerated effect on the consider how yields may increase both in the short
land available for agriculture but many more indirect and long term. They also explore the yield gap; what
effects. Urban populations can access a greater diver- is actually achieved against the best benchmark for a
sity of foods, though this may include meat, dairy particular crop in a particular region. In efficiently
and convenience foods—types of food that may have run operations in industrialized countries, the gap
required more resources for their production or be may only be 20 per cent, and there may only be
less healthy. However, some studies suggest that weak economic incentives to improve yields. However,
income rather than urban or rural location is normally there is considerable variation among different farming
the primary determinant of diet. Urbanization can also operations, with some—for complicated social and
have very positive effects on rural areas and food pro- economic reasons—being very inefficient, and of
duction, in general by increasing national wealth and course great scope for yield increases in many develop-
more specifically by creating markets for food produ- ing countries, especially through increased crop
cers. In developing countries in particular there are fertilization.
often strong financial links between people living in Jaggard and co-workers also review attempts to pre-
cities and the countryside, with remittances from dict crop yields in the future by combining different
urban households financing innovation and yield types of global climate models (GCMs) with crop
growth in farming. growth models. Results to date are informative but
To accurately predict a nation’s food demand, it is not consistent. The nature and extent of CO2 fertiliza-
important to know the full distribution of per capita tion (see also Gornall et al. 2010) and in general how
income and how this is reflected in food purchases, crops respond to climate change is insufficiently
an area of active research in behavioural economics. understood, and while GCMs tend to agree broadly
There is a lower limit to the amount of food an about how increased greenhouse gas concentrations
individual can eat without starving to death and an will lead to rises in temperatures, there is less agree-
upper limit determined by our physiology. These ment about which regions will get more or less rain,
biological facts underlie Engel’s law, which states something that is particularly critical in predicting
that as income increases the proportion spent on yields. The authors select a series of crops in 17 differ-
food declines. There are numerous challenges to ent countries and ask whether the goal of producing
estimating this relationship: moving from micro- to substantially more food in 2050 is feasible, given
macro-economic description—from the behaviour reasonable assumptions about rates of technological
of individuals to the aggregate properties of a advance, efforts to close the yield gap, climate, CO2
population—is complicated by the nonlinearity of the fertilization and (often ignored) ozone pollution.
Engel curve, what economists call the aggregation pro- Their conclusion, hedged with important caveats
blem. One also cannot simply work with per capita about the challenges ahead, is cautiously positive.
behaviours, as household size and in particular the How might we increase the supply of meat and milk
number of children affects the income – food demand to match burgeoning demand? A variety of strategies
relationship. Not all food requires the same resources are explored in the review by Thornton (2010).
for its production and to understand the full conse- Conventional animal breeding is still capable of
quence of increased wealth, there is a need to couple increasing yields, and will be important in addressing
Engel’s Law with Bennett’s Law, the latter describing other goals such as sustainability and better
the shift from starchy staples to more fatty foods as welfare. Modern genomic approaches to breeding
people get richer. will undoubtedly produce further gains, perhaps
Phil. Trans. R. Soc. B (2010)
 2772 H. C. J. Godfray et al. Introduction. The future of the global food system

supplemented by the prospect of genetic modification. Economic Zones. There are particularly complex gov-
Many advances have involved novel crosses, and pre- ernance issues for fisheries exploited largely by poor
serving rare breeds may be a valuable investment for people working in small groups. A better harmoniza-
the future. We have a much better understanding of tion of fisheries and ecosystem management will help
animal nutrition than in the past, but further research protect stocks and there is an important role for non-
is required to develop robust animal growth models to governmental agencies and civil society to champion
help optimize livestock production. Poor nutrition is sustainability in marine fisheries in order to help gov-
a particular problem in developing countries, ernments make difficult decisions, which may have
where livestock often represent a critical component unpopular political and socio-economic ramifications.
of household and community capital. Thornton Inland capture fisheries—the fish and crustaceans
describes a series of important new ideas that are harvested from rivers, lakes, floodplains and
specifically designed to benefit the nutritional status lagoons—are of major importance to many commu-
of livestock kept by very poor communities. In devel- nities, especially those in low-income nations.
oped countries there has been a general decline in Around 15 – 20% of global aquatic food is produced
endemic diseases, although major epidemics, includ- in this sector, though there is a widespread recognition
ing new emergent diseases, continue to be a major that its significance to the economy and food security
threat. Less progress has been made in the tropics, is underestimated because of under-reporting and
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

although with some success such as the probable the dominance of small-scale fishing operations. In
eradication of rinderpest. Animal breeding and low-income countries the food obtained from inland
veterinary advances, as well as better diagnosis and waters tends to be of direct importance as food,
surveillance, will all help farmers keep pace with while in more wealthy counties recreational fishing
evolving pathogens and hopefully reduce the burden has come to dominate.
of disease. Welcomme et al. (2010) explore the threats and
Livestock production is responsible for a significant opportunities for inland capture fisheries. The most
but contested fraction of anthropogenic greenhouse important challenges involve changes to water systems
emissions (e.g. FAO 2006) and will be required to with increasing demands for water from the agricul-
contribute towards mitigation efforts. Switching from tural, domestic, industrial and energy production
ruminants to monogastric livestock may help, as will sectors. Climate change is likely to exacerbate the pro-
technological advances in how intensively maintained blems. Integrated water management policies are
animals are reared. In more extensive systems it may essential to try to balance these competing interests.
be possible to develop systems that both capture A range of management, mitigation and enhancement
more carbon and provide more feedstock. Maintaining strategies are available, for example, enhancement of
viable livestock production will be critical in climate- wild stocks by reared fish, but their success depends
change adaptation, especially for very many poor critically on the political and institutional context of
smallholders whose animals are central to their liveli- the particular fishery. There is a continuum between
hoods. In some of the most marginal agricultural unmanaged capture fisheries through increasing inter-
areas we may even see an increase in pastoralism and ventions to full aquaculture, and while aquaculture is
nomadism if crop production becomes unviable. often seen as an important development option, the
The supply of fishery products caught from the transition from being a fisherman to working in more
world’s coasts and oceans has historically been a key farming-like aquaculture can be a major barrier.
source of high-quality food, and including the fish The development and growth of aquaculture has
used for animal (and fish) feed, plus seaweeds; it is been one of the most remarkable features of the
still the most significant component of the global modern food sector, with production rising steadily
aquatic food industry. It is also a highly complex and in most parts of the world, increasingly supplanting
potentially vulnerable food system, consisting of a capture fisheries as the most important source of fish
mix of industrial operations with significant political and other aquatic food. As Bostock et al. (2010) dis-
influence and small-scale or artisanal fishing, which cuss, there is considerable potential for expansion,
provides an important source of direct food security with the major limiting factors being access to land
and an income safety net for poor people. The major and water as well as adequate market prices to provide
growth in marine fishery capacity over the last a viable return on investment in installation and oper-
50 years has resulted in almost all of the world’s ating costs. Fish-based meals and oils from industrial
stocks being harvested to full capacity or over- fisheries have been used extensively as feed in aquacul-
exploited, with troubling implications for ecosystem ture which raises issues of security of supply and
health, stock resilience and long-term output and value. environmental impact. However, dependence on this
Garcia & Rosenberg (2010) explore the future of type of feed is likely to decrease with a wider range
marine fisheries, arguing that there is little scope for of species being cultivated, especially those from
increased production, though a real risk of further lower trophic levels (non-carnivores). Not unexpect-
declines in catches exists if overfishing continues. edly, a large part of recent investment and technical
Over the last 40 years, the capacity of the global fleet development has focused on higher value species,
to catch fish has increased sixfold and as actual though the real costs of former luxury species such
harvests have remained nearly static, harvest pro- as salmon has declined considerably. Looking ahead,
ductivity has thus declined by six. Critical for the we can expect to see a marked increase in aquaculture
future health of the sector is better governance, and product development involving lower-cost species,
both for high-seas fisheries and those in Exclusive especially in low- and mid-income countries. Research
Phil. Trans. R. Soc. B (2010)
 Introduction. The future of the global food system H. C. J. Godfray et al. 2773

challenges include increasing resource-use efficiency though there are huge regional variations—water
(requiring both biological and engineering inno- dynamics are much more local than CO2 dynamics.
vation), reducing the risk of disease, improving Maintaining ERFs in watersheds such as the
environmental performance and maintaining the Colorado, Nile and Murray-Darling will be particularly
nutritional quality of farmed fish. difficult without affecting agriculture. At least until the
An often overlooked element of the food supply is second half of this century, water supply for agriculture
‘wild foods’—the plant and animal material that will be much more severely affected by competition
people harvest from non-agricultural ecosystems. from other sectors than by climate change. Clearly,
Wild food is collected in developed countries but it increasing water-use efficiency both in food production
is chiefly in low- and middle-income countries that it and in other areas is an important priority.
can be critical for a healthy diet. Bharucha & Pretty For many centuries humans could respond to
(2010) comprehensively review a large and scattered increased demands for food by bringing more land
literature on wild food and find that a remarkably into agriculture but as Smith et al. (2010) show, this
large number of species are used by different commu- is less of an option today. As populations grow and
nities. For example, a meta-analysis of over 20 African urbanization increases, more land is required for
studies finds that a typical rural community uses cities, and there is growing demand for non-food
approximately 100 wild species of plant. In addition, products such as fibre and wood. Any further
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

most rural communities intervene to manage popu- encroachment of agriculture into natural habitats is a
lations of wild species, suggesting that there is not major threat to biodiversity, and limiting deforestation
such a sharp dichotomy between agricultural and or other land-use change that leads to greater emis-
non-agricultural systems. sions is critical to reduce the rise in concentration in
Wild food is an ecosystem service provided by non- greenhouse gases. The threat of climate change has
agricultural environments. As is discussed further in led to a major increase in the fraction of land used
the review by Power (see below), it is important that for biofuels, despite serious doubts about whether
these non-monetarized benefits of natural and semi- first-generation biofuel production actually results in a
natural environments are understood in order to net reduction of greenhouse gas emissions when the
make rational decisions about land use at a time of consequences of this activity are fully accounted for.
increasing pressures on biodiversity. Relatively little new land has been brought into agri-
culture over the last 50 years (though even this
relatively modest conversion has often had major bio-
4. CLIMATE CHANGE AND RESOURCE diversity impacts or affected the livelihoods of poor
COMPETITION and indigenous groups), and the majority of the
Agriculture is globally one of the greatest consumers of gains in production over this period have been due
water and, as Strzepek & Boehlert (2010) discuss, to improved yields. There is little scope for agricultural
shortages of water may have a major effect on food land expansion in Asia and most of Europe, but there
production. Agriculture competes with domestic and may be considerable room for expansion of agricul-
commercial (municipal) consumers, and also with tural land use in South America and some room in
large-scale industrial users. The demand from both sub-Saharan Africa. However, there would be signifi-
these sectors is currently increasing, roughly in line cant environmental and cultural costs to future
with growing global levels of prosperity (though land conversion, especially if it involves the further
water demand eventually plateaus in high-income destruction of rainforest.
countries). Water is also required to maintain Natural ecosystems provide a variety of services to
functioning ecosystems and environmental flow mankind that are seldom included in traditional econ-
requirements (ERFs) are not traditionally included in omic calculations (the benefits of marine and
water calculations. Sufficient environmental flow is freshwater ecosystems through capture fisheries being
critical for freshwater ecosystems (and inland capture exceptions). Power (2010) shows how an explicit con-
fisheries), but also for some terrestrial ecosystems: sideration of these ecosystem services can help better
‘blue’ water in lakes and rivers is hydrologically manage the whole environment. Natural ecosystems
linked to ‘green’ (also called ‘brown’) water in soils. often assist in reducing erosion and improving the
Future water supply will be strongly influenced by cli- nutrient content of soils, as well as regulating hydrolo-
mate change, not least because evapotranspiration gical flows—all of value to farmers. They may also
occurs at a faster rate in a warmer climate. Though cli- maintain healthy populations of pollinators and the
mate change in some areas will be associated with natural enemies of weeds and pests, and hence
higher precipitation, if this occurs as extreme events increase yields in adjacent farmland. Agricultural
(storms and blizzards) it is of less use because of flood- land is also an ecosystem, albeit one strongly influ-
ing and run-off. The world has also been using up enced by man. It too can produce positive ecosystem
groundwater reserves at a rate far above replenish- services, if managed appropriately, for example, flood
ment, a particular concern in shallow aquifers control, wildlife habitats and carbon sequestration.
connected to surface hydrology. Of course, agricultural production can also provide
Strzepek & Boehlert use a watershed level, global negative ecosystem services or, as Power refers to
model to explore the water available for agriculture them, ecosystem disservices, for example, through
under different climate change scenarios. They con- pollution and biodiversity loss.
clude that meeting EFRs is perhaps the single The ecosystem services approach is important in
greatest challenge to agricultural water supply, providing a broader perspective on the value of natural
Phil. Trans. R. Soc. B (2010)
 2774 H. C. J. Godfray et al. Introduction. The future of the global food system

capital and understanding the full societal costs and Predicting energy prices is hugely difficult but scen-
benefits of different forms of food production. It is arios under consideration range from $50 a barrel to
particularly helpful in efforts to try to incorporate $130 or higher; as Woods et al. (2010) explore,
within the food system the positive and negative prices at the high end would have major effects on
externalities of food production, as well as identifying the type of crops grown, their yields and the pressure
‘win – win’ strategies, for example, agricultural prac- to convert new land to agriculture. Incentives to
tices that boost yield and increase sustainability. improve energy efficiency are likely to include much
However, the economic theory underlying ecosystem more on-farm biomass production and energy
services is still in its infancy and needs further develop- generation.
ment to increase its usefulness as a quantitative tool in It has been estimated that agriculture is responsible
decision-making. for about 20 per cent of all greenhouse gas emissions
The world is inevitably committed to some increase and is responsible for a major fraction of the anthropo-
in average global temperatures between now and 2050. genic production of methane and nitrous oxide (Stern
Gornall et al. (2010) explain that most climate models 2007). Further emissions occur indirectly though land
predict rises in temperature that will have mixed effects clearance for food production. Woods and co-workers
on agricultural production, possibly positive in explore opportunities to reduce these emissions, for
medium latitudes but negative in the tropics, especially example, by changed tillage practices, more efficient
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

in areas where agriculture is already at the margin. fertilizer use and production, and by agro-forestry
Changes in rainfall patterns are harder to predict, schemes.
and different regions will experience both higher and
lower precipitation. But as Jaggard et al. (2010) also
note, even where higher rainfall may benefit agricul- 5. CROSS-CUTTING THEMES
ture, if it occurs in high-intensity events much may Over the last 50 years, the world has seen sustained
be lost to run-off. Extreme events in general are growth in trade, including that in food products. The
likely to be an increasing problem for food production, globalization of the food system has occurred owing
with droughts, high temperature extremes, floods and to cheaper transport and communications, but also
(more controversially) tropical storms all likely to because of reductions in trade barriers and agricultural
increase in frequency. Climate change will also affect tariffs. Anderson (2010) explores these trends, and
food production through rises in sea level that risk shows how developed countries continue to subsidize
inundating coastal agriculture, reductions in glacier their agricultural sectors, though at rates that have
cover that might drastically change the hydrology of declined significantly from a peak in the mid-1980s.
rivers critical for irrigating large agricultural areas, Developing countries have historically exploited
and possibly through increases in pest and disease inci- rather than subsidized their agricultural sector,
dence, though the latter is very hard to predict though taxes and other burdens on agriculture are
accurately. declining and recent estimates suggest a net subsidy
However, the effects of climate change are not all in the developing world, albeit at rates still far below
negative: increased CO2 levels can increase yields, those in rich countries. A further major component
especially in C3 plants. Indeed, when the Intergovern- of globalization is liberalization in the rules governing
mental Panel on Climate Change (IPCC) estimated foreign direct investment, a trend that is promoting
the effects of climate change on yields, it predicted the consolidation of the private sector (retail, proces-
modest global increases (IPCC 2007). There are, how- sing and agri-business) into many fewer global players.
ever, important caveats. First, the increase critically The spate of trade restrictions prompted by the
depends on CO2 fertilization, and several recent 2008 food-price spike and the failure to reach a multi-
studies have suggested that this effect may have been lateral trade deal that year shows that increasing trade
overestimated (Ainsworth et al. 2008). Indeed, exactly liberalization is not a foregone conclusion, and indeed
how crops respond to the different components of cli- one possible future scenario is for developing
mate change is a major area of uncertainty. Second, countries, as they become richer, to impose their
the global average hides winners and losers—many own subsidies and trade restrictions. Anderson reviews
areas, particularly in the tropics, are likely to suffer economic models that suggest that this will harm the
radical yield reductions. Finally, even when regions poorest countries and increase volatility. Other factors
benefit from, for example, longer growing seasons, which may influence global food markets are the evol-
these benefits can only be realized if the necessary ution of the structure of the private sector, and the
well-adapted crop varieties, livestock breeds and uncertainties associated with competition for energy
human knowledge and expertise are available. (especially through oil prices and biofuels demand)
Agriculture affects emissions of greenhouse gases and water, and the effects of climate change.
both because it requires inputs and energy derived Food price volatility has long been a feature of
from burning fossil fuels, and because it is a major agricultural markets and Gilbert & Morgan (2010)
emitter of greenhouse gases. Overall, agriculture uses argue that recent fluctuations are not out of
a relative modest 3 per cent of global energy consump- keeping with historical patterns. Volatility in most
tion, a major fraction of which is used in the commodities has actually been lower in the past two
production of nitrogen fertilizer. Energy use by agri- decades than in the 1970s and 1980s, though rice
culture will in the future be affected both by changes remains a significant exception, perhaps because
in energy prices as well as by the introduction of incen- such low volumes are traded on global markets.
tive mechanisms to reduce on-farm energy use. Price fluctuations most often result from shocks to
Phil. Trans. R. Soc. B (2010)
 Introduction. The future of the global food system H. C. J. Godfray et al. 2775

production or consumption, with extreme weather of scenario analysis, combining different assumptions
events and major civil disturbances being the most about exogenous driving factors such as population,
common drivers—though there are interconnections global economic growth and climate change with an
between financial and food commodity markets that endogenous economic modelling engine. The econ-
are still not well understood. The impact of volatility omic models are typically partial equilibrium models,
on countries varies depending upon whether they are where the prices of different types of food in different
net food exporters or importers. For an individual countries connected by trade are determined by
household, the greater the proportion of income spent specifying such things as how demand is driven by
on food, the greater the adverse impact of food-price personal income, supply driven by likely changes in
spikes, as was illustrated by the food riots in a climate, etc. Assumptions are also made about
number of low-income countries in 2008. Price spikes endogenous productivity growth.
can rapidly become a major issue in domestic politics. The studies reviewed are generally relatively opti-
Given the difficulty of predicting future average food mistic about the task of feeding a global population
prices, it is not surprising that forecasting volatility—the of 9 billion, though most predict increased food
fluctuations around the average food price—is even prices and require trade stability to match supply and
harder. The authors find no evidence that volatility is demand in different geographical regions. Weaknesses
higher now than in the past. Looking ahead, the include rather rudimentary model validation and
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

increased frequencies of extreme events with climate deficiencies in the accuracy of the data used to initialize
change will likely increase volatility while a more globa- the models. More work needs to be put into coupling
lized food system can on the one hand buffer local economic, biophysical and climate models. Economic
perturbations but on the other propagate shocks more models by their nature consider commodities
systemically. Volatility will also be influenced by traded on global markets and new approaches are
changes to the governance of the global food system, required to explore more accurately the consequences
for example, the possible introduction of real or virtual of different futures for hunger and environmental sus-
grain reserves that a number of groups have called for in tainability. Nevertheless, despite these shortcomings,
the wake of the 2008 price spike (noting that others such models remain our only tool to try to explore
have argued that such interventions could be at best the synoptic behaviour of the food system.
ineffective and at worst counter-productive in disincen- An important way to increase food supply and to
tivizing any supply response). decrease the environmental consequences of current
Piesse & Thirtle (2010) explore the relationship food production is to reduce waste. It is often stated
between investment in agriculture (research and that between 30 and 40 per cent of food is wasted
development, technology and extension) and growth throughout the food system, though as Parfitt et al.
in yields and other measures of productivity. Since the (2010) point out, the evidence base upon which
Second World War, food production has kept pace these estimates are calculated is weak, particularly
with demand, due largely to scientific and technological with regard to losses in the developing world. There
innovation, first in developed countries, but critically is also little data on the price elasticity of food waste:
then with transmission to developing countries (the that is, how much of the problem would go away if
Green Revolution). There is evidence that the rate of food became more expensive?
increase in yield growth is declining, though Piesse & Food is wasted at all stages of the food chain, from
Thirtle point out that if productivity is measured in production and harvest all the way through to post-
terms of returns on total inputs beyond capital and purchase by the consumer. In developing countries
labour (total factor productivity), this slow-down is the highest losses occur at the post-harvest stage, typi-
less apparent. However, with overall investment in cally owing to such factors as spillage and spoilage
agriculture declining, in part, a result of low prices, con- brought about by inadequate transport and storage
tinued innovation-led growth is far from certain. infrastructure. Lack of capital for investment and
Recent decades have seen both a decline in total poorly functioning socio-economic institutions
investment in research, as well as a switch from the hamper waste reduction, though Parfitt et al. suggest
public to private sectors. It has also seen the transition that waste in traditional or small-scale agriculture
of countries such as Brazil and China to become major may have been over-estimated. In industrialized
investors in agricultural research and investment. countries, substantial waste occurs in households
Private sector investment inevitably leads to more after purchase, though retail, distribution and proces-
restrictions on intellectual property and the authors sing are also responsible for significant amount of
are concerned that this may limit the transmission of waste; there is a marked lack of data on the magnitude
new technology to low-income countries, as well as of waste in catering and public food outlets. A
to less focus on the needs of the poorest countries. reduction in waste early in the food chain has
To plan ahead it is important to have a forecast of occurred in many countries as they move from low-
future global food needs under a variety of different to middle-income status, though these gains are
assumptions or scenarios that are as accurate as poss- sometimes mitigated by increasing waste by consu-
ible. This is difficult as the food system is complex, mers and the retail trade. Encouragingly, a variety of
with dynamics determined by a combination of targeted incentive schemes in high-income countries
physical, biological and socio-economic processes. have demonstrated significant potential for waste
Reilly & Willenbockel (2010) review a number of reduction, given the political will.
exercises, which have sought to explore alternative Food security and health are closely intertwined, as
futures. Most have adopted one of a variety of types discussed by Hawkesworth et al. (2010). They begin
Phil. Trans. R. Soc. B (2010)
 2776 H. C. J. Godfray et al. Introduction. The future of the global food system

by describing the basic nutritional components of a we make and the science we choose to carry out over
healthy diet and how deviations from this can lead to the next few decades will determine whether all the
malnutrition and hunger, or obesity and its associated people living in this world have access to adequate food.
health problems. As also discussed by Kearney, the
health services of many middle-income countries are
having to cope with the double burden of under- and ENDNOTE
over-feeding. Hawkesworth et al. highlight the many 1
The views and opinions of the authors in this volume and this over-
gaps in our knowledge about people’s diets, especially view article do not represent the views of the Government Office for
in developing countries, which hamper better policy Science or the UK Government.
development. They argue that reliance on indirect
proxies such as those based on the balance of a coun-
try’s known or estimated production, imports and REFERENCES
exports, can be seriously misleading and they cham- Ainsworth, E. A., Leakey, A. D. B., Ort, D. R. & Long, S. P.
pion the importance of household and individual 2008 FACE-ing the facts: inconsistencies and interdepen-
surveys based on measuring consumption rather than dence among field, chamber and modeling studies of
availability. There have been significant improvements elevated [CO2] impacts on crop yield and food supply.
New Phytol. 179, 5–9. (doi:10.1111/j.1469-8137.2008.
in collecting data on poverty to address the millennium 02500.x)
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

development goals, which may provide a model for Anderson, K. 2010 Globalization’s effects on world
getting better data on diets. agricultural trade, 1960 –2050. Phil. Trans. R. Soc. B
365, 3007– 3021. (doi:10.1098/rstb.2010.0131)
Bharucha, Z. & Pretty, J. 2010 The roles and values of wild
6. CONCLUSIONS foods in agricultural systems. Phil. Trans. R. Soc. B 365,
The reviews published here describe the size of the 2913– 2926. (doi:10.1098/rstb.2010.0123)
challenge facing humanity in maintaining per capita Bostock, J. et al. 2010 Aquaculture: global status and trends.
food production and in trying to reduce the number Phil. Trans. R. Soc. B 365, 2897–2912. (doi:10.1098/rstb.
of people who suffer hunger and malnutrition. They 2010.0170)
Cirera, X. & Masset, E. 2010 Income distribution trends and
provide a number of grounds for optimism: the high
future food demand. Phil. Trans. R. Soc. B 365, 2821–
likelihood of population and consumption demand 2834. (doi:10.1098/rstb.2010.0164)
reaching a plateau some time during this century, FAO. 2006 Livestock’s long shadow. Rome, Italy: FAO.
and the major opportunities for yield growth through Garcia, S. M. & Rosenberg, A. A. 2010 Food security and
the application of traditional and novel science. They marine capture fisheries: characteristics, trends, drivers,
also stress the significant yield gains that can be and future perspectives. Phil. Trans. R. Soc. B 365,
obtained by working closely with farmers and fisher- 2869– 2880. (doi:10.1098/rstb.2010.0171)
men to deploy existing knowledge and resources Gilbert, C. L. & Morgan, C. W. 2010 Food price volatility.
through better education and greater social and econ- Phil. Trans. R. Soc. B 365, 3023–3034. (doi:10.1098/rstb.
omic equality—closing the gap between realized and 2010.0139)
Godfray, H. C. J. et al. 2010 Food security: the challenge of
current possible levels of productivity and improving
feeding 9 billion people. Science 327, 812– 818. (doi:10.
efforts to reduce waste. Another theme that emerges 1126/science.1185383)
is the importance of taking a ‘competing risks’ Gornall, J., Betts, R., Burke, E., Clark, R., Camp, J., Willett,
approach to regulation in the food system—it is too K. & Wiltshire, A. 2010 Implications of climate change
easy to close off options by applying naive versions of for agricultural productivity in the early twenty-first
the precautionary principle. The world is going to century. Phil. Trans. R. Soc. B 365, 2973–2989. (doi:10.
have to produce more food, and unless much of the 1098/rstb.2010.0158)
Earth’s remaining biodiversity is to be destroyed, this Hawkesworth, S., Dangour, A. D., Johnston, D., Lock, K.,
will need to be done without expanding the area Poole, N., Rushton, J., Uauy, R. & Waage, J. 2010 Feed-
under cultivation. Achieving higher yields from the ing the world healthily: the challenge of measuring the
same acreage without severely impacting the environ- effects of agriculture on health. Phil. Trans. R. Soc. B
365, 3083– 3097. (doi:10.1098/rstb.2010.0122)
ment requires a new way of approaching food
IPCC. 2007 Fourth assessment report of the Intergovernmental
production—sustainable intensification (Royal Society Panel on Climate Change (Contribution of Working
2009; Godfray et al. 2010). The reviews also highlight Group II). Cambridge, UK: Cambridge University Press.
the importance of treating food production in a Jaggard, K. W., Qi, A. & Ober, E. S. 2010 Possible changes
broader context, as one of several major competitors to arable crop yields by 2050. Phil. Trans. R. Soc. B 365,
for freshwater, land and energy, and as integral to 2835– 2851. (doi:10.1098/rstb.2010.0153)
the world’s overarching challenge of mitigating and Kearney, J. 2010 Food consumption trends and drivers. Phil.
adapting to climate change. Finally, they emphasize Trans. R. Soc. B 365, 2793–2807. (doi:10.1098/rstb.
the numerous research questions that require investi- 2010.0149)
gation, not just in the natural sciences but in Kennedy, G., Nantel, G. & Shetty, P. 2004 Globalisation of
food systems in developing countries: a synthesis of
economics, politics and the social sciences.
country case studies. In FAO Food and nutrition paper
In the eighteenth century, Thomas Malthus alerted 83. Globalisation of food systems in developing countries:
the world to the consequences of rapid (geometric) impact on food security and nutrition. Rome, Italy: FAO.
population growth for what we (though not he) would Lutz, W. & Samir, K. C. 2010 Dimensions of global popu-
call food security. Thanks to the demographic tran- lation projections: what do we know about future
sition, we can now envisage a world where population population trends and structures? Phil. Trans. R. Soc. B
and consumption cease to rise. The policy decisions 365, 2779– 2791. (doi:10.1098/rstb.2010.0133)

Phil. Trans. R. Soc. B (2010)

 Introduction. The future of the global food system H. C. J. Godfray et al. 2777

Lutz, W., Scherbov, S., Makinwa-Adebusoye, P. K. & Satterthwaite, D., McGranahan, G. & Tacoli, C. 2010
Reniers, G. 2004 Population-environment-development- Urbanization and its implications for food and farming.
agriculture interactions in Africa: a case study Phil. Trans. R. Soc. B 365, 2809– 2820. (doi:10.1098/
on Ethiopia. In Frontiers of population forecasting (eds rstb.2010.0136)
W. Lutz, J. W. Vaupel & D. A. Ahlburg). New York, Smith, P., Gregory, P. J., van Vuuren, D., Obersteiner, M.,
NY: The Population Council. Havlı́k, P., Rounsevell, M., Woods, J., Stehfest, E. &
Parfitt, J., Barthel, M. & Macnaughton, S. 2010 Food waste Bellarby, J. 2010 Competition for land. Phil. Trans.
within food supply chains: quantification and potential R. Soc. B 365, 2941–2957. (doi:10.1098/rstb.2010.0127)
for change to 2050. Phil. Trans. R. Soc. B 365, 3065– Stern, N. 2007 The economics of climate change. Cambridge,
3081. (doi:10.1098/rstb.2010.0126) UK: Cambridge University Press.
Piesse, J. & Thirtle, C. 2010 Agricultural R & D, technology Strzepek, K. & Boehlert, B. 2010 Competition for water for
and productivity. Phil. Trans. R. Soc. B 365, 3035–3047. the food system. Phil. Trans. R. Soc. B 365, 2927–2940.
(doi:10.1098/rstb.2010.0140) (doi:10.1098/rstb.2010.0152)
Popkin, B. M. 1998 The nutrition transition and its health Thornton, P. K. 2010 Livestock production: recent trends,
implications in lower-income countries. Publ. Health future prospects. Phil. Trans. R. Soc. B 365, 2853–
Nutr. 1, 5 –21. 2867. (doi:10.1098/rstb.2010.0134)
Power, A. G. 2010 Ecosystem services and agriculture: United Nations. 2004 World population. New York, NY:
tradeoffs and synergies. Phil. Trans. R. Soc. B 365, United Nations.
2959– 2971. (doi:10.1098/rstb.2010.0143) Welcomme, R. L., Cowx, I. G., Coates, D., Béné, C., Funge-
Downloaded from https://royalsocietypublishing.org/ on 05 March 2025

Reilly, M. & Willenbockel, D. 2010 Managing uncertainty: a Smith, S., Halls, A. & Lorenzen, K. 2010 Inland capture
review of food system scenario analysis and modelling. fisheries. Phil. Trans. R. Soc. B 365, 2881–2896. (doi:10.
Phil. Trans. R. Soc. B 365, 3049–3063. (doi:10.1098/ 1098/rstb.2010.0168)
rstb.2010.0141) Woods, J., Williams, A., Hughes, J. K., Black, M. & Murphy,
Royal Society 2009 Reaping the benefits: science and the sustainable R. 2010 Energy and the food system. Phil. Trans. R. Soc. B
intensification of global agriculture. London, UK: Royal Society. 365, 2991–3006. (doi:10.1098/rstb.2010.0172)

Phil. Trans. R. Soc. B (2010)