Martina Sartori

Hosting a Mega-Event: Is
it Good or Bad for the
Economy?
General Equilibrium
Models as a Litmus Paper
Test

ISSN: 1827-3580
No. 21/WP/2017

Electronic copy available at: https://ssrn.com/abstract=3046142

 Working Papers
Department of Economics
Ca’ Foscari University of Venice
No. 21/WP/2017
ISSN 1827-3580

Hosting a Mega-Event: Is it Good or Bad for the Economy?

General Equilibrium Models as a Litmus Paper Test

Martina Sartori
Ca’ Foscari University of Venice

Abstract
Hosting mega-events has long been regarded as an opportunity for economic growth, creating
long-lasting benefits and attaining international recognition. Recently, both the scientific
community and the public opinion at large have turned much more skeptical about the impact
of mega-events. Why is it that some events appear successful and other disasters? Why the
perception of impacts has changed over time? To answer these (and other) questions, there is a
need to go beyond the simple narratives and “stylized facts”, to undertake some serious
scientific investigation, based on verifiable data and testable models. Despite the fact that most
recent studies use the same modeling tool, namely some Computable General Equilibrium
(CGE) model, results are sometimes contradictory, thereby hindering the trustiness of the
economic analysis for policy guidance.
In this paper, we show that results are different because assumptions are different and because
the range of effects considered is different, even when the same model is employed.
Furthermore, some critical hypotheses are not often clearly stated. We advocate some kind of
standardization in the process of model building for the economic assessment of mega-events.
Only a transparent and replicable model exercise can serve as a “litmus paper”, to ascertain
whether hosting a mega-event is good or bad for an economy.

Keywords
Computable general equilibrium modeling, methodological issues, mega-events
economic impact assessment
JEL Codes
C68, Z28, Z38
Address for correspondence:
Martina Sartori
Department of Economics
Ca’ Foscari University of Venice
Cannaregio 873, Fondamenta S.Giobbe
30121 Venezia - Italy
e-mail: [email protected]

This Working Paper is published under the auspices of the Department of Economics of the Ca’ Foscari University of Venice. Opinions
expressed herein are those of the authors and not those of the Department. The Working Paper series is designed to divulge preliminary or
incomplete work, circulated to favour discussion and comments. Citation of this paper should consider its provisional character.

The Working Paper Series Department of Economics

is available only on line Ca’ Foscari University of Venice
(http://www.unive.it/pag/16882/) Cannaregio 873, Fondamenta San Giobbe
For editorial correspondence, please contact: 30121 Venice Italy
[email protected] Fax: ++39 041 2349210

Electronic copy available at: https://ssrn.com/abstract=3046142

 Hosting a Mega-Event: Is it Good or Bad for the Economy?
General Equilibrium Models as a Litmus Paper Test

Martina Sartori*

Abstract
Hosting mega-events has long been regarded as an opportunity for economic growth, creating
long-lasting benefits and attaining international recognition. Recently, both the scientific
community and the public opinion at large have turned much more skeptical about the impact
of mega-events. Why is it that some events appear successful and other disasters? Why the
perception of impacts has changed over time? To answer these (and other) questions, there is a
need to go beyond the simple narratives and “stylized facts”, to undertake some serious
scientific investigation, based on verifiable data and testable models. Despite the fact that most
recent studies use the same modeling tool, namely some Computable General Equilibrium
(CGE) model, results are sometimes contradictory, thereby hindering the trustiness of the
economic analysis for policy guidance.
In this paper, we show that results are different because assumptions are different and because
the range of effects considered is different, even when the same model is employed.
Furthermore, some critical hypotheses are not often clearly stated. We advocate some kind of
standardization in the process of model building for the economic assessment of mega-events.
Only a transparent and replicable model exercise can serve as a “litmus paper”, to ascertain
whether hosting a mega-event is good or bad for an economy.

Keywords: computable general equilibrium modeling, methodological issues, mega-events

economic impact assessment.
JEL codes: C68, Z28, Z38.

Acknowledgements
Financial support received by the Department of Economics, Ca’ Foscari University of Venice, is gratefully acknowledged. The
author thanks Prof. Jérôme Massiani for his assistance in reviewing the literature and insightful comments on earlier drafts of the
paper; and Prof. Roberto Roson for his constructive suggestions and help in coding the model. Prof. Larry Dwyer and the
participants to the 20th Annual Conference on Global Economic Analysis (Purdue University, West Lafayette, Indiana, USA) are also
acknowledged for the fruitful discussion.

* Ca’ Foscari University of Venice. E-mail: [email protected]

Electronic copy available at: https://ssrn.com/abstract=3046142

1. Introduction

“Six months after hosting the 2016 Summer Olympic Games, the Rio de Janeiro sport venues are already idle and in disrepair,
raising questions about a legacy that organisers promised would benefit the Brazilian city and its residents.”
The Guardian, 2017

Would you like to get your country hosting the next Olympic Games? Should this question be
asked some years ago, the likely answer would be: “Yes”. Indeed, hosting “mega-events”1, such
as the Olympics or the Expo, has long been regarded as an opportunity for economic growth,
creating long-lasting benefits and attaining international recognition. For instance, the 1964
Tokyo Olympic Games became a hallmark of the economic recovery of Japan after World War II.
Other positive examples may be Los Angeles 1984 and Barcelona 1992 (Preuss, 2004; Brunet,
1995; Zimbalist, 2015; The Guardian, 2002).
Recently, however, both the scientific community and the public opinion at large have turned
much more skeptical about the impact of mega-events (e.g., Baade and Matheson, 2002, 2016;
Horne, 2004; Pillary and Bass, 2008; Barclay, 2009; Zimbalist, 2015; The Guardian, 2012, 2017). It
is now recognized that a mega-event may subtract financial resources to alternative projects,
possibly providing higher social returns. Some studies find that a mega-event is rarely
profitable (Coates and Humphreys, 2008; Baade and Matheson, 2016), as economic benefits
hardly compensate for the substantial costs and risks incurred (Barclay, 2009).
According to Baade and Matheson (2004), following the 1994 soccer World Cup in the United
States host cities experienced a net economic loss rather than the predicted gain. Giesecke and
Madden (2011) show that Sydney 2000 generated a real consumption loss of $2.1 billion. The
Athens 2004 Summer Olympic Games are regarded as one key factor behind the Greek debt
crisis (The Guardian2, 2012).
Host cities are often left with specialized infrastructure that has little use beyond the mega-
event, so that they may face heavy long-term expenses for the maintenance of the so-called
“white elephants” (Baade and Matheson, 2016). For example, many of the venues of Athens
2004 have fallen into disrepair and Brazilian’s Olympic infrastructure is experiencing the same.
(The Guardian3, 2017).
So, are mega-events good or bad for the hosting economy? Why is it that some events appear
successful and other disasters? Why the perception of impacts has changed over time? To
answer these (and other) questions, there is a need to go beyond the simple narratives and
“stylized facts”, to undertake some serious scientific investigation, based on verifiable data and
testable models.
Yet, and despite the fact that most recent studies use the same modeling tool, namely some

1 Mega-events are defined as “ambulatory occasions of a fixed duration that attract a large number of visitors, have a large mediated
reach, come with large costs and have large impacts on the built environment and the population” (Müller, 2015).
2 “No one knows how much the Olympics cost Greece – although many think they played a major role in producing the debt that spurred the

country's economic downfall.”

3 “The iconic stadium has fallen into a state of abandonment and has been closed to tourists due to a dispute between the stadium operators, the

Rio state government, and Olympic organisers over $1m in unpaid electricity bills and management of the venue.”

2
Computable General Equilibrium (CGE) model, results are sometimes contradictory. One may
say that these divergences may be due to some fundamental differences in the economic
environment, as well as in the nature of events and shocks. This is only partially true as, for
instance, studies carried out for the same mega-event exhibit opposite findings, thereby
hindering the trustiness of the economic analysis for policy guidance.
The simple point we want to make here is this: results are different because assumptions are
different and because the range of effects considered is different, even when the same model is
employed. Furthermore, some critical hypotheses are not often clearly stated. Indeed, there is
no single way to make numerical simulations with a CGE model, which may differ in terms of
shocks and/or closure rules.
The success of a recipe depends on the ingredients, which must be listed outside the package of
any food product. Analogously, we advocate some kind of standardization in the process of
model building for the economic assessment of mega-events. Only a transparent and replicable
model exercise can serve as a “litmus paper”, to ascertain whether hosting a mega-event is good
or bad for an economy (and why, how much, etc.).
To this end, after reviewing the literature (Section 2), we discuss in Section 3 of this paper the
key effects at play and (Section 4) how they can be simulated in a CGE setting. In Section 5, we
present an illustrative model and simulation exercises, based on the proposed reference
framework. A final section draws some conclusions.

2. CGE modeling of mega-events

The traditional approach for the impact assessment of large-scale investments is the Input-
Output analysis (e.g., Lee and Taylor, 2005). Recently, I-O models have been progressively
replaced by Computable General Equilibrium (CGE) models, in all those contexts (including the
assessment of mega-events) where changes in relative prices and in the structure of the
economic system matter. Indeed, after the pioneering contribution by Madden and Crowe
(1998), the number of studies using CGE models for the economic analysis of mega-events has
been steadily increasing.
More generally, CGE models have become a standard tool of empirical economic analysis and
policy assessment. They are extensively employed in many fields, most notably to assess
international trade, fiscal and environmental policies, but their adoption for the assessment of
mega-events is relatively recent.
The discussion on the drawbacks of I-O analysis dates back to Crompton (1995), who identified
few typical problems in I-O studies applied to sports events (e.g., the omission of substitution
effects). Since then, there has been a growing literature recognizing that the assumptions
imposed by the I-O approach generate a systematic over-estimation of economic impacts (see,
e.g., Dwyer et al. 2004, 2006; Jago and Dwyer, 2006). As Madden (2006) puts it: “the problems
with I-O analysis are of such a fundamental nature that it should no longer be used in the
analysis of sporting ventures, particular not for mega-events of the scale of the Olympic Games”.
With an assumed unlimited supply of factors, an implicit assumption of I-O models is infinitely
elastic supply curves. As such, I-O models are the multi-sectoral equivalent of Keynesian
macroeconomic models: the economy always finds the necessary resources to meet any increase

3
in final demand, including the one triggered by investments for the realization of a mega-event.
There is no crowding out, and any increase in demand will yield only positive indirect effects.
By ignoring this important aspect, a systematic over-estimation of impacts is clinched, because
the potential negative economic impact induced by a positive demand shock cannot be
measured by construction (Dwyer et al., 2004; Matheson, 2009). Furthermore, for temporary
demand shocks, such as those associated with mega-events, the release of local resources may
be important determinants of the overall economic outcome.
In this setting, CGE analysis has typically been favored (Dwyer et al., 2004, 2005).4 Dwyer et al.
(2005) estimate the economic impact of a major sport event (a car race) at the national level,
using both a standard I-O analysis and a CGE model. They find that net aggregate benefits
computed with the standard I-O model are about two times larger than those estimated with
the general equilibrium model.
CGE models build upon the Walrasian theory of general equilibrium and on the key concepts of
market clearing and optimizing behavior of rational economic agents. A CGE model traces the
circular flow of income and represent the economy as an integrated system, in which markets
influence each other, resource are constrained, prices are allowed to vary and demand and
supply must balance.5 From a mathematical viewpoint, a CGE is a large non-linear system, to be
solved with appropriate mathematical algorithms, and with parameters calibrated on real
world data. Counterfactual simulations are typically carried out by changing exogenous
parameters in a comparative static setting.
Since the contribution of Madden and Crowe (1998) on the economic impact of the Sydney 2000
Summer Olympics, the number of studies using CGE models has increased. Most studies focus
on international large sports events, like the Summer Olympic Games (Madden, 2002, 2006;
Giesecke and Madden, 2007, 2011; Blake, 2005; Li et al., 2011; Li et al., 2013), the FIFA Soccer
World Cup (Bohlmann and van Heerden, 2005, 2008; Saayman and Rossouw, 2008) and the
Glasgow 2014 Commonwealth Games (Allan et al., 2017). Few papers analyze other
international and national sport events, like car races (Dwyer et al., 2004, 2005).
Some of these studies are particularly relevant to the purpose of this paper, because they
quantify the sensitivity of results to different model specifications and offer a portfolio of
simulation cases. Some other studies only provide partial analyses, as some economic effects are
not fully taken into account (Li et al. 2011; Li et al., 2013; Allan et al., 2017). Contrary to I-O
models, most CGE models assume full employment of available resources, thereby ruling out
the employment of unused primary factors (Giesecke and Madden, 2007, 2011). As noted by
Madden (2006), the way in which labor market is modeled has an especially large influence on
the results of a CGE simulation.
The study by Blake (2005) on the London 2012 Olympics shows that the impact of the Games

4CGE analysis is also considered superior to other evaluation methods, like cost-benefit analysis (CBA). Dwyer and Forsyth (2009)
argue that CBA cannot estimate e.g. the inter-industry impacts of an event, impacts on exchange rates, the shadow price of labor
and capital. Andersson et al. (2008) and Li and Jago (2013) stress, instead, the critical assumptions needed behind the estimation of
both benefit- and cost-side of a CBA, especially regarding the quantification of non-economic elements. Finally, macro-econometric
modeling cannot capture the interrelationships of different industries in an economy (Russo, 2009).
5This does not imply that in CGE models primary resources (i.e., labor, capital) are always fully employed or that trade
deficit/surplus are not allowed. This issue will be further discussed in Section 4.

4
may vary considerably across different sectors of the UK economy. Sectors that expand include
construction, passenger land transport, business services, hotels and restaurants. Sectors that
contract include manufacturing, agriculture, fishing and other services. This differential sectoral
impact is an example of the so-called “the Dutch disease”: higher competitiveness in some
industries causes lower competitiveness elsewhere in the economy.6
A couple of works by Giesecke and Madden (2007, 2011) estimate the impact of the 2000 Sydney
Olympics. They employ a recursive-dynamic multiregional CGE model, covering the
Australia’s eight states and territories, and identify three routes of direct effect on the economies
of New South Wales (where Sydney is located) and Australia: (i) construction of the Games
facilities and associated financing, financed via direct taxation of New South Wales households;
(ii) operation of the Games; and (iii) tourism spending by interstate and foreign Games visitors.
The papers by Giesecke and Madden are based on previous studies carried out by Madden
(2002, 2006), who discusses in depth the effects of different modeling assumptions on
simulation results. The simulation of the Sydney Olympics economic effects is done under
alternative scenarios on increased exports demand (more or less optimistic) and labor market
conditions (more or less tight) characterizing the three phases of the Games (i.e., pre-event,
event-year and post-event). The author finds that the assumed degree of slackness in the short-
run labor market and the portion of foreign demand in the direct Olympics expenditure are
some major determinants of the size and sign of the overall economic impact.
While the studies on the Sydney Olympic Games agree that a mega-event, even of the size of
the summer Olympics, is unlikely to provide any substantial boost to either the national or host-
region economy, two studies on the economic impact of Beijing 2008 Olympics arrive at
qualitatively different results (Li et al., 2011; Li et al., 2013).
The extent to which differences in results are due to model assumptions is investigated in Li et
al. (2013) and Allan et al. (2017). Li et al. (2013) assume that firms compete in a Dixit–Stiglitz
imperfectly competitive setting, and results show that the welfare impacts under imperfect
competition are higher than when perfect competition is assumed. Allan et al. (2017) use a
multi-period Scottish CGE model to estimate the effects of the Glasgow 2014 Commonwealth
Games, finding that results are very sensitive to the degree to which economic agents are
forward-looking and on the timing of pre-announcement of the event.
Bohlmann and van Heerden (2005, 2008) and Saayman and Rossouw (2008) consider the 2010
Soccer World Cup in South Africa, and its various impacts (i.e., infrastructures development,
increased tourism and financing implications), recognizing that the simulation results of the
shocks applied to the economy depend to a large extent on the model closure.
To sum up, the key insight emerging from a review of the literature is that model results can be
substantially different, despite the use of a common modeling approach (CGE), because of the
nature of the exogenous shocks considered and of the (sometimes hidden) assumptions and

6The Dutch disease (Corden and Neary, 1982) refers to the apparent relationship between the increase in exploitation of natural
resources and a decline in the manufacturing sector. An increase in revenues from natural resources (or inflows of foreign aid) will
make a given nation’s currency stronger compared to that of other nations, resulting in the nation’s other exports becoming more
expensive for other countries to buy, making the manufacturing sector less competitive. While it most often refers to natural
resource discovery, it can also refer to any development that results in a large inflow of foreign currency, like an increase in tourism
receipts.

5
closure rules. In the rest of the paper, we shall try to discuss how, why and how much the
various “ingredients” of the “model recipe” may affect the simulations output.

3. Economic processes triggered by a mega-event

Mega-events are atypical tourism activities where the scale of investment in bid development,
complexity of decision making involved and the potential impacts are great and lasting (Li and
McCabe, 2013). As pointed out by Faulkner (1993), “holding large scale events implies a range
of economic, social, cultural, political and environmental impacts”. These features make a
mega-event somewhat different from other short run Keynesian demand-pushing expenditure,
like the construction of new transport infrastructures.
Consider first the multiplicity of effects generated by a mega-event and the time when they
occur. Two main categories of effects can be identified: expenditure effects and productivity
effects.
Examples of expenditure effects are: investments to build event-related venues and facilities
(e.g., sport buildings in the case of sport events, other temporary facilities) and renovation of
infrastructure (e.g., transportation, information technology); operational expenditures (e.g.,
ceremonies and programs, media and press center, security, administration and event
promotion); tourist inflows accompanied by increased expenditure. Funding of expenditure and
investment can be based on foreign, domestic private and domestic public sources. This
distinction is important when it comes to accurately estimate the income and welfare effects
generated by the mega-event, because only foreign expenditure/investment are new money
injected into the host economy, while the other two are just national transfers between
alternative uses (Li, 2012). In other words, without holding the event, event-related foreign
expenditure/investment would not be generated, whereas domestic private and public
expenditure/investment would still be used for other purposes.
The productivity effects include: increased productivity because of improved infrastructure,
higher competitiveness of domestic goods and services in international markets, rising export
demand (Rose and Spiegel, 2010).
Productivity effects are related the so-called “intangibles” (Atkinson et al., 2008). For instance,
Baade and Matheson (2016) stress that Olympic Games are perceived as a great international
sporting event that brings increased utility to the population of the host country, arising from
non tangible factors like “national pride”. The non-pecuniary “feel good” effect experienced by
local citizens in terms of enhanced civic pride and national identity and the general excitement
from the event experience is an instance of positive intangibles; degradation of natural fauna,
noise pollution and congestion are examples of negative intangibles. Both should be accounted
for in the estimation of the overall welfare impact (Li and McCabe, 2013). Giesecke and Madden
(2011) indeed find that the Sydney Olympic Games had a beneficial effect on the Australian
economy only if intangible benefits were considered to be large enough. However, while
tangible benefits/costs can be measured in terms of the amount of income earned or expenditure
incurred, intangibles are hard to quantify, as they do not possess an explicit monetary value.
Several studies have attempted to quantify the intangible benefits of the Olympic Games
through the use of contingent valuation methodology, which constructs a set of survey

6
questions designed to elicit the monetary value people place on whether certain events occur or
do not occur (see e.g., Atkinson et al., 2008; Walton et al., 2008).
Figure 1 summarizes the main economic effects generated by a mega-event.

EXPENDITURE - Investment in event-

EFFECTS related infrastructure,
Economic
venues and facilities
- Operating expenditure impacts of the
- International tourism mega-event
expenditure
- Income generation
- Job creation
- Relative price change
PRODUCTIVITY - improved infrastruct.
EFFECTS
- Employment change
- improved competitiv.
- Output change
- rising export demand
- Dutch disease
- Welfare change
- Government balance
INTANGIBLES - national-pride effects
change
- country image legacies
- Financial inflows change
- environmental and
- Change in terms of trade
social costs
-…

Figure 1. Economic effects attributable to a mega-event and their possible effects on the economy.

Regarding the temporal distribution of the various effects, Madden and Crowe (1998) identify
three distinct phases. During the pre-event phase some domestic resources are diverted towards
event-related activities (a demand shift), like the construction and upgrading of venues,
accommodation and transport infrastructures. This phase has a variable duration, depending on
the type of event under consideration. For example, in the case of the Summer Olympic Games,
it could start around five years before the year of the event, at the time when the country is
awarded with the hosting of the event and the construction activities begin.
The second phase corresponds to the period when the event occurs. This is the time of expenses
for operations and international tourism. The last phase is the post-event period, during which
the socio-economic legacy (if any) becomes apparent. For Li and McCabe (2013) legacy includes
“tangible and intangible elements of large-scale events left to future generations of a host
country where these elements influence the economic, physical and psychological well-being at
both community and individual levels in the long-term”.

4. Simulating the economic effects of a mega-event with a CGE model

CGE models are particularly suited to run “what-if” scenario simulations, to depict how an
economy would look like if a policy had been implemented or a shock had been occurred. In the
case of a large event assessment, these simulations are intended to gauge the effects in the
hosting country of a change in expenditure and/or productivity. Not all impacts can be easily
captured within a CGE framework, though. For example, it would be difficult to model traffic
congestion, noise or to incorporate other localized externalities associated with the event. When
it comes to evaluate some impacts on welfare, the level of aggregation of the model matters.
Depending on the nature of the mega-event taken into consideration, economic effects may be

7
very limited in time and quite localized, often restricted to a specific city or region.
Proper model design is necessary to get meaningful and informative simulation exercises. Some
key questions are: which changes in the model structure are needed to better describe the
economy under investigation? Which exogenous parameters, for which sectors, should be
perturbed and what is their economic interpretation? To what extent the size and the sign of the
results are assumption-driven and affected by the model set up? Are there available data to
realistically gauge the magnitude of the exogenous shocks? These are what we refer here as
“ingredients of the model recipe”.
A mega-event induces a change in the patterns of consumption and investment. This structural
modification can be captured in a CGE model through the adjustment of some demand shifting
parameters. Another expenditure effect is due to the increase in the number of tourists in the
hosting country, occurring during the event. Extra international tourism spending is often
regarded as the main advantage to the host economy (Dwyer et al., 2005). To simulate such an
impact, one should consider that more income, earned abroad, is spent in the hosting country;
the pattern of final consumption changes, with higher demand for hotels, restaurants,
transports and other services. Roson and Sartori (2014) suggest to model this effect by
increasing the net inflow of international income transfers, while simultaneously augmenting
the demand for goods and services typically consumed by tourists. Another possibility is
introducing a new sector (tourism exported sector, producing tourism products and services)
and possibly more representative consumers (international tourists, demanding for tourism
products and services) (Li, 2012).
Intangible effects related to the national identity and “brand” can also be quite relevant. To the
extent that these effects translate into higher perceived quality of domestic goods and services
abroad, a tangible impact is the increase in the export demand. Also, a better image may result
in a higher inflow of international tourists, lasting after the event is concluded.
It is essential that a CGE model is designed so as to capture the relevant characteristics of the
economy and to capture most of the effects coming into play when the realization of a mega
event is simulated. This is a somewhat more technical issue, to which the next section is devoted.

4.1. Modeling issues: closure rules and working assumptions

A closure in a CGE model refers to the partition between endogenous and exogenous variables
in the system. An example of a closure rule is the endogenous wage determination through
labor market clearing, which is alternative to fixing real or nominal wages. Other instances are
the determination of the public sector or trade balances. Not surprisingly, the debate on the
appropriate closing rules to be chosen in the various circumstances has been at the heart of the
CGE literature since its very beginning (see, e.g., Shoven and Whalley, 1984; Robinson, 1991).
Closing rules may influence the results of numerical simulations in a significant way. Horridge
(2000) discusses a number of criteria that should guide the choice of the model’s closure. The
time horizon of the simulation is a first aspect to consider (e.g., short- versus long-term). Some
economic variables, like the capital stock, need time to adjust to a new equilibrium, affecting the
way factor markets should be modeled. The time dimension is also important when modeling
the impacts of a mega-event, because the many effects do not occur simultaneously. During the

8
pre-event phase, the expansion of demand for investment and the increase in financial inflows
usually calls for a model closure with flexible government and trade balances, as well as with
some degree of slackness in the labor market. Instead, these assumptions may not be justified
when modeling the post-event phase, where, for example, a trade surplus may be necessary to
repay the interests burden on the cumulated foreign debt.
A second criterion has to do with the appropriate assumptions regarding the circular flow of
income and spending in the economy. For instance, many CGE models impose the equality of
savings and investments (as it would be the case in a closed economy). When the saving rate is
given, total savings and investments are endogenously and simultaneously determined. In a
different context (e.g., during the pre-event phase), it may be more appropriate to set
investment spending, possibly making the savings rate endogenous.
The same holds for the government balance and the balance of trade, often referred to as macro-
closure. The government balance closure defines whether government savings, that is the
difference between government revenues and expenditure, are endogenous or exogenous, and
whether government spending is fixed. If government spending is fixed exogenously and
revenues depend on a variety of endogenous tax instruments, government savings (possibly
negative) are determined endogenously. In the case of a revenue-neutral government closure,
spending and revenue are both fixed, while tax rates must be, of course, endogenous. In another
alternative closure government savings are a flexible residual while all tax rates are fixed, or
government savings and direct tax rates are both fixed and government consumption is the
adjusting variable. Clearly, this closure rule is a critical aspect to be carefully assessed when
simulating a mega-event. If the event is mainly financed through new taxes and the government
balance is retained fixed at its initial equilibrium level, then tax rates should be exogenously
fixed and government spending endogenously adjusts to maintain the fixed balance.
The assumptions on the balance of trade closure are also very important. This closure defines
whether financial net inflows (identically equivalent to the foreign trade deficit) are exogenous
and the exchange rate is endogenous, or vice versa. An exogenous current account closure fixes
the supply of foreign savings, that is the current account deficit or surplus, at its initial level and
lets the exchange rate adjust to maintain that level. In this case, an increase in real national
income will push up the real exchange rate, with possible negative consequences on the relative
competitiveness for some industries. Exchange rate movements may be important determinants
of the economic impacts of large events, especially in small open economies. The alternative
closure is a fixed exchange rate that makes the balance of trade account endogenous. This
closure appears to be appropriate when the initial phase of a mega-event is simulated, because
funding resources at this stage may come abroad, at least in part.7
Clearly, the choice about the closure rules depends on the purpose of the modeling exercise.
Bohlmann and van Heerden (2008), who study the 2010 Soccer World Cup by means of a
comparative static CGE model, adopt a modified version of the closure described in Horridge
(2000): capital stock, technological structure, tax rates and investment levels are hold fixed,
whereas household consumption, employment level and the trade balance are endogenously

7 In the CGE literature, a number of macro-closures are discussed. See, e.g., Lofgren et al. (2002) and Burfisher (2011).

9
obtained. This is due to the need to simulate the funding process through higher taxes, the
building of new stadiums and other infrastructure, and higher productivity. In Giesecke and
Madden (2011), baseline and counterfactual scenarios are simulated under different closures, to
uncover the contribution of a mega-event in the variation of some macroeconomic variables.
Many effects generated by a mega-event are a consequence of a change in the level and
structure of the final demand, including private and public consumption, as well as investment.
The extent to which this change in demand induces a different utilization of scarce primary
resources depends on how factors endowments/supply are handled in the CGE model. The
standard assumption is that primary resources (e.g., labor and capital) are supplied in fixed
amount. However, excluding the possibility of employing unused local resources may severely
underestimate some positive impact generated by a mega-event. In this respect, the hypotheses
underlying CGE and I-O models are totally opposite, and are at the basis of the large
discrepancies in the results obtained with these two classes of models.
This phenomenon is especially evident in the labor market. An expansion of demand may
activate unused labor resources (unemployment) and therefore generate real income growth,
which could be the case before and during the realization of a mega-event. To model this
process in a CGE setting, some labor supply functions could be introduced. For instance, a wage
curve, that is a relationship between real wage and unemployment rate, is the additional
equation a model needs to determine the unemployment rate endogenously (the difference
between total labor supply and equilibrium employment). Boeters and Savard (2011) illustrate
the many different ways in which the labor market can be modeled within a CGE model, and
Dwyer et al. (2000) discuss the sensitivity of results to the different assumptions. Madden (2006)
shows that the degree of slackness in the labor market has a large influence on the results of
CGE simulations dealing with mega-events, confirming the findings of other studies (Roson,
1999; Bontout and Jean, 1998; Pant and Warr, 2016), where CGE models have been used in other
contexts.
The effective endowment of production factors can also change through a variation in factor
productivity, which modifies the output per factor. Or when production factors move from less
to more productive sectors: factor reallocation can indeed affect the system productivity. To the
extent that a factor is allocated more efficiently, this results in higher productivity. Iregui (2003)
shows that the elimination of global restrictions on the mobility of labor generates efficiency
gains that could be of considerable magnitude. However, any increase in an element of final
demand triggers many economic processes besides factor reallocation, and the final impact on
the systemic productivity is not necessarily positive. Furthermore, the empirical literature
highlights the existence of severe costs of factor reallocation, mainly caused by non-
transferability and losses in skills, which hinders the mobility between sectors (Figura and
Wascher, 2010).
As for factors endowment, CGE models do need to make assumptions about factors mobility
across the different production activities within a country. To the best of our knowledge, CGE
studies on mega-events do not discuss the implications of different factor mobility hypotheses,
unlike other research fields where CGE models have been extensively employed (i.e.,
international economics, development economics; see e.g. De Melo, 1977; Kilkenny and

10
Robinson, 1990; Iregui, 2003; Savard, 2005). One standard assumption is that both labor and
capital are perfectly mobile between sectors and allocated such that wages ad rents are
equalized across sectors. The assumption of a single homogenous factor market may, however,
overestimate the economic impact of an increase in an element of the final demand. The degree
of factor mobility across sectors does affect the economy’s production response to a demand
change, through the slope of industry supply curves. All else equal, the more mobile the factors
are, the flatter the supply curve is and the larger the supply response to any type of economic
shock. If hosting a mega-event pushes up the demand for investment, the impact on total
production will be larger under perfect factors mobility assumption.
In presence of barriers to inward factors mobility (e.g., factor reallocation costs) or segmented
factors market, different levels marginal productivity imply different wages and rents among
sectors in equilibrium (Flaig et al., 2013). In this case, inter-sectoral factor reallocation must be
governed explicitly in the model. One possibility is introducing a migration function governing
the movement of the factor between sectors (McDonald and Thierfelder, 2009). Migration would
depend on the change in the relative price of the factor across sector, and the responsiveness of
migration to price changes is determined by a migration elasticity parameter. For example,
Deepak, West, and Spreen (2001) define two types of labor, high- and low-skilled labor, with
different migration propensities. The introduction of a segmented labor market requires,
however, assembling a proper dataset, with information about the sectoral employment of the
various types of labor considered and the substitution elasticities.
The degree of factors mobility also matters in relation to the time horizon of the simulation
exercise. While full factor mobility is a reasonable assumption in case of long-run simulations,
short-run simulations would be better carried out under the hypothesis of imperfect factor
mobility, at least for those production factors which are typically partially mobile (e.g., capital).
Hosting a mega-event generates both short- and long-run effects, which take place in different
phases (Madden and Crowe, 1998). As consequence, simulating the long-run impact of
productivity effects would require different factor mobility assumptions then simulating the
short-run impact of a change in consumption patterns.

4.2. Simulation phases: the phase-static CGE modeling solution

The sign and magnitude of the economic impacts generated by a mega-event are determined by
the nature of the event, its context and timing. To illustrate the modeling issues associated with
the various mechanisms at work, we can distinguish three phases, which may be analyzed by
means of a relatively simple CGE setting, in a set of comparative statics simulations. In the
following, we briefly discuss what kind of processes is considered in each phase, and how they
could be simulated into the model. The types of shock, the closure rules adopted and the
expected outcomes are summarized in Table 1.

4.2.1. Pre-event phase

This phase is characterized by the construction of facilities and infrastructure for the event. A
key aspect, determining the overall impact on the economic system, is how the investment plan
is funded. Financial resources may either come from abroad (i.e., international investments or

11
transfers), or internally, through additional savings or taxes, both reducing consumption levels.
In our simulations pertaining to this phase, government and trade balances are endogenous,
whereas the exchange rate and the baseline government spending are held exogenous.
The extra demand for public and private investment generates a pressure on available resources,
pushing up production costs of domestic goods as well as their market prices. Consumers and
firms substitute some domestic goods with imports, getting relatively cheaper. The trade
balance deteriorates which, from a financial perspective, amounts to an inflow of foreign
financial capital.
When more resources are employed in the event-related sectors of the economy, they are
partially drawn away from productive activities elsewhere in the economy (the “crowding out”
effect). As the relative prices of all inputs change, other activities are discouraged. The net effect
on the overall economic welfare and national income is therefore uncertain and may well be
negative, even in the presence of unemployment.

4.2.2. Event phase

During the realization of the event, consumption patterns change with an expansion of event-
related goods and services. Total consumption levels may also increase, with a corresponding
decline in savings. Also, an inflow of international tourists may take place, thereby boosting the
demand for some consumption items. Some displacement effects may come into play as well:
potential visitors and local residents could respond to higher prices of local services by either
going elsewhere or leaving the area for the duration of the event.
New infrastructure and capital stock is now available, since the pre-event construction phase is
completed: marginal productivity of labor and other factors gets higher. If labor (or other
primary resource) is not fully employed, then the economy can expand.

4.2.3. Post-event phase

The good image, possibly gained by the host country, could translate into improved
competitiveness of (some) domestic goods and services in international markets. Quality-
adjusted productivity rises, positively affecting domestic output, final demand and national
income. This effect overlaps with the one already observed during the realization of the event,
linked to the higher stock of capital infrastructure.
After the event, some interest services on the foreign debt may have to be paid back. This may
require a surplus in the government and trade balances.

12
 Table 1. Modeling phases to simulate a mega-event by means a static CGE model.
Phase Main shock(s) Closure Expected outcome
Pre- - expansion of the demand - flexible government - increase in the financial inflows and
event for investment expenditure and trade balance to appreciation of the real exchange rate (trade
in event-related industries allow for mega-event deficit)
financing - increase of household savings and decreased
- sufficient degree of demand for household consumption
slackness in the labor - structural change in the patterns of final
market demand (crowding out effects)
- full factor mobility - loss of competitiveness in international
markets
- lower competitiveness of non-related event
industries
- uncertain effect on GDP
Event - new money from - flexible government - expansion of tourist/mega-event-related
international tourists and trade balance industries
expenditure on event - sufficient degree of - crowding out effects on household
activities (increase in slackness in the labor consumption
international transfers) market - increase in the marginal productivity of labor
- reshuffling of household - sufficient degree of induced by capital stock increase
spending toward event- factor mobility - positive effect on national income due to new
activities; capital stock and international income inflows
- increase in the stock of - reduction of unemployment
capital (investment of the
- increase in government revenues from higher
previous phase)
national income
Post- - increase in the demand of - fixed government - positive effect on total production,
event exports, not induced by and trade balance employment and national income due to new
changes in relative prices surplus capital stock, increase in productivity and
or income - sufficient degree of exports
- interests on foreign and slackness in the labor - improved competitiveness in international
national debt are paid back market markets
- same increase in the stock - full factor mobility
capital from the previous
phase
- intangibles (benefits/costs
without explicit monetary
value)

5. A simple illustrative CGE simulation exercise

To better illustrate how the various processes operate in the different phases, we perform some
simulation exercises with a rather standard, comparative static, single country CGE model. The
economy is here disaggregated into ten industries (agriculture, energy sectors, manufacture,
construction, trade, transport, communication, recreations, government services and other
services), two primary factors (labor and capital), a representative household, a public and a
foreign sector. The structural parameters of the model are estimated assuming that the economy
is in equilibrium in the calibration period. Non-calibrated parameters include the elasticities of

13
substitution; some of them have been taken from the GTAP model (Corong et al., 2017).
Structural parameters of the model are obtained (wherever possible) from a national SAM of
Italy, extracted from the GTAP9 dataset (Aguiar et al., 2016). A complete description of the
model equations and the dataset is provided in the Appendix.8
The pre-event phase is simulated by increasing the demand for investment expenditures. As
mega-events differ in the structure and level of investment required, two different simulations
are considered. In simulation 1A, investment expenditure is increased by 5% in manufacture,
constructions, government services, trade and hotels, and other services. In simulation 1B,
investment expenditure increases in manufacture and trade sectors only, but the total value of
new investment is the same. Total government expenditure is fixed at its baseline level, as well
as the exchange rate. Both trade and government balances are therefore flexible and
endogenously determined.
Construction in the pre-event phase makes additional infrastructural capital available during
and after the event. We therefore assume that the stock of capital increases (+0.5%) in the
following two phases. Additionally, during the event consumption expenditures in event-
related industries (i.e., trade and hotels, transport, communication and recreations activities) are
assumed to grow by 2%.
The post-event phase is simulated through three exogenous shocks. First, the international price
of exports is increased by 0.5%, to simulate the increased foreign willingness to pay for
domestic goods; second, total factor productivity is raised by 0.1%, to simulate the “brand
image” effect; and third, we impose a surplus in both trade and government balances, to repay
the interests on government and foreign debt. Total government expenditure and the exchange
rate are now endogenously determined. For the sake of simplicity, we assume that factors
markets are flexible in all phases, so that labor and capital are fully mobile across sectors. Under
the partial mobility assumption, factors would likely be differently allocated across sectors,
affecting factors’ marginal productivity.
The magnitude of the exogenous shock is relatively arbitrary, but an effort was made to keep it
inside a plausible range for a typical mega-event, like the hosting of Olympic Games. The
simulation described is therefore only one example of the many “recipes” one could simulate.
For example, one could assume that the increase in the stock of capital concerns only the event-
phase, and in the post-event it falls in disrepair.
Results of the simulation exercise are reported in Tables 2 and 3.

8 The model has been coded and implemented with the GAMS language. Code files are available upon request.

14
 Table 2. Results from an illustrative simulation exercise
Pre-event phase Event phase Post-event
Macroeconomic Variables
(sim. 1A) (sim. 1B) (sim. 2) phase (sim. 3)
Household consump. (%var) -1.36 -1.29 0.53 1.19
Total investment (%var) 4.04 2.68 -2.40 1.20
Govern. spending (%var) 0 (exog.) 0 (exog.) 0 (exog.) 0.07
Exports (%var) -0.12 1.18 1.27 1.27
Imports (%var) 0.15 0.23 -0.62 0.85
(Exports - Imports)/GDP (%var) -0.08 0.3 0.6 0.1 (exog.)
(Taxes – Spending)/GDP (%var) 0.02 ≈0 0.04 0.1 (exog.)
Consumer price index (%var) 0.05 -0.15 -0.31 -0.60
Return to capital (%var) 0.07 -0.25 -0.49 0.10
Exchange rate (%var) 0 (exog.) 0 (exog.) 0 (exog.) -0.87
Industry output (%var) -0.03 0.25 0.44 1.08
Real GDP (%var) -0.06 0.20 0.65 1.13
Unemployment (%var) 0.45 -1.53 -3.06 -6

Table 3. % changes in total output volume by sector.

Pre-event phase Event phase Post-event
Sectors (sim 2)
(sim 1A) (sim 1B) phase (sim. 3)
Agriculture -1.0 -0.52 0.53 1.03
Energy sectors -0.06 0.79 1.05 1.24
Manufacture 0.30 3.17 -0.002 1.14
Government Services -0.18 -0.03 0.23 0.65
Constructions 3.13 -1.93 -1.56 1.14
Trade and hotels -0.97 -0.45 1.43 1.27
Transports -0.67 -0.32 1.22 1.21
Communication -0.70 -0.57 1.10 1.15
Recreational activities -1.16 -0.99 1.42 1.18
Other Services -0.20 -0.40 0.27 1.07

Real GDP varies in opposite directions in simulations 1A and 1B (pre-event phase). To

understand why, consider how the labor market is modeled and the degree of labor intensity of
the different sectors.
Total labor supply is given and is the sum of total labor demand and involuntary
unemployment. A wage curve relating real wage to unemployment determines the
unemployment rate endogenously (see Eq. 26 of the Appendix). The presence of unused labor
resources to employ is crucial for the real income growth: GDP increases if labor demand
expands. Indeed, labor demand decreases in simulation 1A but increases in 1B.
To understand why this is the case, consider that any additional element of final demand (extra

15
investment, private or public consumption) must be financed (internally or externally). This is a
major difference between I-O and general equilibrium settings. In I-O models, final demand is
fully exogenous, so any extra expenditure directly adds to the national income. In a general
equilibrium framework, national income is given by the value of the primary resources, so any
increment in an element of final demand must be accompanied by a decrement in some other
element. Consequently, when “new” investment is considered in simulations 1A and 1B, one
should correctly interpret the scenario as one in which total final demand has simply changed
its internal composition, with more demand for some industries and less demand for other
industries.
Clearly, the total effect on labor demand depends on the relative labor intensity of the various
sectors involved. Ranking the industries in terms of labor intensity is not a trivial exercise, but
some information may be obtained by inspecting the labor/capital ratios in our models, which
are shown in Table 4. It appears that manufacturing, government services, trade and hotels
employ relatively more labor than capital. The opposite occurs for transports, other services,
constructions, recreational activities, agriculture and communication. Therefore, real GDP
decreases in simulation 1A because it simulates a relative shift in final demand, which mainly
benefit capital-intensive industries (whereas 1B is the opposite case).

Table 4. Labor/capital ratios (Italy). Source: GTAP v9 dataset

Sectors L/K
Agriculture 0.76
Energy Intensive Sectors 0.96
Manufacture 1.26
Government Services 1.18
Constructions 0.41
Trade and hotels 2.62
Transports 0.21
Communication 0.85
Recreational activities 0.75
Other Services 0.27

In simulation 1A the country shows a trade deficit, as the relative price of domestic goods
increases. The trade deficit mirrors a net inflow of financial capital from abroad. The opposite
occurs in simulation 1B. Only in simulation 1A the resources to finance new investment
expenditures come from both taxation and financial inflows.
In the event phase, simulations show an overall positive impact for the economy activity, which
is stimulated primarily by the increase in the stock of capital. General price level declines, the
international competitiveness of domestic goods and services improves, positively affecting
total exports and the trade balance. Real GDP and employment levels both increase, as well as
total industry output, whose costs of production are now relatively cheaper.
Results of the post-event phase are qualitatively similar to those generated by simulation 2, as

16
they both hinge on higher productivity levels. Larger industrial productivity and increased
foreign demand enhance the positive impact on GDP, employment and industry output.
At the industry level, the simulations of the three phases show significant differences (Table 3).
Some sectors expand, some other contract, depending also on the phase taken into account.
The expansion of the construction and manufacturing industries during the pre-event phase
comes at the expense of all the other sectors (but the energy sectors, that show a positive
increase in simulation 1B). A similar but less evident effect occurs in the event-phase, where the
output of two sectors decreases despite the increase in the stock of capital. Higher
competitiveness in some industries causes lower competitiveness elsewhere in the economy (i.e.,
the Dutch disease). The whole productive structure of the country changes, with consequences
in terms of income and wealth distribution.
Nonetheless, after the event production expands in all sectors. This may look as a rather
positive finding, suggesting that hosting mega-events is always good for the economy, at least
in the long run. But this is simply a logical consequence of the investment, creating new capital
stock. So the question would be better cast as: are returns on event-related infrastructure higher
than alternative investments? Would the economy grow more by hosting a mega-event, or by
improving road infrastructure, higher education institutions, etc.? Obviously, our illustrative
simulation exercise cannot help in this matter. It may only be worth noticing that the
perspective of a large important event is often politically necessary to implement infrastructural
projects, which would be beneficial in any case (e.g., a new underground line).

6. Concluding remarks
Getting back to the original question: would you like to get your country hosting the next
Olympic Games? To answer, you may consider that hosting countries can reap certain benefits,
like an increase in trade, foreign investment, travel and tourism-related spending. They can also
take advantage from the improvement in communication and transportation systems, buildings
and general infrastructure, which increases the productive capacity of the economy. On the
other hand, countries must commit to a considerable investment and many analyses conclude
that hosting a mega-event is rarely profitable. Legacies are also heterogeneous, prone to
political interpretation and difficult to quantify.
CGE models may turn out to be the right tool to sort the various effects at play and to ascertain
the likely net effect, in the short and in the long term. But for the CGE models to work, they
must be appropriately designed to serve as the correct test bed. In other words, one has to
choose the right closure rules and the right variables to shock.
A serious scientific investigation needs testable models, clearly stated assumptions, verifiable
data, accurate interpretation of the results, as well as (in this case) understanding what a CGE
model can and cannot do. Indeed, only a transparent and replicable model exercise can serve as
a “litmus paper”, to ascertain whether hosting a mega-event is good or bad for an economy.

17
References
Aguiar, A., Narayanan, B., and R. McDougall. 2016. “An Overview of the GTAP 9 Data Base”. Journal of Global
Economic Analysis, 1(1): 181-208. doi: 10.21642/JGEA.010103AF.
Allan, G.J., Lecca, P., and Swales K. 2017. “The impact of temporary but anticipated tourism spending: An application
to the Glasgow 2014 Commonwealth Games”, Tourism Management, 59: 325-337.
Andersson, T.D., Armbrecht, J. and Lundberg, E. 2008. “Impact of Mega-Events on the Economy”, Asian Business &
Management, 7: 163-179.
Armington, P. 1969. A theory of demand for products distinguished by place of production. International Monetary Fund
Staff Papers, XVI, 159-78.
Atkinson, G., Mourato, S., Szymanski, S. and Ozdemiroglu, E. 2008. “Are We Willing to Pay Enough to ‘‘Back The
Bid’’?: Valuing the Intangible Impacts of London’s Bid to Host the 2012 Summer Olympic Games”, Urban Studies, 45:
419-444.
Baade, R., and Matheson, V. 2002. “Bidding for the Olympics: Fool’s Gold?”, in C. P. Barros, M. Ibrahimo and S.
Szymanski (eds.) Transatlantic Sport: The Comparative Economics of North American and European Sports, Cheltenham:
Edward Elgar.
Baade, R., and Matheson, V. 2004. “The quest for the Cup: Assessing the economic impact of the World Cup”,
Regional Studies, 38(4), 343–354. doi:10.1080/03434002000213888
Baade, R., and Matheson, V. 2016. “Going for the Gold: the Economics of the Olympics” Journal of Economic Perspective,
30(2): 201-218.
Blake, A. 2005. The Economic Impact of the London 2012 Olympics, Discussion Paper 2005⁄05, Christel DeHaan Tourism
and Travel Research Institute, Nottingham University.
Bohlmann, H.R., and van Heerden, J.H. 2005. The impact of Hosting a Major Sport Event on the South African Economy,
University of Pretoria, Department of Economics Working Paper Series, 2005-09.
Bohlmann, H.R., and van Heerden, J.H. 2008. “Predicting the economic impact of the 2010 FIFA World Cup on South
Africa”, International Journal of Sport Management and Marketing, 3(4): 383-396.
Bontout, B. and Sebastien. J. 1998. Wages and Unemployment: Trade-Off under Different Labour Market Paradigms, CEPII
(Working Paper) Document de travail N. 98-13.
Brunet, F. 1995. “An economic analysis of the Barcelona’92 Olympic Games: resources, financing and impact", in
Miquel de Moragas & Miquel Botella, The Keys to Success: the social, sporting, economic and communications impact of
Barcelona’92. Barcelona: Servei de Publicacions de la UAB, pp. 203-237.
Burfisher, M.E. 2011. Introduction to Computable General Equilibrium models. Cambridge University Press.
Corden, W.M. and Neary, J.P. 1982. “Booming sector and de-industrialisation in a small open economy”, The
Economic Journal, 92: 825-848.
Corong E., Hertel, T.W., McDougall, R.A., Tsigas, M.E. and van der Mensbrugghe, D. 2017. “The Standard GTAP
Model, Version 7”, Journal of Global Economic Analysis, 2(1): 1-119.
Crompton, J.L. 1995. “Economic Impact Analysis of Sports Facilities and Events: Eleven Sources of Misapplication”,
Journal of Sports Management, 9: 14–35.
de Melo, J.A.P. 1977. “Distortions in the Factor Market: Some General Equilibrium Estimates”, The Review of
Economics and Statistics, 59(4): 398-405.
Dwyer, L., P. Forsyth, J. Madden, and Spurr, R. 2000. “Economic Impacts of Inbound Tourism under Different
Assumptions about the Macroeconomy”, Current Issues in Tourism, 3(4): 325–63.
Dwyer, L., P. Forsyth, R. Spurr, and Ho, T. 2003. “The Contribution of Tourism to a State and National Economy: A
Multi-Regional General Equilibrium Analysis”, Tourism Economics, 9(4): 431–48.
Dwyer, L., Forsyth, P. and Spurr, R. 2004. “Evaluating tourism’s economic effects: new and old approaches”, Tourism
Management, 25: 307–17.
Dwyer, L., Forsyth, P., and Spurr, R. 2005. “Estimating the Impacts of Special Events on an Economy,” Journal of
Travel Research, 43: 1-9.

18
Dwyer, L., Forsyth, P., and Spurr, R. 2006. “Assessing the Economic Impacts of Events: A Computable General
Equilibrium Approach”, Journal of Travel Research, 45: 59-66.
Dwyer, L., and Forsyth, P. 2009. “Public sector support for special events”, Eastern Economic Journal, 35: 481–499.
Figura, A., and Wascher, W. 2010. “The Causes and Consequences of Sectoral Reallocation: Evidence from the Early
Twenty-First Century”, Business Economics, 45(1): 49-68.
Flaig, D., Grethe, H. and McDonald, S. 2013. Imperfect labor mobility in a CGE model: does factor specific productivity
matter? Paper presented at the 16th Annual Conference on Global Economic Analysis, Shanghai, China, June 12-14,
2013. Available at https://www.gtap.agecon.purdue.edu/resources/res_display.asp?RecordID=4214.
Giesecke, J.A., and Madden, J.R. 2007. The Sydney Olympics, seven years on: an ex-post dynamic CGE assessment.
CoPS General Paper No. G-168. Centre of Policy Studies, Monash University, Melbourne.
Giesecke, J.A., and Madden, J.R. 2011. “Modelling the economic impacts of the Sydney Olympics in retrospect e
Game over for the bonanza story?”, Economic Papers, 30: 119e287.
Hertel, T.W. (Ed.) 1997. Global Trade Analysis: Modeling and Applications. Cambridge University Press.
Horne, J. 2004. “The global game of football: the 2002 World Cup and regional development in Japan”, Third World
Quarterly, 25(7): 1233–1244.
Horridge, M. 2000. ORANI-G: a generic single-country computable general equilibrium model, CoPS Working Paper OP-93,
Centre of Policy Studies, Monash University.
Iregui, A.M. 2003. Effiency gains from the elimination of global restrictions on labour mobility: An analysis using a
multiregional CGE model, WIDER Discussion Papers // World Institute for Development Economics (UNU-WIDER),
No. 2003/27, ISBN 9291904376.
Jago, L., Dwyer, L., 2006. Economic Evaluation of Special Events: A Practitioners’ Guide, Common Ground Publishers,
Altona, Australia.
Kilkenny, M. and Robinson, S. 1990. “Computable General Equilibrium Analysis of Agricultural Liberalization:
Factor Mobility and Macro Closure”, Journal of Policy Modeling, 12(3): 527-556.
Lee, C.K, and Taylor, T. 2005. “Critical reflections on the economic impact assessment of a mega-event: the case of
2002 FIFA World Cup”, Tourism Management, 26: 595-603.
Li, S.N. (2012) “A simple framework for evaluating the welfare impacts of a large event”, Tourism Analysis, 17(4): 495-
508
Li, S.N. and Jago, L. 2013. “Evaluating economic impacts of major sports events – a meta analysis of the key trends”,
Current Issues in Tourism, 16(6): 591-611. http://dx.doi.org/10.1080/13683500.2012.736482.
Li, S.N., Blake, A., Cooper, C., 2011. “Modelling the economic impact of international tourism on the Chinese
economy: a CGE analysis of the Beijing 2008 Olympics”, Tourism Economics, 17(2): 279-303.
Li, S., Blake, A., and Thomas, R. 2013. “Modelling the economic impact of sports events: The case of the Beijing
Olympics”, Economic Modelling, 30, 235e244.
Li, S.N., and McCabe, S. 2012. "Measuring the socio-economic legacies of mega events: concepts, propositions and
indicators”, International Journal of Tourism Research, 15: 388-402.
Lofgren, H., Harris, R., and Robinson, S. 2002. A Standard Computable General Equilibrium (CGE) Model in GAMS.
International Food Policy Research Institute, Washington.
Madden, J.R. and Crowe, M. 1998. “Estimating the Economic Impact of the Sydney Olympic Games”. Mimeo.
Available at https://www.researchgate.net/publication/23732594.
Madden, J.R. 2002. “The Economic Consequences of the Sydney Olympics: The CREA⁄Arthur Andersen Study”,
Current Issues in Tourism, 5: 7-21.
Madden, J.R. 2006. “Economic and fiscal impacts of mega events: a general equilibrium assessment”, Public Finance
and Management, 6(3): 346-394.
Matheson, V. 2009. “Economic Multiplayers and Mega-Event Analysis”, International Journal of Sport Finance, 4: 63-70.
McDonald, S. and Thierfelder, K. 2009. STAGE_LAB: An Applied General Equilibrium Model with Enhanced Labour
Markets. Mimeo.

19
Müller, M. 2015. “What makes an event a mega-event? Definitions and sizes”. Leisure Studies,
http://dx.doi.org/10.1080/02614367.2014.993333.
Pant, H.M. and Warr, P.G. 2016. “Modelling Involuntary Unemployment in Applied GE Models”. Paper presented to
the 19th Annual Conference on Global Economic Analysis, June 15-17, The World Bank, Washington DC. Available at
https://www.gtap.agecon.purdue.edu/resources/res_display.asp?RecordID=5091.
Pillary, U. and Bass, O. 2008. “Mega-events as a Response to Poverty Reduction: The 2010 FIFA World Cup and its
Urban Development Implications ”, Urban Forum, 19:329–346 DOI 10.1007/s12132-008-9034-9
Preuss, H. 2004. Economics of the Olympic Games, Edward Elgar Publishing.
Robinson, S. 1991. “Macroeconomics, Fonancial variables, and Computable General Equilibrium Models”, World
Development, 19(11): 1509-1525.
Rose, A.K. and Spiegel, M.M. 2011. “The Olympic Effect”, The Economic Journal, 121: 652-677.
Roson, R., and Sartori, M. 2014. “Climate Change, Tourism and Water Resources in the Mediterranean: a General
Equilibrium Analysis”, International Journal of Climate Change Strategies and Management, 6(2): 212-228.
Russo, B. (2009). Innovation and the long-run elasticity of total taxable income. Southern Economic Journal, 75(3),
798-828.
Saayman, M. and Rossouw, R. 2008. “The Economic Value of the 2010 Soccer World Cup”, Acta Commercil, 8(1): 1-14.
Savard, L. 2005. “Poverty and Inequality Analysis within a CGE Framework: A Comparative Analysis of the
Representative Agent and Microsimulation Approaches “,Development Policy Review, 23(3): 313-331.
Shoven, J.B, and Whalley, J. 1984. “Applied General Equilibrium Models of Taxation and International Trade: An
Introduction Survey”, Journal of Economic Literature, 22(3): 1007-1051.
The Guardian. 2002. https://www.theguardian.com/politics/2002/dec/08/athletics.olympics2012. Web access:
September 8th, 2017.
The Guardian. 2012. https://www.theguardian.com/sport/2012/may/09/athens-2004-olympics-athletes-home. Web
access: September 8th, 2017.
The Guardian. 2017. https://www.theguardian.com/sport/gallery/2017/feb/10/rios-olympic-venues-six-months-on-in-
pictures. Web access: August 28th, 2017.
Zimbalist, A. 2015. Circus Maximus: The Economic Gamble Behind Hosting the Olympics and the World Cup. Brookings
Institution Press

20
Appendix

A1. The model

Model parameters
!"#$! = minimum consumption level by sector
!"! = share parameter in the LES utility function by sector
! = marginal propensity to save
!"! = share parameter in the Cobb-Douglass investment function by sector
!"! = share parameter in the Cobb-Douglass government function by sector
!"! = CES production function share parameter by sector
!"! = CES production function efficiency parameter by sector
!"! = CES capital-labor substitution-elasticities by sector
!"!" = input-output technical coeffcient
!"! = CES Armington production function share parameter by sector
!"! = CES Armington production function efficiency parameter by sector
!"! = CES Armington capital-labor substitution-elasticities by sector
!"! = CET production function share parameter by sector
!"! = CET production function efficiency parameter by sector
!"! = CET capital-labor substitution-elasticities by sector
!! = tax rate on total output by sector
! = Phillips parameter of the wage curve
!"#!! = initial international price of imports
!"#!! = initial international price of exports
!"#$%&'! = initial consumer price index
!"#$%! = initial unemployment level
!! = initial wage rate
!"! = initial stock of labor
!!! = initial consumption level by sector
!!! = initial prices of composite commodities

Model variables
! = return to capital
! = wage rate
!! = prices of composite commodities
!"! = domestic producer prices of commodities
!""! = price of domestic output delivered to home market
!"! = export prices in national currency
!"! = import prices in national currency
!"#$%&' = consumer price index
!" = capital stock (endowment)

21
!" = labor stock (endowment)
!! = domestic sales composite commodity
!"! = gross domestic output
!! = exports
!! = imports
!""! = domestic output delivered to home market
!"#$ = trade balance
!! = capital demand
!! = labor demand
!! = consumer demand
!"#$ = consumer expenditure
!"#$% = involuntary unemployment
! = household income
!" = household savings
! = total savings
!! = investment demand
!"#$ = government balance
!! = government demand by sector
!"#" = total government demand aggregate
!"#$ = total tax revenues

Model equations
Consumer demand for commodities
!! !! = !! !"#$! + !"! !"#$ − ! !"#$! !! (1)
Household savings
!" = !" (2)
Capital demand function by sector
!"! (!!" ) !"! !!!"! !"! !!!"!
!! = !"! ! ! [!"! ! + 1 − !"! ! ]!"! !!!"!
(!"! !"! ) (3)
Labor demand function by sector
!"!
!! = (1 − !"! )!"! ! (!!"!) [!"! ! !!!"!
+ 1 − !"! !"!
! !!!"!
]!"! !!!"!
!"! !"! (4)
Zero profit condition
!"! ∙ !"! = (!!! + !!! + ! !"!" !"! !! ) ∙ (1 + !"! ) (5)
Total Savings
! = !" + !"#$%&'×!" ∙ !"#$ + !" ∙ !"#$ (6)
Investment demand function by sector
!! !! = !"! ∙ ! (7)
Government demand function by sector
!! !! = !"! ∙ !"#" (8)
Total tax revenues
!"!
!"#$ = ! !! ∙ !"! ∙ (9)
!!!!

Government balance

22
!"#$%&' ∙ !"#$ = !"#$ − !"#" (10)
Export supply
!"! !"! !"! !"! !!!"! !"! !!!"!
!"! !!!"!
!! = ∙ ∙ !"! ∙ !"! + 1 − !"! ∙ !""! (11)
!"! !"!

Domestic supply of domestic good

!"! !!!"! !"! !"! !!!"! !"! !!!"!
!"! !!!"!
!""! = ∙ ∙ !"! ∙ !"! + 1 − !"! ∙ !""! (12)
!"! !""!

CET zero profit condition

!"! !"! = !"! !! + !""! !""! (13)
Import demand
!! !"! !"! !"! !!!"! !"! !!!"!
!"! !!!"!
!! = ∙ ∙ !"! ∙ !"! + 1 − !"! ∙ !""! (14)
!"! !"!

Demand for domestic good

!! !!!"! !"! !"! !!!"! !"! !!!"!
!"! !!!"!
!""! = ∙ ∙ !"! ∙ !"! + 1 − !"! ∙ !""! (15)
!"! !""!

Armington zero profit condition

!! !! = !"! !! + !""! !""! (16)
Market clearing for labor
! !! = !" − !"#$% (17)
Market clearing for capital
! !! = !" (18)
Market clearing for goods
!! = !! + !! + !! + ! !"!" !"! (19)
Balance of payments
! !! !"#!! − ! !! !"#!! = !"#$ (20)
Export price equation
!"! = !" ∙ !"#!! (21)
Import price equation
!"! = !" ∙ !"#!! (22)
Laspeyres consumer index
! !! !!!
!"#$%&' = (23)
! !!! !!!

Total household income

! = !"# + ! !" − !"#$% (24)
Household expenditure on commodities
!"#$ = ! 1 − ! (25)
Wage curve
! !! !"#$% !"#$%!
=! −1 (26)
!"#$%&' !"#$%&'! !" !"!

23
 A2. Data

Table A1 reports the I-O table used in this study to calibrate the CGE model. Intra-industry flows, household consumption,
investment and government consumption by sector are the same as provided by the GTAP9 dataset. The amount of taxes collected,
imports and exports flows are determined such that the government balance and the trade balance are both in equilibrium in the
calibration year.

Table A1. Italian I-O table used in this study (M$ 2011).
En. Trade, Recr. Other Hh. Gover.
Agric. Intens. Manuf. Gov.Serv. Construc. hotels Transp. Comm. Activ. Serv. Cons. Invest. Cons. Exports
Agric. 76484 5806 8989 3186 157 37711 1559 19 995 2710 112892 112 197 29367
En.Intens. 23624 248738 129422 22383 64612 30928 43463 3989 4196 27474 88881 261 87 179122
Manufac. 3758 31123 154544 15335 14671 23550 9942 3709 1324 14282 238590 118933 1470 112922
Gov.Serv. 1653 4720 3820 36434 3057 5483 954 811 671 4983 65619 1309 444504 19266
Construc. 1172 3605 3056 6208 22279 4928 2393 5117 497 13289 9572 246058 1 22001
Trade, hotels 10765 27876 28836 10821 7994 49089 8107 1750 2369 17366 252617 6663 1056 42325
Transports 5706 12315 12785 3070 6861 26168 36044 2633 1026 15872 49099 344 3713 22576
Comm. 1164 4063 3966 6683 3190 9715 3211 2089 1547 15411 30010 2 3 4437
Recr.Activ. 2345 3880 3009 3763 1158 4850 356 763 16443 4387 66453 2599 2920 4731
Other Serv. 19825 62154 66178 52005 36547 112291 23513 16053 14572 176612 238533 45038 368 32062
Labor 37931 93063 92819 171199 33969 46166 6098 15783 26101 106008
Capital 49820 97077 73845 145127 82579 17626 29531 18680 34799 395364
Taxes 16570 93638 49962 97804 41101 56804 10465 9658 8386 69931
Imports 29367 179122 112922 19266 22001 42325 22576 4437 4731 32062
Source: adapted from GTAP v9 datase

24