economic impacts of the olympic games through …...! 1! economic impacts of the olympic games...

97
1 Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable 1984 Los Angeles Olympic Games, some cities began to compete for and host the Olympics in part to garner economic growth and development. The purpose of this paper is to study current economic cost and benefit analysis, before quantifying if and when benefits exist for the 1992 Barcelona, 1996 Atlanta and 2000 Sydney Games. To assess the economic impact of hosting the Olympics, the host state/region will be compared to a same-country state/region that did not host the Games through the examination of infrastructure, prestige and general financial growth models over a nine-year period. Apart from one section—construction—from the 1992 Barcelona Games, all models appear to show that host states/regions do not have significantly different changes in growth compared to the control states/regions.

Upload: others

Post on 09-Mar-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

  1  

Economic Impacts of the Olympic Games through State Comparison

Samantha Edds*

April 2012

Abstract

After the financially profitable 1984 Los Angeles Olympic Games, some cities began to compete for and host the Olympics in part to garner economic growth and development. The purpose of this paper is to study current economic cost and benefit analysis, before quantifying if and when benefits exist for the 1992 Barcelona, 1996 Atlanta and 2000 Sydney Games. To assess the economic impact of hosting the Olympics, the host state/region will be compared to a same-country state/region that did not host the Games through the examination of infrastructure, prestige and general financial growth models over a nine-year period. Apart from one section—construction—from the 1992 Barcelona Games, all models appear to show that host states/regions do not have significantly different changes in growth compared to the control states/regions.

Page 2: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  2  

Contents 1 Introduction 2 Background: History and Economics 3 Literature Review 4 Methodology 4.1 Methodological Considerations 4.2 Ideal Data Collection 4.3 Independent Variable Ideal Data Collection 4.4 Data Collection 4.4.1 Spain Variables 4.4.2 United States Variables 4.4.3 Australia Variables 5 Model and Analysis 5.1 Model Implications Overview 5.2 1992 Barcelona Games Analysis 5.2.1 Spain Construction 5.2.2 Spain Tourism 5.2.3 Spain Financial Services 5.2.4 Spain Conclusions 5.3 1996 Atlanta Games Analysis 5.3.1 U.S. Construction 5.3.2 U.S. Tourism 5.3.3 U.S. Financial Services 5.3.4 U.S. Conclusions 5.4 2000 Sydney Games Analysis 5.4.1 Australia Construction 5.4.2 Australia Tourism 5.4.3 Australia Financial Services 5.4.4 Australia Conclusions 6 Summary and Conclusions 7 Policy Implications

Page 3: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  3  

References Appendix A: Comparison Explanations A.1 Spain A.2 United States A.3 Australia Appendix B: Methodological Changes B.1 Spain B.2 United States B.3 Australia Appendix C: Regressions C.1 Spain C.2 United States C.3 Australia

Page 4: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  4  

Introduction

In recent memory, hosting the Olympic and Paralympic Games* has been considered

highly desirable for host cities. Along with irresistible lure of economic prosperity, prestige and

other long-term benefits, cities have thus vied aggressively for the coveted hosting role. To help

entice their constituents and plead their case to the International Olympic Committee (IOC),

cities commission economic impact studies to estimate the seemingly vast economic benefits for

different industries through hosting, from increasing tourism rates to job creation.

Economic impact studies attempt to measure the costs and benefits of hosting the Games;

however, they use complicated modeling and do not account for the impact of the IOC to

potentially induce cost coverage and distort incentives, which may lead the studies to incorrectly

model the economic costs and benefits of the Olympics. The IOC solicits bids, and as Syzmanski

and Giesecke and Madden† show, it appears host cities bid away all or almost all future benefit

because the IOC “can induce host cities and governments to underwrite most of the costs while

keeping for itself a large part of the revenues.”‡ Given the IOC’s actions it seems that becoming

an Olympic host city diminishes the positive economic effects originally anticipated at the

beginning of the bidding process and detailed in the proposal. Additionally, the IOC has

allegedly committed fraud and accepted bribes, further distorting the incentives to host.§ Yet,

                                                                                                               * Samantha Edds, Undergraduate, International Studies Department, University of Chicago, 5801 S. Ellis Ave., Chicago, IL 60637, e-mail: [email protected] * This is the official International Olympic Committee (IOC) designation. From this point forward in the paper I will employ similar terms, such as Olympics or Olympic Games. † James A. Giesceke, and John R. Madden. "Modelling the Economic Impacts of the Sydney Olympics in Retrospect ? Game Over for the Bonanza Story?*.” Economic Papers: A journal of applied economics and policy 30, No. 2 (2011): 230. ‡ Szymanski, Stefan. Playbooks And Checkbooks: An Introduction To The Economics Of Modern Sports. Princeton : Princeton University Press, 2009: 160. § Ibid.158.

Page 5: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  5  

despite the IOC’s bidding process and overestimated benefits, some economists believe hosting

the Olympics still has significant benefits.**

Evaluating these benefits, however, has its own set of challenges. First, economic effect

and benefit need to be defined to provide a clear understanding of what this paper measures

compared to other types of studies. Economic effects of the Olympics include costs and revenues,

otherwise known as benefits, which exist because of the Olympic Games. For example, in order

to host the Olympics there is often a large amount of construction completed solely for the

Games. This is a cost of the Olympics. As with any other structure, ones used for the Olympics

will then be occupied, which generates revenue. One way Olympic benefits are measured is

through consumer spending, otherwise known as expenditures. People who come to the host city

for the Olympics will spend money on activities such as hotels, the Games, eating, and shopping,

all of which generates revenue. Additionally, because of the Olympics, people may later visit the

host city and/or firms may move there, leading to more construction activity that can be seen as a

benefit of the Olympics. Thus, the economic effects of the Olympics, for this study, are created

by changes in the growth and development of infrastructure (construction), prestige (tourism),

and financial services.†† However, as will be discussed below, calculating costs and revenues for

the Olympics is not just calculating all visitors’ spending throughout the Olympics and

comparing it to the cost of construction and changes completed solely for the Olympics; there are

complicated pitfalls that can lead to overestimated benefits.

In addition to the IOC’s ability to skew incentives, which is not registered by economic

impact studies, mathematical inaccuracies further distort the studies’ stated benefits.

                                                                                                               ** Jeffrey M. Humpreys and Michael K. Plummer, “The Economic Impact On the State of Georgia of Hosting the 1996 Summer Olympics” (diss., The University of Georgia, 1995), 1-133, http://www.terry.uga.edu/selig/docs/olympics.pdf (accessed March 8, 2012). ††  From this point forward discussion of construction will refer to infrastructure, tourism will refer to prestige and financial data will refer to financial services.

Page 6: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  6  

Mathematical overestimations, such as the substitution effect‡‡ crowding-out effect,§§ or

mathematical multiplier*** calculations, misrepresent the predicted costs and benefits of

economic impact studies because they incorrectly measure these effects, which are often

perpetuated when applied to the host city’s overall economy. If focusing on development and

international awareness, it is likely these studies could convince cities to bid for the right to host

future Olympics based on falsities. In order to avoid these inaccuracies, this paper uses a

different form of measuring cost and benefits with a relative comparison model, which does not

require distinction between types of visitor spending and mathematical multipliers. Instead, the

paper studies relative changes that look at the economic impact of specific variables with all

things held equal and compares differences. The comparative model offers a more relative

measure of economic effects, and it will avoid many mathematical measurement issues leading

to a more accurate study.

Since modeling error and the IOC’s role are ignored or incorrect, impact studies will

overestimate the positive economic effect of the Olympic Games on host cities. This leads to the

question: how can the monetary impact of the Olympics be correctly modeled? To correct for

inherent flaws, economists have begun to run post-Olympic models. Examining costs and

benefits through post-Olympic models will allow authors to reconcile the differences in pre-

                                                                                                               ‡‡ Some people have visits that coincide with the Olympics, or were slightly altered to time the visit with the Olympics, but regardless these visitors would have traveled to the host  city. Because of this, the visitors  would have spent money on things such as eating, hotels and tourist attractions independent of the Olympics. Therefore the visitors are substituting one activity, in this case watching an Olympic event, for another, what they would be doing if the Olympics were not in the host city. Matheson, Victor. “Upon Further Review: An Examination of Sporting Event Economic Impact Studies.” Sports Journal 5, no.1 (2002): 2. §§ The Olympics acts as deterrence likely due to reports of potential crowding and congestion that divert visitors and spending away from the host city. Ibid.2.  ***  Money  spent  on  the  Olympics  that  circulates  through  the  economy,  is  created  into  a  mathematical  multipler  to  account  for  the  effect  of  each  dollar  that  continues  to  circulate  through  the  economy,  acting  as  a  stimulus.  Studies  often  do  not  distinguish  dollars  that  stay  in  the  local  economy  and  dollars  that  will  go  to  corporations  and  nonlocal  vendors,  and  thus  do  not  count  toward  the  money  multiplier.  Ibid.2.  

Page 7: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  7  

Olympic impact studies; accurate data and new models account for the issues Matheson stated.†††

Authors have moved towards precise comparison models with specific variable selection instead

of general modeling such as Computable General Equilibrium modeling (CGE)‡‡‡ to account for

model inaccuracies. While there are many studies that now use comparative modeling, how can

these studies best provide cities with a clear picture of the economic effects of hosting the

Olympics?

Given the issues and inaccuracies of current economic effect studies, does measuring for

changes in prestige, infrastructure, and general financial growth produce a more accurate picture

of the economic gains that result for Olympic host cities?

The purpose of this study is to make cities more aware of the discrepancy in projected

benefits, so cities are able to better understand the difference between estimated and actual

benefits when they consider bidding. Understanding modeling discrepancy patterns, as well as

the actual economic costs and benefits of hosting the Olympics will allow constituents who stand

to gain from hosting and cities to make decisions with eyes wide open. Knowing the true

opportunity cost of hosting provides cities and developers, hotels and other firms accurate

information to either bid while acknowledging potential economic losses, or to use money that

would be spent on the Olympics on the next best alternative. For example, funds spent on the

Games could be used for other infrastructure, transportation networks, parks and recreation or

even on crime prevention that with the existence of the Olympics would not exist.

                                                                                                               ††† Ibid.230. ‡‡‡  CGE  modeling  is  one  of  the  main  ways  to  estimate  the  potential  economic  effects  a  host  city  receives  as  a  result  of  hosting  the  Olympics.  It  involves  studying  the  disaggregated  forms  of  the  economy  through  sector  examination,  creating  a  model  that  Kasimati  notes  is  difficult  to  compute. Kasimati, Evangelia. "Economic Aspects and the Summer Olympics: A Review of Related Research." International Journal of Tourism Research 5.6 (2003): 433-44. Print.  

Page 8: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  8  

While there exists a number of post-Olympic studies and comparative model use,

particularly the works by Billings and Holladay, Giesecke and Madden, and Rose and Spiegel,

my model aims to create greater specificity and accuracy, by comparing similar Olympic Games’

host cities to same-country sister-cities at a statewide level to control for countrywide

cofounding effects. Because methods of recordkeeping do not provide city level specific data for

my dependent or independent variables, they will be state or regional specific, based upon case

and definition. Since countrywide effects are held constant and each city is the largest within the

comparable studied area, the models may be able to see effects attributable solely to the

Olympics Games. I will use a comparison model, similar to the approach of Giesceke and

Madden, and Rose and Spiegel because there is less likely to be inaccuracies from data collection

and mathematical modeling as with CGE modeling. Compared to other models, my study model

attempts to provide greater specificity through relative comparisons and greater accuracy with by

holding countrywide effects constant and modeling that does not require difficult to calculate

multipliers and effects such as substitution.

After running the models, results show that only the 1992 Barcelona Games construction

model has a positive economic effect that could be attributable to the Olympics; all other

construction, tourism and financial services models show either no difference or less change in

growth from the control models. In a number of cases, the control model has greater change in

economic growth than the Olympic model. Because there is likely an amount of inaccuracy due

to measurement changes and the indication of identity variables,§§§ it is not possible to say with

total certainty that the 1992, 1996 and 2000 Olympic Games lead to essentially no positive

economic growth, and in fact often hamper industry growth.

                                                                                                               §§§ Identities happen when the independent variables are completely correlated with the dependent variables, thus there exists no difference between them.

Page 9: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  9  

To best understand and analyze the study’s model of the economic effects of the

Olympics, first the historical background and current models will be described, then an

explanation of why this study is using a slightly different approach, followed by a description of

ideal and actual data, model analysis, results and implications. The literature review will focus

on types of economic cost and benefits studies that look at the Games retroactively, often

through comparative modeling, on which my model is based, instead of predictive modeling to

understand the actual costs and benefits of the Olympics. My model is derived from these current

comparative modeling studies, with the aim of providing further specificity and accuracy. After

describing the study’s model, there will be a section on ideal data collection to better explain the

shortcomings and aim of this paper, followed by a description of actual proxy data assessed in

the model.

The results section will interpret the study’s model in an attempt to better understand

trends within the Olympics and a relative comparison of hosting the Olympics versus not hosting

for changes in construction, tourism, and financial growth. In my conclusion I will further

discuss the limitations to my analysis, including possible measuring inaccuracies and problems

that may affect results, before presenting a section on policy and research implications. The

historical background of the Olympics and current models provide a basis for the study’s model,

which aims for greater specificity and accuracy; results and a conclusion that describes further

policy implications will help the constituents and cities considering bidding in the future better

understand the costs and benefits of hosting the Olympics.

Page 10: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  10  

2 Background: History and Economics

Each Olympic Game takes place under different circumstances; there are a litany of

reasons why cities host the Olympics, some, such as Beijing, are meant to show city and country

power, while others host for international awareness and yet others host for economic gains. This

paper focuses solely on economic gains as the main reason for hosting the Olympics, thus the

conclusions drawn about economic effects and feasibility will not be applicable to host cities

such as Beijing.

The Olympics Games began in 1896 as a small tradition fusing athletic feats and

international unity. Over time, the Games have evolved into a display of wealth and economic

development. The myths of economic prosperity and development**** have become prominent

only after the 1984 Los Angeles Olympics, the only time in which a city did not face competition

from other bidders. Below is a short history of the past Olympic legacies to better understand

how the Games have evolved and why the paper studies the 1992, 1996 and 2000 Games.

Understanding the ever-changing history of the Olympics will show the reader why the 1992,

1996 and 2000 Olympics are the most similar and represent a shift in reason for hosting to

economic development, which remains one of the host city focal points.

After World War II, the Olympics until 1968 had fairly low attendance as the world

began to rebuild and adjust to changes. In 1968 the Summer Games were held in Mexico City.

The legacy of the Games is not positive, during 1968 there is the War in Vietnam and anti-war

protesters, political and civil rights issues and assassinations in the U.S., and general protests

throughout the world.†††† During the 1972 Games in Munich, Israeli athletes were murdered.

                                                                                                               ****  Jeffrey M. Humpreys and Michael K. Plummer, “The Economic Impact On the State of Georgia of Hosting the 1996 Summer Olympics.”:1-2.  ††††  Allen Sanderson, “Olympics Lecture” (lecture, University of Chicago, Chicago, IL, January 6, 2012).

Page 11: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  11  

The 1976 Montreal Games will be remembered for overrun costs and the poor execution

of the Games.‡‡‡‡ Because of this, in 1978, when the host city for 1984 was selected, Los

Angeles was the only city to bid, meaning it did not have to ‘out bid’ another city, keeping its

costs for hosting low, while taking advantage of the high-quality university facilities from

schools such as UCLA and USC.§§§§ Many of the structures constructed solely for the Olympics

are rarely used again after the Olympics given their size, nature of facilities and maintenance

costs, however because Los Angeles had so many Olympic, or near Olympic ready facilities, the

cost of the Olympics was very low. During this time, given the issues from Mexico City to

Montreal, hosting the Games was seen as a burdensome task instead of an honor. After Montreal,

China and much of the West boycotted the 1980 Moscow Games because the Soviets invaded

Afghanistan.***** Similarly, when the 1984 Los Angeles Games happened, the Soviet block

boycotted the Games, further contributing to the small, but inexpensive Olympics. Los Angeles

generated a profit, albeit smaller than usually described.†††††

However, Seoul, the host of the 1988 Olympic Games, was too far into the building and

planning process of their Games to consider the economic development and growth aspect of

hosting. Unlike Seoul, Barcelona was able to realize their economic development goals by

entering into the Common Market/EU and using the Olympics as a catalyst; this is the first

Olympic Games that urban and economic development was considered when bidding to host the

Games.‡‡‡‡‡ Similarly, the 1996 Atlanta Games, focused on development and economic growth,

as did the 2000 Games in Sydney. However, in Sydney most of the Olympic facilities have little

                                                                                                               ‡‡‡‡ Ibid. §§§§ Ibid. ***** Ibid. ††††† Ibid.  ‡‡‡‡‡  Allen Sanderson, “Olympics Lecture” (lecture, University of Chicago, Chicago, IL, January 6, 2012).  

Page 12: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  12  

to no use, but have very high year maintenance costs. §§§§§ 2004 Athens had crippling

expenditures, overrun costs from security concerns due to 9/11/2001 and increasingly elaborate

and expense facility construction in an attempt to advertise the city as a tourist destination.

The 2008 Beijing Games had enormous costs and the government largely decided

everything, to showcase the city and country.****** The next Olympics in the summer of 2012 are

in London; Rio de Janeiro will be the host city in 2016. Given the shift in hosting reasons, and

the level of extenuating different circumstances, I chose the 1992, 1996 and 2000 Games because

they are the most similar Games and represent change and the focus on economic development.

                                                                                                               §§§§§  Dennis Passa, A Year Later Sydney Areas Go Unused, Free Lance Star (Sydney), September 16, 2001.  ******  Allen Sanderson, “Olympics Lecture” (lecture, University of Chicago, Chicago, IL, January 6, 2012).  

Page 13: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  13  

3 Literature Review

Examining current economic modeling and Arthur Andersen’s consulting report provides

the basis for this study’s model and results. Arthur Andersen’s report predicts economic effects

of the Sydney Games, and highlights past trends found in the Atlanta and Barcelona Games.

Atlanta and Barcelona models show the overall economic situation likely has a great impact on

any Olympic effect, meaning any affect attributable to the Olympics is too small to overcome

larger economic trends. While this will be important to consider when examining my results,

current economic modeling studies guide my choice of a same-country state level comparative

model of construction, tourism, financial services and manufacturing variables. By examining

current economic modeling, my study will be able to use greater precision and accuracy than

current models; Arthur Andersen’s study will provide a framework to better gauge and study my

results.

Arthur Andersen, a global consulting firm, conducted a study the year before the Sydney

Olympic Games, which is often cited in later Olympic literature as a reliable source that presents

the difficulty of distinguishing an Olympic effect. The study offers insight into some overarching

economic effects that may be seen in my models comparing NSW to Victoria. The study claims

a $6.5 billion dollar extra economic activity in Australia, with $5.1 alone in New South Wales at

net present value.†††††† Because of this, it appears there would be significant increases in tourism,

construction, and financial services industry incomes in New South Wales compared to Victoria,

as indicated by the changes in predictor variables. However, the study simultaneously notes

NSW income (gross state product) is $77 million dollars lower in the 6 years after the Games

than if the Games had not been held in Sydney, which may impact government willingness to

                                                                                                               ††††††  CREA/Arthur Andersen. “Economic Impact Study of the Sydney 2000 Olympic Games.” Sydney: Arthur Andersen (1999): 1-7.Print.  

Page 14: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  14  

help fund projects, and in turn impact some parts of the construction industry.‡‡‡‡‡‡ According to

the study, if correct, my models should not see a significant change in employment for New

South Wales because of the Olympics. While there appears to be an overall small net positive

economic effect of hosting the Olympics, it is not equal in every industry. If the study’s

employment numbers are correct, the idea that the Olympics leads to greater employment

because more people are drawn to visit the area and live in the area where the Olympics took

place may be misguided and the Olympics actually have a small, short-term effect. The Arthur

Anderson study offers trends that, if true, should be seen in this paper’s model results, such as

varying levels of increased GDP among construction, tourism, and financial services.

The Arthur Andersen study also describes some of the economic trends that the company

discovered about the Atlanta and Barcelona Games. In Atlanta, the study claims there is a tight

labor market, so the city could not maximize potential gains for employment, and furthermore,

the Olympics were unable to overpower Georgia’s already struggling economy at the time.§§§§§§

Knowing that Georgia’s economy was already under pressure at the time of the Olympics may

override any potential Olympic effect. However, in Barcelona, the study notes that the Olympics

and economic activity need to host may have stimulated the economy, which before the

Olympics was in an economic downturn.******* Another possibility for economic stimulation is

Spain entering the EU, which likely had a large, positive economic impact. Because the models

in this study compare autonomous communities within Spain, the EU effect is held constant, and

a significant change in construction, tourism, financial services or manufacturing industry in

Catalonia compared to Madrid AC might be attributable to the Olympics. Similarly, the province

of Barcelona’s (a subset of Catalonia) employment rate went from 23.6% to 11.7% from 1986 to                                                                                                                ‡‡‡‡‡‡  CREA/Arthur Andersen. “Economic Impact Study of the Sydney 2000 Olympic Games.”6.  §§§§§§  ibid.5-6.  *******  ibid.5-­‐6.  

Page 15: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  15  

1992.††††††† By holding countrywide effects constant it may be possible that significant

employment change exists for the construction, tourism and financial services industries. As the

study shows, each Olympics Game takes place under different circumstances and larger

economic shifts and activities may cover or amplify any potential Olympic effect. While Arthur

Andersen’s study is helpful in providing a framework with which to gauge results, it does not

detail a model to create and interpret results, which is why I will examine current post-Olympic

studies.

A number of authors have created post-Olympic studies using different methods and

variables to show the imprecision in economic impact studies and provide a better picture of the

relative economic costs and benefits that host cities receive. Giesecke and Madden use a form of

more accurate CGE modeling, through post-Olympic analysis, judging four major areas of

economic activity, yet their model derives data from different sources and has to account for

difficult to accurately compute data, such as the mathematical multiplier.‡‡‡‡‡‡‡ Rose and Spiegel

use a more accurate model, examining trade openness through country level comparison to a

candidate country,§§§§§§§ yet using only one variable at the country level does not allow for the

examination of more specific Olympic economic effects seen at the city level. Billings and

Holladay use a city level comparison model of host cities to finalist cities, examining population,

two measure of Gross Domestic Product (GDP)******** and trade openness to best account for all

economic indicators, however they are less able to account for countrywide confounding

                                                                                                               †††††††  CREA/Arthur Andersen. “Economic Impact Study of the Sydney 2000 Olympic Games.”5-6.  ‡‡‡‡‡‡‡  James A. Giesceke, and John R. Madden. "Modelling the Economic Impacts of the Sydney Olympics in Retrospect.”: 230.  §§§§§§§ Andrew K. Rose, and Mark M. Spiegel. "The Olympic Effect*." The Economic Journal 121.553 (2011): 664-665. ********  Gross  Domestic  Product:  A  country’s  final  output  of  goods  and  services  produced  within  a  certain  timeframe. Australian Board of Statistics, “Glossary,” Australian Board of Statistics,http://www.abs.gov.au/Ausstats/[email protected]/glossary/5220.0 (accessed March 9, 2012).  

Page 16: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  16  

effects.†††††††† Because of this, and given data availability, my model focuses on a variety of state

level variables thought to best see effects attributable to the Olympics through a same country

comparison.

Giesecke and Madden use CGE modeling for a post-Olympic study, which makes it more

accurate than predictive studies, but still requires complicated mathematical calculations, and

likely generates some model inaccuracies. They examine whether predicted economic benefit

exist for the 2000 Sydney Games by examining Australia and New South Wales state through

CGE modeling, the same type of modeling typically used to conduct economic impact studies.

Their study is most comparable to pre-Olympic evaluations because it uses the same modeling

methodology to best evaluate predicted and actual benefits, although they have two scenarios:

Sydney hosting the Olympics and Sydney not hosting the Olympics. ‡‡‡‡‡‡‡‡ The comparison of

the Olympic model to the non-Olympic model allows the authors to compute relative costs and

benefits by providing outcomes for all possible scenarios. By using the same type of modeling

post-Olympics, the authors are able to correct for modeling errors such as visitor spending

methodology that cause inexactness in CGE models used for the final bid. Giesecke and Madden

found a real consumption loss of $2.1 billion for Sydney.§§§§§§§§ While the authors are able to

determine the monetary loss of hosting the Olympics in Sydney, their model requires very

precise data measurement that otherwise exacerbates the multiplier, one of the most important

parts of the model, for accurate outcomes. The study also derives data from multiple authors,

who might have different data collection and measurement methods, which makes the data and

variable interactions overall less reliable. With such a complicated model there are many ways

                                                                                                               ††††††††  Stephen B. Billings and J. Scott Holladay. "Should Cities Go for the Gold? The Long-Term Impacts of Hosting the Olympics.” Economic inquiry (2011): 1-278. Print.  ‡‡‡‡‡‡‡‡ Giesecke and Madden. “Modeling the Economic Impacts of the Sydney Olympics in Retrospect?” 230. §§§§§§§§ Ibid.230.  

Page 17: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  17  

data can be contaminated and lead to less accurate results. While Giesecke and Madden provide

a comparison to most economic impact studies by using more accurate post-Olympic CGE

modeling, their study still requires measuring for the substitution effect, among others, and uses

data from different authors, which likely creates some inaccuracies in their model.

Giesecke and Madden’s model uses variables that show the economic impact of the

Olympics at both a city and countrywide level. The variables are ones that “encompasses the four

key economic dimensions of the Games: (i) Games operations; (ii) construction of Games

facilities and associated financing; (iii) spending by interstate visitors in the Games year; and (iv)

spending by foreign visitors in the Games year.”********* Examining four major economic

dimensions allows the authors to test seemingly important costs and benefits that will evaluate

overall economic effects from a variety of viewpoints. Games operations will help measure the

receipts and contribution to GDP by the Olympics, which is the main measure of a country’s

economic growth. Construction and financing variables help show the direct impact of the

Games. For example, Olympic facilities were often built solely for the Olympics, never to be

used again, or built on a larger size and time scale. Spending by interstate and foreign visitors

shows the effect of the Olympics in GDP from interstate visitors. All of the variables measure

the real levels of costs and benefits from the Olympics on the city and country in a financial

sense. However, given that almost every study is composed differently, there could be some

level of bias among the variables that could skew outcomes. Additionally, collection methods for

receipts††††††††† are often measured incorrectly because of the “crowding-out” effect.‡‡‡‡‡‡‡‡‡

Overall the variables appear to measure all different aspects of the Olympics at both a city and

country level, albeit in a very complicated model.                                                                                                                *********  Giesecke and Madden. “Modeling the Economic Impacts of the Sydney Olympics in Retrospect?” 225.  †††††††††  Receipts  measure  customer  spend  ‡‡‡‡‡‡‡‡‡  Victor Matheson. “Upon Further Review: An Examination of Sporting Event Economic Impact Studies.”: 2.  

Page 18: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  18  

Rose and Spiegel use a country-to-country comparison methodology instead of CGE

modeling to provide a more relative evaluation for the economic effects of hosting the Olympics.

By comparing the country of a host city to the country of a candidate city, the authors are able to

see the whether the benefits of hosting the Olympics are negligible if the candidate country

receives similar benefits. Rose and Spiegel examine the effects of the Olympics through

increasing exports by quantifying the differences in trade openness between finalist and host

countries for 1956-2006.The authors find that trade openness increases about 20% for each

country.§§§§§§§§§ Rose and Spiegel show while trade benefits do exist for the country in which the

Olympics take place, they similarly exist in countries with a finalist host city.********** Because of

this, the cost of the Olympics becomes more apparent as well as the amount of benefit received

by the host country attributable to the Olympics. However, Rose and Spiegel’s model is based on

the country data. Because of this, it is very possible that the Olympic or candidate city has little

overall effect on the results, because the country has other big cities and the Olympics is such a

small percentage in city, let alone state or country revenue that it will not be visible. Similarly,

the authors’ model is more likely to have confounding factors that contribute to the country’s

trade openness which are not caused by the Olympics and are likely more related to trade

openness. While Rose and Spiegel’s model creates a relative comparison to a candidate country,

therefore bypassing more complicated mathematical modeling required for CGE, comparing

trade openness at a countrywide level may cover up any effect from host and candidate cities that

could be seen through a city level comparison.

Unlike Giesceke and Madden, Rose and Spiegel use only one variable, trade openness, in

their model to determine the relative costs and benefits of the Olympics. The authors’ use of                                                                                                                §§§§§§§§§ Andrew K. Rose, and Mark M. Spiegel. "The Olympic Effect*." The Economic Journal 121.553 (2011): 664-665. **********  Ibid.664.  

Page 19: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  19  

trade openness shows how active a country is on an international level, and trade is intimately

tied to economic growth and GDP. Trade openness may show relative costs and benefits

attributable to the Olympics on a countrywide scale, but not a citywide scale. It is a somewhat

indirect way to show economic growth and GDP increase, if that is the authors’ intent. In general

the model does not appear to be adequate because it is testing only form of economic growth,

GDP and relative costs and benefits. Additionally, trade openness only tests a countrywide

indicator, not city level ones that may provide more specific insight into how and what industries

the Olympics significantly affect. While trade openness indicates international awareness and

legitimacy, because it is the only variable, and measured at a countrywide level, there are

different types of economic costs and benefits, which cannot be measured that show a more full

picture of the Olympics’ economic impact.

For greater accuracy, Billings and Holladay use city level comparisons to illustrate the

relative economic effects of hosting the Olympics for specific variables. The authors compare

host and candidate cities to test whether the host city gains long-term economic growth

attributable to the Olympics.†††††††††† Their use of city level comparison provides the most direct

picture of real economic costs and benefits relative to a control city, because regional and

countrywide confounders are not included in the data, which could hide or dilute the effect of the

host and control cities. One issue comparing host cities to finalist cities is potential confounding

effects, such as countrywide events that effect one country and the costs or benefits outcome, but

not the other. Confounding effects make the Olympic or finalist city data seem more significant

compared to the other. While the authors may need to better account for potential confounders in

                                                                                                               ††††††††††  Stephen B. Billings and J. Scott Holladay. "Should Cities Go for the Gold? The Long-Term Impacts of Hosting the Olympics.”:1-278.  

Page 20: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  20  

their model to produce the most accurate results their use of city level comparison provides the

most specific relative comparison model.

Billings and Holladay use variables that are theoretically similar to Giesecke and Madden,

testing both city and country level economic effects. The authors test two measures of population,

real GDP per capita and trade openness on the city-level, comparing host cities and

finalists.‡‡‡‡‡‡‡‡‡‡ The authors found there was little to no statistical significance between the two

groups, from the 1970s onwards because of escalating costs and the ability of the International

Olympic Committee to further increase costs, cities to bid away benefits.§§§§§§§§§§ The use of

population is likely trying to examine either if employment increased significantly because of the

Olympics or if more people moved to a city because of an increase in popularity and awareness,

statistic deemed important by many economists. Real GDP per capita measures the change in

total goods and services produced in the country over a certain time span per person in the given

population, which is a way to test if city and country income increased because of hosting. As

mentioned in Rose and Spiegel’s model trade openness is another way to measure economic

growth and real GDP. Again, the variables used all take slightly different approaches in an

attempt to form a full picture and understanding of the relative cost and benefits of the Olympics

at a city level with population measures, and at a country level with real GDP per capita and

trade openness.

                                                                                                               ‡‡‡‡‡‡‡‡‡‡ Stephen B. Billings and J. Scott Holladay. "Should Cities Go for the Gold? The Long-Term Impacts of Hosting the Olympics.”: 1-278.  §§§§§§§§§§  Ibid.2.  

Page 21: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  21  

4 Methodology

While the authors’ studies continually evolve in specificity through comparative

modeling and accuracy through greater variable reach to explain the depth of economic effect

attributable to the Olympics, my study attempts to provide a greater level of specificity and

accuracy. Giesecke and Madden’s model uses difficult to compute effects such as the

mathematical multiplier and uses data from multiple sources, which may affect their variables.

Rose and Spiegel offer a more specific model through country level comparison of Olympic

countries to candidate countries, but only use one variable, trade openness, to measure the effects

of the Olympics. The best model appears to be Billings and Holladay, comparing Olympic cities

to finalist cities through population, two measures of GDP and trade openness, accounting for

many different areas where the Olympic effect may be seen. My model attempts to similarly use

a wide variety of variables in a comparative model, but comparing host cities to same-country

sister-cities to hold countrywide effects constant.

I will quantify the economic value of prestige, infrastructure, and financial services to

determine if there exists an economic payoff by comparing Olympic cities at the state level to

same-country states. While the study planned on providing greater specificity in order to produce

a more precise evaluation of the Olympic Games’ economic effects through same-country sister

city analysis, the methods of record keeping does not provide the necessary data for city level

comparisons. My study will still use comparative analysis instead of CGE modeling for greater

precision because there is less chance of model error through the seemingly simplified approach

and greater accuracy through relative comparison, except with comparative states or regions.

Comparing host and same country sister states or regions offers a somewhat precise picture of

Page 22: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  22  

economic effects attributable to the Olympics, because countrywide effects are held constant,

which may offer a more precise model than Billings and Holladay.

4.1 Methodological Considerations

Economic impact studies are predictive and have inherent bias because they are created

in favor of hosting the Games, misrepresenting and overestimating many benefits through

theoretical and mathematical errors. Olympic impact studies used to be computed by simplistic

Input-Output models***********, but more recently economists have been using CGE models,

which are relatively more accurate, but still have a large number of deficiencies.†††††††††††

Matheson’s work “Mega-events: The Effect of the World’s Biggest Sporting Events On Local,

Regional, and National Economies,” examines the main effects that the modeling used in impact

studies ignores. Modeling typically ignores or overstates the substitution effect,‡‡‡‡‡‡‡‡‡‡‡

understates or ignores the “crowding-out” effect,§§§§§§§§§§§ especially in already popular tourist

destinations, and incorrectly calculates the effect of expenditures by overestimating the

mathematical money multiplier. ************ Aside from intangibles, such as prestige or personal

utility, the majority of cities describe their reason to host based on urban economic development,

or the potential to increase international awareness. However, given modeling inaccuracies and

the IOC’s role in bidding, cities will often not obtain their desired outcome.

I initially chose control cities located in the same country as host cities based upon

greatest similarity to host cities at the time of the Olympics in size, growth rate, and

                                                                                                               ***********  Input-­‐Output  models  (I-­‐O  models)  use  production  and  input-­‐output  tables,  but  ignore  regression  analysis  that  is  used  for  behavior  equations,  such  as  substitution  and  crowding  out  effects.  Evangelia Kasimati, “Economic Aspects and the Summer Olympics: A Review of Related Research,” International Journal of Tourism Research 5 (2003): 433-44, http://people.bath.ac.uk/ecpek/index_files/Paper1.pdf. ††††††††††† Ibid.442-444. ‡‡‡‡‡‡‡‡‡‡‡  Victor Matheson, “Mega-events: The Effect of the World’s Biggest Sporting Events On Local, Regional, and National Economies,” College of Holy Cross Faculty Research Series (October 2006): 1-31, http://casgroup.fiu.edu/pages/docs/2744/1277904942_matheson_events.pdf. §§§§§§§§§§§  Ibid.10  ************  Ibid.8-9.  

Page 23: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  23  

industry.†††††††††††† However, because the data are not available, I will compare the states or

regions, which are the most comparable. Because the host or sister city is the largest city among

the studied area, an effect due to the Olympics may be visible, and the variables still inherently

control for countrywide confounders, such as Spain entering the EU in 1986.

The economic effects of the Olympics are evaluated over a nine-year time period, four

years before the Olympics, the year the Games occurred, and four years after the Olympics to

best judge both short and long-term effects. The nine-year period is chosen based on the work of

Matheson, which notes the Olympic effect eventually become ambiguous after the Games

conclude.‡‡‡‡‡‡‡‡‡‡‡‡

This study’s use of dependent variables, construction, tourism and financial services, is

still fairly theoretically similar to Giesecke and Madden, with the aim of capturing any Olympic

effects by studying the change and growth in a wide variety of industries. However, as

mentioned earlier, the dependent variables will instead be state or regional comparisons.

                                                                                                               ††††††††††††  See  Appendix  A  ‡‡‡‡‡‡‡‡‡‡‡‡    Victor Matheson, “Mega-events: The Effect of the World’s Biggest Sporting Events On Local, Regional, and National Economies,” College of Holy Cross Faculty Research Series (October 2006): 1-31, http://casgroup.fiu.edu/pages/docs/2744/1277904942_matheson_events.pdf.

Page 24: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  24  

4.2 Ideal Data Collection

While this paper aims to examine the monetary impact of the Games through the

potential increase in construction, tourism, and financial services industry growth at a

comparative city level, any changes attributable to the Olympics may be too small to see. As all

of the host and sister cities have large economies, factors beyond the Games can also drive

growth in construction, tourism, and financial services income. In construction, for instance, at

the same time Olympic development takes place, hundreds of other buildings and projects are

happening at the same time. If Games related construction is a small part of the city’s

construction GDP, change on a state/region level will be even harder to see. For example,

Barcelona, the largest city in Catalonia, makes up 30 percent of Catalonia’s GDP,§§§§§§§§§§§§

which may not be enough to see Barcelona’s construction economic impact on the region. The

same idea applies to tourists visiting the city and spending money, as well as financial holdings

and investment. Hotels, restaurants and tourist attractions are constructed and used mostly for

tourists visiting the city, not because of the 10-day Olympic event. If economic growth is hard to

see on the city level, it will be difficult to see changes attributable to the Olympics, even when

cities such as Atlanta make up 65 percent of state GDP.************* Overall, for all industries

studied in this paper, the Olympics are only a small proportion of that industry’s GDP relative to

other factors and the overall economy. Thus, it is likely, compared to the main driving forces of

the construction, tourism, and financial services, the Olympics has a very small impact that may

not be visible at a city or state level.

However, the main factors that drive construction, tourism and financial services GDP

are in a sense held almost constant because they will similarly impact the industries before and

                                                                                                               §§§§§§§§§§§§  See Appendix A ************* See Appendix A  

Page 25: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  25  

after the Olympics. Most people will similarly visit and move to host cities completely

independent of the Olympics, as they did before the Games were announced. Because these

trends will likely be steady before and after the Games, any small variation in GDP may be due

to them, because other changes are in a sense constant. If there is a true difference in

construction, tourism or financial services GDP caused by hosting there should be a significant

difference between the changes in GDP from host to sister city because the Olympic city should

have more people come to visit, stay for longer periods of time, and move to the city beyond

normal growth patterns and trajectories.

A same-country sister city comparison methodology will best show if any change in

studied industries is due to the Olympics, however, there was not enough useable data to run

complete models for any of the Olympic Games. In the Spain, the United States, and Australia

construction, tourism and financial services data were not available at a city level for any

necessary 9-year periods, as confirmed by government sources††††††††††††† and independent

economists‡‡‡‡‡‡‡‡‡‡‡‡‡ in each country.§§§§§§§§§§§§§ Because city level data are not available, the

study will provide state/region proxies, which are most similar. However, ideal data will also be

described to best understand proxy data shortcomings. While city level data is ideal, based on

data availability, state level data most similar to city level data is used.

                                                                                                               †††††††††††††  In  an  email  message  to  the  author  on  January  17,  2012  the  Bureau  of  Economic  Analysis  economist  Ralph  Rodriguez  confirmed  that  GDP  by  metropolitan  area  series  only  exists  from  2001-­‐2010.  Ralph  Rodriguez,  email  message  to  author,  January  17,2012.  ‡‡‡‡‡‡‡‡‡‡‡‡‡  In  an  email  message  to  the  author  on  January  19,  2012  Spanish  economist  Levi  Perez  confirmed  that  municipality  data  likely  does  not  exist  for  1988-­‐1996.  This  was  later  confirmed  by  a  different  Spanish  economist  Jaume  Garcia.  Levi  Perez,  email  message  to  the  author,  January  19,  2012.  Juame  Garcia,  email  message  to  the  author,  February  6,  2012.  §§§§§§§§§§§§§  The  Australian  Board  of  Statistics  (ABS)  confirmed  that  statistical  area  data  does  not  exist  for  the  necessary  dates.  Australian Board of Statistics, “National Regional Profile: Past and Future Releases,” Australian Board of Statistics, http://www.abs.gov.au/AUSSTATS/[email protected]/Lookup/Past+Future+Issues2006-2010?OpenDocument&tabname=Past%20Future%20Issues&prodno=LGA17200&issue=2006-2010&num=&view=&(accessed March 9, 2012).    

Page 26: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  26  

Ideal dependent variable data would be aggregated GDP for construction, tourism, and

financial services. Construction GDP would include building construction, heavy engineering,

and special contractors’ construction income.************** Within the variable there would be

three costs: all infrastructure built solely for the Olympics, the opportunity cost of building the

infrastructure solely for Olympic use, and infrastructure that would be built regardless of the

Olympics, but may have been altered or created on an expedited timescale. Opportunity cost is

money that would have been spent elsewhere if the Games did not occur because the stadiums

and facilities built solely for them would not be necessary. Thus, projects that would have

happened are also counted in construction GDP to account for opportunity cost. Opportunity cost

is often overlooked in Olympic economic modeling, but it is important to consider because cities

could have spent money used for the Olympics elsewhere, which is why it would be included in

construction GDP. Additionally, after the event, construction GDP may grow because, as a long-

term effect, more people are moving to the host city because of Olympic awareness, which can

be seen through deviation in typical growth trends.

Tourism GDP would include income from hotels and other accommodations, restaurants,

gaming, and tourist attractions. While tourism is used to explain prestige there are other

measures which are also considered prestige, such as trade and the number of new conferences.

Data for these ideas do not exist for the necessary years at the state level, thus prestige will be

narrowly defined as tourism. However, it seems likely that if a city increases prestige and

therefore awareness from hosting the Olympics, there will be a large increase in visitors and

tourist expenditures. When people visit a city to see the Games, aside from money spent on the

physical events, they spend money on accommodations, eating out at different restaurants, going

                                                                                                               **************  Heavy engineering includes highways and other infrastructure; special contractors include contractors hired for specific specialty building tasks.  

Page 27: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  27  

to see other events such as plays and local professional sports, and visit various visitor sites such

as museums and theme parks. All of this contributes to the study’s tourism GDP variable that

captures spending due to the Olympics, as well as spending that would otherwise occur. If the

Olympics increase the change in tourism GDP, as with any change in GDP, it will be visible

through normal growth pattern divergences.

Financial services GDP would include income from financial services such as depository

institutions and investment offices income. Financial service is used as a long-term economic

impact variable. In theory, hosting the Olympics leads to an increase in awareness and popularity,

and possible later on an increase in general growth because more people are moving to the host

city and more people are visiting the area. Financial services tries to capture the idea that more

people are moving into the area and possibly investing in the host city, thus the banks,

investment and holding offices will have a significant change in GDP over time aside from usual

growth trends.

4.3 Independent Variable Ideal Data Collection

For infrastructure, my predictor variables would ideally be city construction employment

rate per year, the number of construction firms in business per year, property values, residential

construction permits, and non-residential construction permits. These variables would help paint

a picture of the changes in the city’s construction industry GDP. Construction employment rate

and the number of construction firms in business per year both drive construction GDP; the more

companies in the business will likely lead to an increase in GDP because the industry is more

productive. Similarly, the more people employed in the industry, the more buildings and

structures they can produce. Higher property values may indicate a level of citywide growth and

desirability to locate businesses and homes there, implying more building starts and an increase

Page 28: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  28  

in GDP. A similar, but more specific measure than property value is residential and non-

residential permits, which show the number of new construction projects for individuals and

businesses. All of the independent construction variables should show increases or decreases in

construction GDP change.

There are a number of other predictors, which affect construction GDP, and need to be

held constant to isolate the effects of the chosen predictor variables. Because I am comparing

same-country sister cities, countrywide factors that would influence construction GDP, such as

Spain entering the EU, which would likely effect factors such as construction material prices and

trade openness, will be held constant. However, there still may be differences between cities that

need to be taken into consideration to make an accurate and unbiased comparison. If possible I

would hold steady dissimilarities in land and material prices. If they are significantly higher in

one city compared to the other there may be a decrease in building and infrastructure creation,

decreasing construction GDP. Control variables will hold constant outside effects that cause

difference in construction GDP among comparable states due to non-Olympic effects.

For prestige, my predictor variables would ideally be city level employment and number

of firms per year for the hospitality industry (hotels and accommodations), restaurants,

entertainment (amusement parks and tourist attractions) and recreation (sporting and

entertainment events), number of hotels and occupancy rate per year, and the number of flights

to the city as a final destination. These variables would help demonstrate the changes in the

city’s overall tourism GDP. Employment and number of tourist related firms in business per year

both raise tourism income because visitors need places to stay, restaurants to eat, and activities to

keep them interested. Generally, a city with more hotels, restaurants and attractions will have

something for all types of visitors. Since wealthier tourists spend more money, cities will likely

Page 29: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  29  

try to continuously upgrade or build new facilities, which will increase tourism expenditures.

Because hotels are only built when there is need—also reflected by a high occupancy rate—an

increase in the number of hotels and accommodations, should be reflected by an increasing

economic growth rate. The number of flights arriving in a city per day, assuming the city is a

destination city and not a connection city (for example a large percentage of international and

domestic flights connect through Atlanta everyday where Atlanta is not the final destination) is

another good indicator of visitors, because international and many out of state tourists will fly

into the city. All tourism independent variables aim to demonstrate the change in industry GDP

and hope to isolate any affects attributable to the Olympics.

To see if an Olympic effect exists for tourism, I need to control for other major factors

that will affect GDP. As previously mentioned, countrywide factors such as flight prices and ease

of travel for other EU member countries caused by Spain entering the EU will affect both

Catalonia and Madrid AC and thus will not influence tourism GDP. However, there still may be

differences between cities that affect tourism related GDP. If possible I would hold steady

differences in substitute destinations, number of major cities nearby. Substitute destination and

number of major cities nearby are closely related because depending on location, one city may

split the visitor market share because there are other similar popular locations nearby, while the

other city dominates their market share. Controlling for outside between states not attributable to

the Olympics will help isolate the potential Games related impact on tourism expenditures.

For the financial services industry, my predictor variables would ideally be city level

financial services employment, including depository and non-depository institutions, trading

firms, and international banks, the number of banks and similar firms in business by year, and

number and type of investments made per year. These variables would help show the changes in

Page 30: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  30  

a city’s financial industry income, which is used as a long-term indicator of hosting the Olympics.

It may be that more people or firms chose to move to/invest in the host city because of the

Olympics. An increase in banking firms and employment (depository and non-depository

institutions) reflects an overall demand from citizens and businesses to hold money. Trading

firms and hedge funds handle domestic and international investors, which shows both domestic

and international individual business economic growth, and may imply an increased interest in a

certain city or the overall state. International banks may show the international awareness of a

city or state. As with an increasing number of firms, an increase in employment also indicates

positive financial services and overall economic growth, and increased state awareness at the

national and international level. Similarly, the number of investments per year made by

individuals and businesses reveals an interest and commitment to economic growth. Financial

services independent variables should explain growth in financial services GDP.

To isolate an Olympic impact on financial services GDP, I will control for non-Olympic

related effects that impact financial services GDP. The biggest potential difference between

cities is economic growth. Economic growth drives all other industries such as tourism and

construction that use financial services; it is already held constant through same-country

comparisons.

Page 31: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  31  

4.4 Data Collection

Because data are not available at the city level, I will be using proxy data from

states/regions. For dependent variables in Spain I will use autonomous communities (Catalonia

for Barcelona and Madrid AC for Madrid).†††††††††††††† In the United States I will study statewide

dependent variables (Georgia for Atlanta and North Carolina for Charlotte). Similarly in

Australia, I will examine variable data among state/territories (New South Wales for Sydney and

Victoria for Melbourne).‡‡‡‡‡‡‡‡‡‡‡‡‡‡ Because the dependent variables are measured fairly

similarly and all aim to show industry income (GDP,§§§§§§§§§§§§§§ adjusted gross added

value*************** and adjusted factor income),††††††††††††††† the terms GDP and income will be

used to represent all different types of industry income.

4.4.1 Spain Variables

Below dependent and independent variables and sources are described for each Olympic Game.

In Spain all of the dependent variables for Catalonia and Madrid AC use 1990

euros‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡ as the base year. Data from 1988-1996 are inflation-adjusted to 1990 Euros for

best comparison.

                                                                                                               ††††††††††††††  In  size  similar  to  a  large  U.S.  state  ‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Also  similar  to  the  size  of  some  U.S.  states  §§§§§§§§§§§§§§  Value  added  in  production  by  labor  and  capital  within  a  state.  Bureau of Economic Analysis, “Gross Domestic Product (GDP) by State,” Bureau of Economic Analysis, http://bea.gov/regional/definitions/nextpage.cfm?key=Gross%20domestic%20product%20(GDP)%20by%20state (accessed March 9, 2012).  ***************  Output  minus  Intermediate  Consumption  at  purchaser  price  (accounts  for  over  90%  of  GDP). Eurostat, “Glossary: Gross Value Added at Market Prices,” Eurostat,http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Glossary:Value_added (accessed March 9, 2012).  †††††††††††††††  Equivalent  to  gross  domestic  product  less  taxes  plus  subsidies  on  production  and  imports. Australian Board of Statistics, “Glossary,” Australian Board of Statistics,http://www.abs.gov.au/Ausstats/[email protected]/glossary/5220.0 (accessed March 9, 2012).  ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  In  1990  the  Euro  did  not  exist.  For  methodological  changes  to  create  the  1990  ‘Euro’  values  see  Appendix  B.  

Page 32: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  32  

The proxy infrastructure variable for Spain is construction and civil engineering

works gross added value at market prices, hereby known as construction GAV.§§§§§§§§§§§§§§§

The prestige variable is a gross added value at market prices aggregate of

recuperation,**************** repair, trade, accommodation, restaurants, foodstuff, beverages, and

tobacco.

The financial variable is the gross added value at market prices of imputed production

of bank services,†††††††††††††††† hereby known as the financial GAV.

The construction independent variables for Catalonia and Madrid AC demonstrate

changes in construction GAV at the autonomous community level. The predictor variables are

new residential building permits, total gross investment‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡ and total net capital stock

for§§§§§§§§§§§§§§§§ construction, engineering, rental of buildings, and residential

capital.***************** As explained in the ideal data section, building permits can show growth in

construction income. Investment and capital stock investment imply a long-term commitment to

economic growth for individuals and businesses through equipment and new structure creation,

which similarly suggests positive growth in the amount of businesses and people coming to the

autonomous community for business and travel.

                                                                                                               §§§§§§§§§§§§§§§ See footnote 57 for explanation of gross added value at market price. **************** Instituto Nacional de Estadistica, “Spanish Regional Accounts,” Instituto Nacional de Estadistica,http://www.ine.es/jaxi/menu.do?type=pcaxis&path=/t35/p010/a1996&file=pcaxis (accessed March 9, 2012). †††††††††††††††† Interest receivable minus interest payments. Dennis Fixler, “Measuring the Services of Commercial Banks in the Nipas: Changes in Concepts and Methods,” OECD, www.oecd.org/dataoecd/10/13/24333573.doc. ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Public  and  private  un-­‐depreciated  total  invest  value  of  business  assets  in  this  industry.  Business Dictionary, s.v. “Gross Investment,” http://www.businessdictionary.com/definition/gross-investment.html (accessed March 12, 2012).  §§§§§§§§§§§§§§§§  Net  capital  stock  is  the  sum  of  the  written-­‐down  values  of  all  the  fixed  assets  still  in  use  is  described  as  the  net  capital  stock. OECD Stats, “Net Capital Stock,” OECD Stats, http://stats.oecd.org/glossary/detail.asp?ID=1752 (accessed March 9, 2012).    *****************  Instituto Nacional de Estadistica, “Spanish Regional Accounts,” Instituto Nacional de Estadistica,http://www.ine.es/jaxi/menu.do?type=pcaxis&path=/t35/p010/a1996&file=pcaxis.  

Page 33: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  33  

The tourism independent variables for Catalonia and Madrid AC explain growth

through change in autonomous community tourism GAV. Predictor variables are the number of

hotel establishments, as well as the total gross investment and total net capital stock for food,

beverages, hotels and restaurants. As seen in the ideal data section the number of hotel

establishments can show increases in tourism GDP change. Investment and capital stock

aggregated for tourism strongly imply long-term development and growth for major domestic

and international tourism industries in hotels, restaurant and related tourism sales which could be

caused by an increased in destination viability.

Financial independent variables for the autonomous communities of Catalonia and

Madrid AC will determine change in financial GAV. The predictor variables include number of

banks and credit institutions, and investment and capital stock for credit and insurance

institutions. As seen in the ideal data section the number of firms can impact growth in financial

income. Investment and capital stock for credit and insurance institutions indicates overall

economic stimulation and a long-term trend in growth and productivity for the autonomous

community.

4.4.2 United States Variables

In the United States all of the dependent variables for Georgia and North Carolina are

measured in real GDP by state, adjusted for inflation from 1992-2000 to 1997 US dollars.

The infrastructure variable for the United States is general contractors and operative

builders, heavy construction, except buildings, and special trade contractors.†††††††††††††††††

                                                                                                               †††††††††††††††††  Heavy  engineering  includes  highways  and  infrastructure  projects;  special  contractors  include  specific  building  projects  not  included  under  general  contractors.  Bureau of Economic Analysis, “Construction,” Bureau of Economic Analysis,http://bea.gov/regional/definitions/nextpage.cfm?key=Construction (accessed March 9, 2012).  

Page 34: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  34  

The prestige variable is a real GDP aggregate of air transportation, transit and ground

passenger transportation, arts, amusement and recreation (performing arts, spectator sports,

motion pictures, museums and related services), and accommodations.‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡

The financial variable is a real GDP aggregate of depository and non-depository

institutions, securities and commodities brokers, and holding and other investment

offices.§§§§§§§§§§§§§§§§§

The independent variables that indicate changes in construction industry GDP are

industry employment per year, the number of construction firms in business per year, and

residential building permits. Industry employment and number of firms includes general

contractors and operative builders, heavy construction, except buildings, and special trade

contractors.****************** As discussed in the ideal data section employment, number of firms,

and residential permits can indicate economic growth, and lead to an increase in construction

GDP.

The independent variables that predict changes in tourism related GDP are tourism

industry employment per year, the number of tourism related firms in business per year, and the

number of hotels and occupancy rates per year. Employment and number of firms are based upon

data aggregated from: eating and drinking places, hotels and other lodging places, motion

pictures, amusement and recreation services, museums, botanical and zoological

                                                                                                               ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Bureau of Economic Analysis, “Interactive Data,” Bureau of Economic Analysis, http://bea.gov/itable/ (accessed March 9, 2012).  §§§§§§§§§§§§§§§§§  Bureau of Economic Analysis, “Securities, Commodity Contracts, Investments,” Bureau of Economic Analysis,  http://bea.gov/regional/definitions/nextpage.cfm?key=Securities,%20commodity%20contracts,%20investments  (accessed  March  9,2012).  ******************  Bureau of Economic Analysis, “Interactive Data,” Bureau of Economic Analysis, http://bea.gov/itable/.  

Page 35: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  35  

gardens.†††††††††††††††††† As seen in the ideal data section employment, number of firms, number

of hotels and occupancy rates all can show growth or contraction in tourism GDP.

Independent variables that show changes in financial services GDP include financial

employment and the number of firms in business per year aggregated from depository and non-

depository institutions, and holding and other investment offices.‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡ As shown in the

ideal data collection, employment and number of firms could lead to an increase in that state’s

financial services GDP and reflects an overall economic growth among businesses and

individuals.

4.4.3 Australia Variables

In Australia, the dependent variables for New South Wales (NSW) and Victoria are

measured as total factor income at current prices from 1996-2004, with a base year of 1996.

The proxy infrastructure variable is construction total factor income at current

prices.§§§§§§§§§§§§§§§§§§

The proxy prestige variable is the total factor income at current prices of

accommodation and food services and arts and recreation services******************* by state.

The proxy financial variable is the total factor income at current prices of financial

and insurance services.†††††††††††††††††††

The independent construction variables for NSW and Victoria show the change in

industry income at the state/territory level through construction employment and dwelling unit

                                                                                                               ††††††††††††††††††  Ibid.  ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Ibid.  §§§§§§§§§§§§§§§§§§  Australian Board of Statistics, “Glossary,” Australian Board of Statistics,http://www.abs.gov.au/Ausstats/[email protected]/glossary/5220.0 (accessed March 9, 2012).  *******************  Ibid.  †††††††††††††††††††  Ibid.  

Page 36: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  36  

approvals, which are similar to residential construction permits. As noted in the ideal data

collection employment and construction permits indicate changes in construction income.

Independent tourism variables for NSW and Victoria show changes in industry

income at the state/territory level through an aggregate of accommodations and food services,

arts and recreation employment, number of hotels, and average length of stay at accommodations.

Employment statistics and the number of hotels, as indicated in ideal data collection, will show

changes in tourism GDP. The average length of stay at accommodation implies people are

spending more money in the state because they are staying longer, which leads to industry

growth.

The financial independent variables for Australia explain changes in financial

services income through financial and insurance employment. Increases in employment may

indicate an increase in businesses, possibly driven by an influx of people moving to the area due

to the Olympics, as reflected in financial services expenditures.

In Spain, the United States and Australia all independent variables are used to show

growth or contractions in dependent variables’ industry GDP. By comparing relative changes, it

is hoped to see if short and long-term economic effects of the Olympics, which might cause

changes in visitors and firms who move to the Olympic area, exist.

Page 37: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  37  

5 Model and Analysis

In this study, each of these Olympic Games has three types of linear regression models.

Regressions are run on each state for construction, tourism and financial models, and then

compared to the same-country, sister-state models. All regression equations, models and

diagnostic tests can be seen in Appendix C.

Each model has a number of potential issues from mathematical changes to model

assumptions. Mathematical changes that make data compatible for analysis can be found in

Appendix B. Because measurement adjustments made for the study or through country data

systems—such as the U.S. data measurement switching from the SIC to NAICS—some

regressions may exhibit influential outliers, which means the removal of that point will

significantly change and normalize results.‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡ Some of these changes, particularly

for Spain, may lead to autocorrelation—where data are dependent on time and space—which

will be overlooked because autocorrelation will not inhibit any effect the Olympics have on

change in industry GDP. Additionally, each model ignores the linear model assumption of

multicollinearity§§§§§§§§§§§§§§§§§§§ because while using highly correlated variables will not show

the importance of the individual variables, the study is looking for an overall change in GDP due

to the Olympics, which, if existent will still be visible with multicollinearity. Overall, necessary

mathematical compatibility changes might lead to autocorrelation and multicollinearity, which

will be disregarded based on study aims.

                                                                                                               ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  An  example:  If  the  data  1995  Barcelona  tourism  GDP  is  an  influential  outlier,  it  is  significantly  different  from  all  other  years  in  change  in  GDP,  meaning  it  is  inherently  different  from  the  other  data  points.  In  this  paper,  the  cases  appear  to  be  due  to  measurement  differences  and  changes.    §§§§§§§§§§§§§§§§§§§ Multicollinearity: two predictor variables are very highly correlated

Page 38: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  38  

Other model considerations include the type and number of predictor variables. While

every variable combination is run in each model, models with R-squared values above .95 were

not considered because real-world data with numbers that high indicate an identity, meaning the

predictors completely explain the outcome/ are the same as the outcome, which given the

variables tested, does not seem logically possible.******************** Additionally, since data are

being compared across models each pair needs to use the same model for true comparison.

Because of the R-squared values and model comparisons, many models will not include all

possible variables.

Finally, only countrywide effects and variables in the model will be held constant. Effects

such as Spain entering the EU, which should have a much larger effect on industries than the

Olympics, will be held constant because all states are within the same country. Similarly the

overall economic growth trend should be held constant. As the Arthur Andersen study

notes,†††††††††††††††††††† at the time of their Olympic Game the state of Georgia and Spain are in

cycles of economic downtown, which perpetuate throughout the country. If one state is

experiencing stunted or negative economic growth, it is generally accepted that the other states

within that country will experience a similar trend. Overall, all major differences between states

from economic growth to countrywide effects will be held constant.

                                                                                                               ******************** In an email message to the author on March 10,2012 Allen Sanderson noted models with .95 or above R-squared values can be considered an identity and variables used in this study will not explain all of the outcome variables. Allen Sanderson, email message to the author, March 10, 2012.    ††††††††††††††††††††  CREA/Arthur Andersen. “Economic Impact Study of the Sydney 2000 Olympic Games.”5-6.  

Page 39: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  39  

5.1 Model Implications Overview

The study’s Olympic model results suggest that the economic effects of the Olympics to

create positive economic growth may be a myth. Regressions run to see the economic impact of

the Olympics through change in industry GDP unsurprisingly varied for each set of Games; as

noted in the history section, while these Games are the most similar, at the same time each Game

takes place under unique circumstances. Furthermore, because each set of regressions for the

different Olympic Games do not have the same models, the models are not completely

comparable. Overall, despite model differences, it appears the 1992 Barcelona Games could have

lead to a positive change in construction GDP growth, but all other regressions and models for

the Barcelona, Atlanta, and Sydney Games showed no difference or negative growth when

compared to control state/region models.

Page 40: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  40  

5.2 1992 Barcelona Games Analysis

5.2.1 Spain Construction

Catalonia Madrid Spain Construction

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Year 406,197 .001 213,258 .031 Final Model

Residential Construction Permits

Adjusted R-squared Value .8722 .9284

Catalonia experiences a significantly higher positive change in construction income

compared to Madrid that may be attributable to the Olympics. Year, which tracks changes in

time, significantly predicts some of the change in construction GDP for both Catalonia and

Madrid. For Catalonia, with each year increase, GDP increases by 406,197 euro. Over the course

of the model, from 1988 to 1996 GDP will increase by 3,249,576 euro, which accounts for

60.15% of the change in Catalonia’s 1996 GDP. While the effect of individual years cannot be

examined due to model issues, overall, year accounts for a significant amount of the change in

GDP when adjusted for inflation. If attributable to the Olympics, the GDP change could be due

to Olympic building in pre-Olympic years, and increased awareness post-Olympics which may

lead to an influx of tourists and people moving to the area, and, necessarily, more construction.

Ideally, the model would allow a year-by-year breakdown, to look for specific trends, such as

those described above. While year is a significant variable in the model, it must also be

compared to Madrid to see if Madrid experiences a similar change, which would imply hosting

Page 41: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  41  

the Olympics does not have a significant impact on construction GDP, because both countries

experienced similar growth.

The corresponding model for Madrid also indicates that year has a significant impact on

construction GDP. For every year increase, construction GDP increases by 213,258 euro, which

over an eight-year change is 1,706,064 euro, or 32.88% of the change in Madrid’s 1996 GDP.

Year account for about 28% more of Catalonia’s amount of change in GDP compared to

Madrid’s change, which suggests that over the nine-year span Catalonia has more growth, which,

holding countrywide indicators constant, could be due to the Olympics. While a yearly

breakdown cannot be assessed, it seems Olympic facility construction, and possibly a rise in

firms and people moving to Catalonia as a result of the Olympics, could account for some of the

difference in growth.

5.2.2 Spain Tourism

Catalonia Madrid Spain Tourism

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Number of Hotel Rooms

205.78 .017 380.40 .000 Final Model

Total Gross Investment

Adjusted R-squared Value .5256 .9203

The number of hotels does not significantly impact Catalonia or Madrid change in

tourism GDP growth. The model for Catalonia and Madrid tourism shows that for both regions,

the number of hotel rooms explains a significant change in tourism GDP. Models including year

were considered an identity of tourism GDP, and will not be considered. When the number of

hotel rooms increases by 1%, GDP increases by 205.78 euro. Compared to overall GDP—at its

Page 42: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  42  

lowest in 1988 is over 4 million euro—it appears that even a 50% increase in number of rooms

(from 181,262 rooms to 271,893 rooms, which if one hotel is taken to be 250 rooms is an

increase in 1,087 hotels‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡) is only an increase of 10,294 euro, or far less than 1%

of tourism GDP. Because hotel rooms explain less than 1% of the amount of change in tourism

GDP, it appears that the Olympics did not have an effect on hotel and possibly tourism growth.

However, Madrid’s model will be examined to see if hotel rooms have a greater impact on the

change in Madrid’s tourism GDP.

The number of hotel rooms accounts for a significant change in Madrid’s tourism GDP.

For a 1% increase in number of hotels, tourism GDP increases by 380.4 euro. With a 50%

increase in 1988 hotels rooms from 33,677 to 50,516 rooms, or the equivalent of 202 new

hotels,§§§§§§§§§§§§§§§§§§§§ hotel rooms account for less than 1% of the amount of change in tourism

GDP. Similar to Catalonia, Madrid did not show that hotels account for some of the change in

GDP growth, which suggests that the Olympics did not have a visible effect on the change in

tourism GDP growth; however, the number of hotels does not account for other major aspects of

the tourism industry and results must be interpreted with caution.

                                                                                                               ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Hotel  refers  to  a  general  place  of  accommodation  that  includes  bed  and  breakfasts,  motels,  hotels,  and  other  places  of  accommodations.  §§§§§§§§§§§§§§§§§§§§  Using  250  rooms=1  hotel  

Page 43: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  43  

5.2.3 Spain Financial Services

Catalonia Madrid Spain Financial Services

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Year 230,394 .001 641,811 .022 Final Model

Number of Banking Institutions

659 .024

Adjusted R-squared Value .8351 .7498

The financial services model indicates that Madrid experiences 51% greater change in

growth in nine years that Catalonia, implying the Olympics could have hindered Catalonia’s

industry growth. Both year and number of banking institutes are significant variables for

Catalonia. For each year increase, GDP increases by 230,294 euro. Over the nine-year period

Catalonia GDP increases by 1,842,352 euro, which account for 41% of the change in financial

services GDP. Additionally, a 1% increase in the number of banking institutes (69 banking

institutions for 1988) increases GDP by 659 euro. Over the nine-year period studied, the number

of banks increases by about 1000 at its peak, but then declines over time. Given the banking

trends, it seems highly unlikely to have a more than a 10% increase in number of banks over a

nine-year period (676 banks, and only a 6590 euro increase), which suggests that the effect of

banking institutions is very small and can be discounted. Over time, more people and firms may

come to Catalonia because of an increase of popularity, international awareness, or some other

reason, which then leads to a heightened demand for banking services. However, the results must

be compared to Madrid to see the model in a relative context.

Madrid’s model shows year causes a significant amount of change in financial GDP. A

year increase leads to an increase in GDP by 641,811 euro. By the end of the nine-year period,

Page 44: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  44  

the yearly change accounts for 92% of the change in financial services GDP. This is 51% more

than in Catalonia, which implies that Madrid had a much larger growth rate over this period

while not hosting the Olympics. Madrid’s much larger growth rate compared to Catalonia as

seen by the impact of year on the change in financial services GDP might infer that the Olympics

negatively impacted Catalonia’s financial services growth rate. At the same time, it seems likely

that the Olympics causing people/firms to move into an area and invest in local institutes may be

small and therefore difficult to see on a regional level, thus differences may be due to factors not

considered in the scope of the study.

5.2.4 Spain Conclusions

The Barcelona Olympics seems to have a positive influence the construction industry, no

impact on tourism, and a negative effect on financial services for Catalonia. Change in time

accounted for about 38% more of the increase in Catalonia’s construction GDP growth than

Madrid’s. Tourism models show no difference between Catalonia and Madrid, which indicates

the Olympics do not significantly impact the number of hotels, and through this a lack of tourism

demand growth. Finally, financial services implied that Madrid experience greater growth

without hosting the Olympics than Catalonia.

Page 45: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  45  

5.3 1996 Atlanta Game Analysis

5.3.1 U.S. Construction

Based upon the narrow variable, residential permits, the U.S. construction models show

negligible significance on the change in construction GDP. Thus, non-residential building is not

measured, which is vital to the Olympics and would likely explain more change in industry

income. As seen by the small coefficient, a 1% increase in number of residential construction

permits (460 new permits for 1996) leads to a 179 US dollar increase in GDP. A 100% increase

in the number of construction permits, which seems unrealistic for a 9-year period (there was

about a 50% increase from 1992-2000), still only leads to a 17,900 US dollar increase. This

increase accounts for far less than 1% of change in construction GDP over the studied timeframe.

It appears that based on model constraints and the final model, any Olympic effect from the

change in construction income does not exist, or is not visible. The North Carolina model shows

similar results; a 1% increase in number of residential construction permits (499 new permits)

produces an increase in GDP by 176 US dollars. Even with a 100% increase in permits, there

will be no visible effect on the change in construction GDP. Given that residential building

permits only capture the long-term effect caused by people moving to an area overtime, it does

not capture a potential Olympic effect from before the Olympics, which might be seen through

U.S. Construction Georgia North Carolina

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Final Model Number of Residential Permits

179.12 .000 176.36 .030

Adjusted R-squared Value .9010 .8954

Page 46: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  46  

an increasing number of firms (considered an identity), or nonresidential construction

permits.********************* While no difference can be seen between Georgia and North Carolina

in the long run, there may be better models to judge whether there is a short-term Olympic effect.

5.3.2 U.S. Tourism

The number of tourism related firms leads to a insignificant change in industry GDP

growth for both Georgia and North Carolina. Most models run for Georgia and North Carolina

either have R-squared values that imply an identity, or when variables are significant, they

account for far less than 1% of changes in tourism GDP. While year appeared to have a large

effect for both Georgia and North Carolina, every model with year appeared to be an identity.

However, a number of variable changes logically occur over time, especially number of hotel

rooms, employees, and number of firms (from hotels to amusement industries). Since the study is

looking at changes in tourism GDP, the number of firms should provide the most sensical,

concrete change, because an increase in hotel rooms eventually leads to new hotels, but that does

not account for other tourism industries, and increases in employees will lead to new firms, yet

the number of employees to create a new firm vary. Given R-squared constraints and the levels

of data, the model predictor is number of firms. For a 1% increase in the number of firms (205

new firms), there is a corresponding GDP increase of 380.5 US dollars. A 20% increase in the

                                                                                                                 

U.S. Tourism Georgia North Carolina

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Final Model Number of Firms

380.5 .001 191.5 .030

Adjusted R-squared Value .7850 .4412

Page 47: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  47  

number of firms (in 1988 4,111 firms) would lead to a GDP increase of 456.6 US dollars. If the

number of firms increases by 100%, the GDP increase would still be less than 1% of the change

in tourism GDP.

Georgia’s modeling results are similarly mirrored by North Carolina, where the number

of hotel firms also represents less than 1% of the change in tourism GDP, showing no difference

between the model. Based on number of tourism businesses the Olympics did not have any

significant impact Georgia’s tourism expenditures; at the same time North Carolina did not

experience significant growth by not hosting.

5.3.3 U.S. Financial Services

North Carolina experiences a significantly larger change in financial services growth over

time than Georgia. Each year increase leads to a 506,232 US dollar increase in Georgia’s

financial GDP. Over the 9-year timeframe there is a 4,049,856 US dollar increase in financial

GDP, accounting for 37.2% of the growth in Georgia’s financial GDP. Over time, people may

choose to invest or move into Georgia as a result of the Atlanta Olympic Games increasing

international presence and awareness.

For each year change in North Carolina there is a 2,140,296 US dollar increase in GDP.

For the study timeframe, there is a 17,122,368 US dollar increase in GDP, which explains 91.5%

U.S. Financial Services Georgia North Carolina

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Year 506,232 .000 2,140,296 .000 Final Model

Number of employees

-295.0 .027

Adjusted R-squared Value .9424 .9158

Page 48: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  48  

of the change in financial GDP growth for North Carolina. This is a 54.3% greater increase in

financial GDP growth than in Georgia. While the number of employees in financial services

appears significant for North Carolina it is unimportant. For a 1% increase in employees (765

new employees, which is 7 new firms)††††††††††††††††††††† there is a decrease in GDP by 295

dollars. With a 100% increase in employees (76,451 employees or 765 new firms) there is still

only a decrease in GDP by 29,500 US dollars, or about 1/10th of a percent decrease in change in

GDP growth. Because the number of employees has such a small impact on change in GDP

growth it will be considered insignificant. Overall, compared to North Carolina, it appears the

Olympics may have stunted Georgia’s long-term financial services growth.

5.3.4 U.S. Conclusions

Overall, the Atlanta Olympics did not appear to create a positive change in economic

growth for construction, tourism or financial services. There are a number of model issues and

identity variables, which are excluded from models. The lack of differences in construction GDP

change due to residential construction permits indicates the Atlanta Games did not bring more

firms and people to the area in the long-run, and thus did not positively impact on the

construction industry. Similarly, there is not a significant difference in tourism GDP change

based on the number of firms, showing the Games did not attract a significant number of new

tourism businesses; finally, financial services expenditures appears to be negatively impacted by

the Olympics.

                                                                                                               †††††††††††††††††††††  When  100  employees  equal  1  new  firm.  

Page 49: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  49  

5.4 2000 Sydney Olympic Game

5.4.1 Australia Construction

Time series changes in construction income show Victoria has more economic industry

growth than NSW. For each year increase from 1996 to 2000 there is a 515,927 Australian dollar

increase in GDP. Over the course of the 9-year period there is a total increase in GDP of

4,127,416 Australian dollars, which accounts for 23.6% of the change in construction GDP

growth. Because there is not a year-by-year model available, pre-and post-Olympic trends cannot

be quantified, although it seems construction would heavily increase in the years leading up to

the Olympic Games because of facilities and infrastructure projects, and might drop back to

normal rates after the Games.

Victoria’s construction expenditures indicate a larger change in growth than NSW’s. For

each year increase, construction GDP increases by 522,186 Australian dollars. Over the course of

the study there is a 4,177,488 Australian dollar increase in construction GDP due to year, which

explains 37.2% of the change in Victoria’s construction GDP. A change in year over time

predicts 13.6% more of the growth in Victoria’s construction GDP, when compared to NSW,

which suggest Victoria has a larger growth without hosting the Olympics, implying that hosting

the Olympics may have had a negative impact on Sydney construction growth.

Australia Construction NSW Victoria

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Year 515,927 .016 522,186 .012 Final Model

Number of residential construction permits

Adjusted R-squared Value .5651 .7702

Page 50: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  50  

5.4.2 Australia Tourism

Sydney experiences 2.9% less change in growth in tourism GDP than Victoria indicating

the Olympics may have deterred visitors or somehow hindered growth over time. In NSW, every

year increase leads to an increase in GDP by 245,754 Australian dollars, which over the 9-year

time period leads to a GDP increase of 1,966,032 Australian dollars. The GDP change over the

9-year period accounts for 6.7% of the growth in industry income.

However, Victoria experiences greater change in growth by 2.9% over the course of nine-

years, indicating a small difference between the regions. A one-year change leads to an increase

in GDP by 202,559 dollars, which is 1,620,472 Australian dollars, or 9.6% of the change in GDP

growth. This indicates that the Olympic year and after the Olympics may have actually deterred

visitors from visiting Sydney, had some other negative impact on visitor rates, or Victoria was

able to better capitalize on tourism over time.

Australia Tourism NSW Victoria

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Year 245,754 .007 202,559 .035 Final Model

Average Stay

Adjusted R-squared Value .7810 .8812

Page 51: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  51  

5.4.3 Australia Financial Services

The Australian financial model for NSW and Victoria only has two predictors available,

all of which appear to be identities of financial GDP, which makes interpreting the values

unnecessary because of real world constraints that indicate other variables better truly predict the

change in financial services GDP. The financial services model is already difficult to evaluate for

an Olympic effect because it is solely a long-term effect, which combined with the inability of

the model to differentiate between year-by-year trends makes it difficult to see whether pre-and

post-Olympic tendencies differ. If the Olympics lead to a change in growth in financial services

GDP it seems likely that it is because of an increase in the number of people moving and

investing in an area, which leads to more banking institutions and employees. If year is a

significant variable, for NSW, a year increase leads to a 1,399,820 Australian dollar increase in

financial GDP over the 9-year period, which account for 37.7% of the change in NSW’s

financial GDP growth. Victoria shows a slightly higher amount of overall growth; while a year

increase leads to an 845,870 Australian dollar increase in GDP, or 40.0% over the 9-year study.

Overall, modeling identity issues, variable selection and the aggregate year variable may cover

any potential significant differences between the two states.

Australia Financial NSW Victoria

Coefficient P-Value (.05)

Coefficient P-Value (.05)

Year 1,399,820 .000 845,870 .000 Final Model

Number of Employees

Adjusted R-squared Value .9573 .9622

Page 52: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  52  

5.4.4 Australia Conclusions

While there are some model identity issues, overall, the models indicate that Victoria

experiences significantly more growth than NSW for construction and tourism industries, which

implies the Olympics do not lead to positive economic change in those industries.

6 Summary and Conclusions

Overall the regression models run may show a narrow scope of results. As mentioned in

the model and analysis section, there are a number of mathematical and model changes that

could be problematic, and interpretation must proceed with caution. Because there could be some

amount of inaccuracy, it is not possible to be completely certain that the 1992, 1996 and 2000

Olympic Games lead to almost no positive economic growth, and in some cases hinder growth.

Only the 1992 Barcelona Olympic Games model indicates a possible positive economic

effect due to the Olympics through change in construction GDP. All other models show that the

Olympics does not have a significantly positive impact on the change in industry growth when

compared to the control model, and in some cases, the control model accounts for greater change

in growth than the Olympic model. However, it must be noted that only three industries were

studied: construction, tourism and financial services, which try to represent the idea of

infrastructure, prestige and financial growth that many cities vie for when submitting bids to host

the Olympics, thus non-economic reasons are ignored.

Page 53: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  53  

7 Policy Implications

Policy implications may be limited because of the somewhat narrow scope of the project,

combined with potential error due to mathematical changes and variable issues discussed in the

model analysis section. Additionally, because the variables across Olympic models differ, all of

the Games are not completely comparable. With these issues in mind, all interpretations must be

examined with extreme caution. On the whole, aside from the 1992 Barcelona Games

construction, it appears the Olympics did not have a positive impact on change in GDP for

construction, tourism and financial services for states/regions where the Olympics took place

compared to states/regions where the Olympics did not take place for the 1992, 1996 and 2000

Olympic Games.

Aside from the Barcelona Games possibly positively impacting the economic change in

construction growth for Catalonia, if there is no overall difference in change in construction,

tourism and financial services growth between states, it seems possible that no change exists

between cities, and the Olympics have a very small impact. The Barcelona, Atlanta and Sydney

Games were chosen for this study because all of the Games hoped to use the Olympics to

achieve economic growth and development, which means if the Olympics did not lead to that

growth (at least through three major industries studied that seem to target economic

development) the host cities should have chosen to construct facilities most needed that might

lead to substantial economic growth, and pursue other ways to attract more tourist and people to

the area such as renovating historic areas, and building new facilities such as attractions and

schools. Since tourism and financial services are long-term growth indicators, investing in

projects other than the Olympics may bring more people to visit and move into the area. While

the variables have flaws as described above, all of the financial services models and the Sydney

Page 54: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  54  

tourism model, which could be considered the most complete tourism model, showed that the

control state/region leads to greater change in industry GDP growth. This seems to confirm that

if a city, and its state/region are hoping for economic growth and development, pursuing options

other than the Olympics may lead to more growth and development.

Overall, while the analysis results provides a red flag to cities considering bidding for the

Olympics, it does not capture non-economic reasons for hosting the Olympics, and aims to open

the eyes of potential bidders. From the analysis results, it seems constituents, developers, hotels

and those who stand to make a profit (according to economic impact studies) should not take

study numbers at face value that promise vast amounts of extra profit garnered by hosting the

Games. The economic impact studies are misleading; while there could be profit for developers,

restaurants and others involved, the studies greatly amplify the possibilities, as seen through the

relative comparison between host cities and control cities. The purpose of this analysis is

increase constituents’ awareness to the economic impact and development that can or cannot be

expected as a result of hosting the Olympics or other mega-events, such as the World Cup or

Super Bowl. Additionally, the analysis does not account for non-economic reasons to host the

Olympics, such as pride, showing off the city and all it has to offer, and personal utility, such as

the fun of being in an Olympic city and watching the Games. In the end, the results raise

concerns about the economic benefit promised by impact studies to those with a vested monetary

interest, and hopes to help cities consider their reasons for wanting to host the Olympics and be

open with all constituents.

Page 55: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  55  

References

Australian Board of Statistics. “2000 Year Book Australia.” Australian Board of Statistics.http://www.ausstats.abs.gov.au/Ausstats/subscriber.nsf/0/CA25687100069892CA256889000E7506/$File/13010_2000.pdf(accessed March 9, 2012). Australian Board of Statistics. “Glossary.” Australian Board of Statistics.http://www.abs.gov.au/Ausstats/[email protected]/glossary/5220.0 (accessed March 9, 2012). Australian Board of Statistics. “National Regional Profile: Melbourne (statistical Division).” Australian Board of Statistics.http://www.abs.gov.au/ausstats/[email protected]/lookup/205Main+Features12006-2010(accessed March 9, 2012). Australian Board of Statistics, “National Regional Profile: Past and Future Releases,” Australian Board of Statistics, http://www.abs.gov.au/AUSSTATS/[email protected]/Lookup/Past+Future+Issues2006-2010?OpenDocument&tabname=Past%20Future%20Issues&prodno=LGA17200&issue=2006-2010&num=&view=&(accessed March 9, 2012). Australian Board of Statistics. “National Regional Profile: Sydney (statistical Division).” Australian Board of Statistics. http://www.abs.gov.au/ausstats/[email protected]/lookup/105Main+Features12006-2010 (accessed March 9, 2012). Billings, Stephen B. and J. Scott Holladay. "Should Cities Go for the Gold? The Long-Term Impacts of Hosting the Olympics.” Economic inquiry (2011): 1-278. Print. Bureau of Economic Analysis, “Construction,” Bureau of Economic Analysis,http://bea.gov/regional/definitions/nextpage.cfm?key=Construction (accessed March 9, 2012). Bureau of Economic Analysis, “Gross Domestic Product (GDP) by State,” Bureau of Economic Analysis, http://bea.gov/regional/definitions/nextpage.cfm?key=Gross%20domestic%20product%20(GDP)%20by%20state (accessed March 9, 2012). Bureau of Economic Analysis, “Interactive Data,” Bureau of Economic Analysis. http://bea.gov/itable/ (accessed March 9,2012). Bureau of Economic Analysis, “Securities, Commodity Contracts, Investments,” Bureau of Economic Analysis, http://bea.gov/regional/definitions/nextpage.cfm?key=Securities,%20commodity%20contracts,%20investments (accessed March 9,2012).

Page 56: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  56  

CREA/Arthur Andersen. “Economic Impact Study of the Sydney 2000 Olympic Games.” Sydney: Arthur Andersen (1999): 1-7.Print. Eurostat. “Glossary: Gross Value Added at Market Prices.” Eurostat.http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Glossary:Value_added(accessed March 9, 2012). Fixler, Dennis. “Measuring the Services of Commercial Banks in the Nipas: Changes in Concepts and Methods.” OECD. www.oecd.org/dataoecd/10/13/24333573.doc. Humpreys, Jeffrey M. and Michael K. Plummer, “The Economic Impact On the State of Georgia of Hosting the 1996 Summer Olympics” (diss., The University of Georgia, 1995), 1-133, http://www.terry.uga.edu/selig/docs/olympics.pdf (accessed March 8, 2012). Giesceke, James A. and John R. Madden. "Modelling the Economic Impacts of the Sydney Olympics in Retrospect ? Game Over for the Bonanza Story?*.” Economic Papers: A journal of applied economics and policy 30, No. 2 (2011): 1-230. Instituto Nacional de Estadistica. “Spanish Regional Accounts.” Instituto Nacional de Estadistica.http://www.ine.es/jaxi/menu.do?type=pcaxis&path=/t35/p010/a1996&file=pcaxis (accessed March 9, 2012). Kasimati, Evangelia. “Economic Aspects and the Summer Olympics: A Review of Related Research.” International Journal of Tourism Research 5 (2003): 433-44. http://people.bath.ac.uk/ecpek/index_files/Paper1.pdf. Matheson, Victor. “Mega-events: The Effect of the World’s Biggest Sporting Events On Local, Regional, and National Economies.” College of Holy Cross Faculty Research Series(October 2006): 1-31. http://casgroup.fiu.edu/pages/docs/2744/1277904942_matheson_events.pdf Matheson, Victor. “Upon Further Review: An Examination of Sporting Event Economic Impact Studies.” Sports Journal 5, no.1 (2002): 1-5. OECD Stats. “Net Capital Stock.” OECD Stats. http://stats.oecd.org/glossary/detail.asp?ID=1752(accessed March 9, 2012). Passa, Dennis. A Year Later Sydney Areas Go Unused,.Free Lance Star (Sydney). September 16, 2001. Rose, Andrew K., and Mark M. Spiegel. "The Olympic Effect*." The Economic Journal 121.553 (2011): 664-665. Sanderson, Allen. “Olympic Comparisons.” Lecture, University of Chicago, Chicago, IL, November 12, 2011. Sanderson, Allen. “Olympics Lecture.” Lecture, University of Chicago, Chicago, IL, January 6, 2012.

Page 57: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  57  

Szymanski, Stefan. Playbooks And Checkbooks: An Introduction To The Economics Of Modern Sports. Princeton : Princeton University Press, 2009: 160. US Government Bureau. “Georgia Demographic Profile 2000.” Censtats.http://censtats.census.gov/data/GA/04013.pdf (accessed March 9, 2012).

Page 58: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  58  

Appendix A: Comparison Explanations

A.1  Spain    Catalonia-­‐Madrid  Comparison:  Of  the  autonomous  communities  in  Spain  the  most  comparable  to  Catalonia  is  Madrid.  From  1988-­‐1996,  according  to  the  Instituto  Nacional  de  Stadistica,  Catalonia  and  Madrid  had  the  two  largest  GDP  by  autonomous  communities  in  Spain.‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Barcelona  and  Madrid  are  the  biggest  cities  with  their  autonomous  communities,  as  well  as  the  largest  and  most  traversed  nationally  and  internationally.  While  Barcelona  is  30%  of  Catalonia’s  GDP  and  Madrid  is  65%  of  Madrid  AC’s  economy,  both  are  seen  as  premier  tourist  and  business  cities.§§§§§§§§§§§§§§§§§§§§§    A.2  United  States    Georgia-­‐North  Carolina  Comparison:  Only  southern  states  were  considered  for  comparison  given  their  city  characteristics  are  inherently  different  from  northern  cities  and  states.**********************  According  to  the  Bureau  of  Economic  Analysis,  in  the  earliest  year  measured,2002,  Atlanta  and  Charlotte  were  the  largest  part  of  the  state’s  economy.  Atlanta  makes  up  65%  of  Georgia’s  economy,††††††††††††††††††††††  with  Savannah,  the  second  largest  city,  accounting  for  4%  of  state  GDP.‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Charlotte  makes  up  28%  of  North  Carolina’s  economy,§§§§§§§§§§§§§§§§§§§§§§  and  although  it  is  much  smaller  than  Atlanta,  the  next  largest  city  area  (Durham-­‐Chapel  Hill)  account  for  8%  of  the  economy.***********************  Compared  to  other  city/state  ratios,  this  is  the  closest  match.  Furthermore,  their  growth  patterns  are  similar,  from  1990-­‐2000  Georgia’s  population  grew  by  26.4%,†††††††††††††††††††††††  similar  to  North  Carolina’s  21.6%  growth  rate.‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Additionally,  the  states  have  similar  industry  breakdowns  and  growth,  with  focuses  on  manufacturing  and  education,  health  and  social  services  that  make  them  most  comparable.      A.3  Australia    

                                                                                                               ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Instituto Nacional de Estadistica, “Spanish Regional Accounts,” Instituto Nacional de Estadistica,http://www.ine.es/jaxi/menu.do?type=pcaxis&path=/t35/p010/a1996&file=pcaxis.  §§§§§§§§§§§§§§§§§§§§§  Ibid.  **********************  Allen Sanderson, “Olympic Comparisons” (lecture, University of Chicago, Chicago, IL, November 12, 2011). ††††††††††††††††††††††  Bureau of Economic Analysis, “Interactive Data,” Bureau of Economic Analysis, http://bea.gov/itable/.  ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  Ibid.  §§§§§§§§§§§§§§§§§§§§§§  Ibid.  ***********************  Ibid.  †††††††††††††††††††††††  US Government Bureau, “Georgia Demographic Profile 2000,” Censtats,http://censtats.census.gov/data/GA/04013.pdf (accessed March 9, 2012).  ‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡  US Government Bureau, “North Carolina Demographic Profile 2000,” Censtats,  http://censtats.census.gov/data/NC/04037.pdf   (accessed March 9, 2012).  

Page 59: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  59  

New  South  Wales-­‐Victoria  Comparison:  Within  the  states,  Sydney  and  Melbourne  are  the  two  largest  cities  in  the  country.§§§§§§§§§§§§§§§§§§§§§§§    The  state  populations  are  the  two  largest  among  all  Australian  states,  they  have  the  most  air  traffic,  comparable  construction,  manufacturing  and  other  industries.************************    

                                                                                                               §§§§§§§§§§§§§§§§§§§§§§§  Australian Board of Statistics, “National Regional Profile: Melbourne (statistical Division),” Australian Board of Statistics, http://www.abs.gov.au/ausstats/[email protected]/lookup/205Main+Features12006-2010 (accessed March 9, 2012).  Australian Board of Statistics, “National Regional Profile: Sydney (statistical Division),” Australian Board of Statistics, http://www.abs.gov.au/ausstats/[email protected]/lookup/105Main+Features12006-­‐2010 (accessed March 9, 2012).    ************************  Australian Board of Statistics, “2000 Year Book Australia,” Australian Board of Statistics,http://www.ausstats.abs.gov.au/Ausstats/subscriber.nsf/0/CA25687100069892CA256889000E7506/$File/13010_2000.pdf (accessed March 9, 2012).    

Page 60: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  60  

Appendix B: Methodological Changes

B.1 Spain -Dependent Variables: Originally measured in 1986 pesetas. Convert to 1986 USD. 142.2 pesetas=1 US dollar. (Exchange rates based on historic Federal Reserve Bureau numbers tallied by day, and averaged by months for annual rate). 1986 dollar to 1995 dollar. Multiply by 1986 dollar by 1.38. 1995 dollar to 1995 peseta. Multiply by 124.66. (See above exchange rate note) Given that 1995 and 1996 euro (base 1995) were available from the INE’s website, these numbers were used to determine 1988-1994 numbers for each dependent variable. 1995 pesetas and 1996 were divided by 1995 euro and 1996 euro, then vice-versa. The euro/peseta equation was averaged among the two years, then applied to all peseta years to create 1995 base euro for 1988-1996. To check these values, the peseta/euro equation average from above was applied to the new euro base to check for accuracy in equations. However, because one average was applied to each year, which may shift on a year-to-year basis compared to the average, there will likely be autocorrelation where data shows time/space dependent patterns. -1996 dependent tourism numbers did not exist. I averaged changes between 1993,1994,and 1995 numbers, which had a similar trend, then applied that to the 1995 value to create the 1996 value. -’95-’96 Barcelona hotel rooms drastically change, due an unclear change in measurement classification system. I took 35,000 off the 1995,1996 numbers each because the overall trend/average trend is otherwise a small increase each time, if not by the same percentage change difference by year, and decreasing hotel rooms by 35,000 followed the original trend. -’91-’92 Barcelona number of financial institutes. 1988-1990 trends averaged and applied to the 1990 number to create the 1991 number. 1993-1996 trends averaged and applied to the 1993 number to create the 1992 number. - ’90-’92 Madrid house construction orders: I took the average between other years to look for trends, and found a decreasing trend. I then took the average and applied it to the 1993 number to create 1992, and so on to create the 1991 and 1990 value. -’92-’94 Madrid credit institutions: I took the average between each year with a more than 1 bank change, applied 1991 number times average, 1992 number times average and the 1993 number times average. (average=1.043) B.2 United States -Note: Employment (CBP) and dependent variable data (Bureau of Economic Analysis) changes in 1997 from SIC (Standard Industrial Classification) methodology to NAICS (North American Industry Classification System)

Page 61: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  61  

-Data was adjusted… “Many individual NAICS industries correspond directly to industries as defined under the SIC system, most of the aggregate NAICS grouping do not.” Employment data is similar enough for no changes. -For category differences in SIC and NAICS numbers I took the category called Finance, Insurance and Real Estate compared to Finance and Insurance. I took away the subset Insurance and Real Estate sections so the variable would be only financial data, and I subtracted the Insurance subset from the NAICS numbers to make them comparable. -Dependent variables: Each 1997 dependent variable was measured by both the SIC and NAICS. Comparing those numbers showed differences with all data adjusted to 1997. Using the SIC as base numbers, I applied the difference between the SIC and NAICS 1997 numbers, to the 1998,1999 and 2000 NAICS numbers which puts them in SIC form. (This change was applied after all numbers were adjusted for a constant year 1997). B.3 Australia -Dependent Variables: Adjusted to 2000 values

Page 62: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  62  

Appendix C: Regressions C.1 Spain

 

 

Page 63: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  63  

 

 

 

Page 64: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  64  

 

 

Page 65: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  65  

 

 

 

Page 66: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  66  

 

 

Page 67: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  67  

   

 

 

Page 68: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  68  

 

 

Page 69: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  69  

 

 

 

Page 70: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  70  

 

 

Page 71: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  71  

 

 

     

Page 72: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  72  

 

 

Page 73: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  73  

 

 

    C.2 United States

Page 74: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  74  

 

 

Page 75: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  75  

 

 

 

Page 76: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  76  

 

   

Page 77: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  77  

 

 

 

Page 78: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  78  

 

 

Page 79: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  79  

 

 

 

Page 80: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  80  

 

 

Page 81: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  81  

 

 

         

Page 82: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  82  

 

         

Page 83: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  83  

 

 

 

Page 84: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  84  

 

           

Page 85: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  85  

       

 

 

  C.3 Australia

Page 86: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  86  

 

 

Page 87: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  87  

 

 

 

Page 88: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  88  

 

 

Page 89: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  89  

 

 

 

Page 90: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  90  

 

 

Page 91: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  91  

 

 

 

Page 92: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  92  

 

 

Page 93: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  93  

 

 

 

Page 94: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  94  

 

 

Page 95: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  95  

 

 

 

Page 96: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  96  

 

 

Page 97: Economic Impacts of the Olympic Games through …...! 1! Economic Impacts of the Olympic Games through State Comparison Samantha Edds* April 2012 Abstract After the financially profitable

       

  97