Estimation of the Value of Public Goods Generated by Improved Sport Stadiums and Arenas Using the Contingent Valuation Method

Transcription

1 Estimation of the Value of Public Goods Generated by Improved Sport Stadiums and Arenas Using the Contingent Valuation Method Bruce K. Johnson James Graham Brown Associate Professor of Economics Centre College Danville, KY and John C. Whitehead Associate Professor Department of Economics East Carolina University Greenville, NC August 1998 An earlier version of this paper was presented at the annual meeting of the Western Economic Association in Lake Tahoe, Nevada, June 28-July 2, The authors thank Anthony C. Krautmann for his comments and suggestions.

2 Introduction The 1990s have seen a surge in sports stadium and arena building. In mid 1996, the total spending completed and planned for the decade topped $9 billion, with more than 80 percent of the funding provided by state and local governments (USA Today, 1996). Even more new sports building projects have been approved since 1996 and others have been proposed but not yet funded. For instance, Raleigh, NC agreed to build a new arena to attract a hockey team and both the Yankees and the Mets have asked New York City to build new stadiums for them. The building boom continues a trend begun in the 1950s. Most major league sports teams played in privately owned buildings in 1950, including 15 of the 16 baseball teams. By 1989, however, 77 of the 100 teams in major league baseball, football, basketball, and hockey played in publicly owned buildings (Quirk and Fort, ). By building new stadiums, governments hope to prevent existing teams from moving to another city or to attract an expansion team or an existing team to the area. Supporters of public funding for stadiums and arenas usually argue that the buildings will pay for themselves through their direct and indirect benefits. The direct benefits include the income created by the team s production of private goods. They sell tickets, concessions, souvenirs, and other goods and services. These private goods are exclusive and rivalrous. A person cannot attend the game without buying a ticket, and every ticket bought reduces the number of tickets available to others. The indirect benefits come in two types, tangible and intangible. The tangible, indirect benefits derive from the multiplier effect as the direct income generated by sale of the private goods is spent and respent in the metropolitan economy. The income

3 enjoyed by restaurants, bars, and other businesses as a result of the stadium activities is a positive externality not able to be appropriated by privately owned teams or stadiums. The intangible, indirect benefits derive from the production of nonrivalrous and nonexcludable public goods, what some might call fan loyalty and civic pride. They can be enjoyed by anyone, even those who never purchase tickets or attend games. People can listen to games on the radio, watch them on television, read about them in the sports pages, and talk about them with friends. Pennant races and championships can knit an entire city or region together, perhaps more effectively than any other institution. Most objective, scientific studies of publicly financed stadiums have been unable to support the claims that stadiums pay for themselves. Neither the direct nor indirect tangible benefits justify public subsidies of sports buildings. Quirk and Fort (1992), for instance, show that publicly funded sports buildings invariably fail to generate sufficient direct benefits to cover their full opportunity costs. Some fail even to cover their variable costs. Baade and Dye (1990) and Noll and Zimbalist (1997), among others, demonstrate that stadiums and teams fail to stimulate much growth in a region s economy. The public good of civic pride may be the reason so many state and local governments choose to subsidize stadiums despite the overwhelming evidence that the total costs exceed direct and indirect tangible benefits. No study to date has measured the value of benefits generated by the public goods produced by big-time sports. Noll and Zimbalist (1997) and Fort (1997) have suggested the value of these intangible, indirect benefits may well exceed the value of the tangible benefits produced. In this paper we apply the contingent valuation method (CVM, Mitchell and Carson, 1989), widely used to measure the value of environmental public goods, to two

4 publicly funded sports projects in Lexington, Kentucky. i We determine the extent to which people consume the public goods generated by sports, whether the public goods are valued by those who do not attend games, and most importantly, to determine the value of public goods generated by sports teams. We first describe the Lexington proposals. Next we describe the survey instrument, the empirical model, and results. We then discuss the policy implications of this study. Conclusions and suggestions for future research follow. Lexington Proposals The first proposal called for a new basketball arena for the University of Kentucky Wildcats, who have played since 1976 in city-owned Rupp Arena. The second would build a baseball stadium to attract a AA minor league baseball team. The basketball arena arose as an issue a few months after UK won its sixth NCAA national championship in Then-coach Rick Pitino, with apparent support from the university administration and backing from several private donors, proposed building a new, $100 million, university-owned arena to generate greater income for the university, to provide more modern facilities, and to allow the team greater access to its home court for practice. Although Pitino advocated the use of private money, a new arena would impose costs on Lexington taxpayers. They currently subsidize Rupp Arena operating costs and will be paying off the debt from its construction for another 20 years. The loss of Rupp s major tenant would increase the annual subsidy, a subject of much concern in the public discussion that ensued from Pitino s proposal. ii Although the Wildcats are not a major league team, they serve the same purpose in Kentucky. Many observers have noted the avid following UK enjoys, and some have

5 likened Wildcat basketball to a religion. There are no local professional sports teams to root for and, since the university s football team rarely tops the.500 mark, there s only one team that captures the attention of most people around the state. (Feinstein, 1997) If any team engenders fan loyalty and civic pride, it is the Kentucky Wildcats. In 1996 and 1997, Lexington faced requests for public funding of a minor league baseball stadium to cost about $10-$12 million. With a metropolitan population of more than 400,000, Lexington may be the largest metro area in the United States without a professional baseball team. A local developer in 1996 proposed to bring a team from the AA Southern League to town, and the league indicated it would move a team to Lexington as soon as a stadium was ready. The mayor initially supported the idea, then withdrew her support and the project seemed to die. It was revived briefly in 1997 when it appeared that state money, rather than local money, might finance the stadium. The two Lexington projects provide an ideal opportunity to examine several aspects of the public good nature of professional sports. The basketball arena would result in an improvement of the quality of an existing public good, while the baseball stadium would result in the difference between having a public good and not having one. Survey In April 1997, about one week after UK lost the NCAA national championship game, we sent a CVM survey to a sample, purchased from a professional sampling firm, of 500 randomly selected households in Lexington. Fifty of the surveys proved undeliverable. Of the 450 delivered, 230 were returned, a response rate of 51.1 percent. iii Descriptive statistics of the variables taken from the survey and used in the regressions later in this paper appear in Table 1. This section describes additional details.

6 The CVM captures both use and nonuse values (Michell and Carson, 1989). Use value is the portion of willingness to pay motivated by the revealed behavior of attending games. Nonuse value is the portion of willingness to pay that is motivated by less tangible behavior such as talking about sports with friends and family. Our survey asked about two contingent valuation markets designed to elicit willingness to pay for a government policy to (1) build an improved basketball arena and (2) build a minor league baseball stadium. iv The survey asked respondents about their consumption of UK basketball during the recently completed season. The first question asked about their private good consumption--how many games they attended in Rupp Arena. Several others asked about their consumption of nonexcludable public good aspects fan loyalty--of UK basketball. The survey results provide evidence of a substantial public goods component to UK basketball. Of 229 respondents reporting the number of games they attended in the season at Rupp Arena, 138, or about 60 percent, said they attended no games. Another 62, about 27 percent, said they attended only one to three games. Yet, although most people never attended a game, 56 percent of the respondents reported watching at least eleven games on television. Only 7.4 percent claimed to have watched no games. Seventy-two percent of all respondents report they regularly, either a few days per week or daily, read about UK basketball. Seventy-two percent also say they regularly discuss UK basketball with others. About one third of the respondents claim they "live and die" with the Wildcats while 55.5 percent profess to be casual fans. Just under half,

7 49.1 percent, of respondents believe the quality of life in Central Kentucky would fall slightly (28.6 percent) or a great deal (20.5 percent) without UK basketball. Clearly, UK basketball plays a prominent role in the lives of Lexingtonians, even though most never attend a game. The production of University of Kentucky basketball games generates substantial nonrivalrous and nonexcludable public goods. Would people be willing to pay for them? This CVM survey posed a hypothetical situation closely based on the facts of Pitino s arena proposal. The survey said a new arena might have 4,000 more seats and 30 more luxury suites than Rupp and that it might be the largest arena in the country built for college basketball. It suggested that such an arena might enhance UK s recruiting of players and its ability to reach the Final Four and to win national championships. It said that although UK had not asked for public money, the local government might have to provide larger annual subsidies to Rupp Arena if the Wildcats moved out. Fewer than 26 percent of respondents said they believe a new arena would enhance the prestige of UK and help it win a national championship. A newer, bigger arena would attract more of the respondents to UK games, but nearly half of them--105 of 224--said they would attend no games in a new arena. A key portion of any CVM survey asks people if they are willing to pay (WTP) for an enhancement in the quality or quantity of a public good. This survey s initial willingness to pay question is discrete choice: Would you be willing to pay $X per year out of your own household budget in higher taxes to help pay for a new arena? Respondents were presented with one of four different values for X, $1, $5, $10, or $25. Then, the willingness to pay values were elicited with a payment card format: What is

8 the most you would be willing to pay out of your own household budget per year to make a new arena possible? The potential responses were: Zero, Between $0.01 and $4.99, Between $5 and $14.99, Between $15 and $29.99, Between $30 and $49.99, Between $50 and $75, and More than $75. Coding the interval data at the midpoints of the intervals generates the dependent variable WTPUK for the University of Kentucky arena (Table 1). Perhaps because so few believed a new arena would help UK, more than two thirds of the respondents said they would be unwilling to pay higher taxes for a new arena, even though a large majority consume the public goods generated by the basketball team. Of the 72 respondents willing to pay higher taxes, 26 said the reason was that they like to attend UK games. Thirty-eight gave public good reasons--they like to watch on TV or discuss it with others, they think it would improve the quality of life in the area, or they say it would make them proud to have such an arena in Lexington. Of the 169 who offered reasons for their opposition to higher taxes, only 2.4 percent said it was because they cared nothing for UK basketball. The most popular answer, selected by 36.7 percent, was that taxes should not be used for a basketball arena. Others thought Rupp Arena is fine as it is or should be renovated. A little more than 11 percent thought UK made enough money from basketball to pay its own way while a similar portion thought they paid too much in taxes already. The survey also presented respondents with a hypothetical situation based on the proposal to build a minor league baseball stadium with public money. The survey said the stadium would seat 6,500, would resemble an old-fashioned baseball park, and would host 60 to 70 games per summer.

9 Because Lexington has had no minor league baseball team for decades, the survey asked no questions about past consumption of baseball. Respondents were asked how many games they thought they would attend each year. Then they were asked the same two WTP questions as in the basketball section. The WTPBB variable was coded in exactly the same fashion as the WTPUK variable described above. As with the arena, most respondents, 63.3 percent, said they would oppose higher taxes for a baseball stadium. Of those willing to pay, 46.3 percent said they wanted to attend games, a private good motivation. Few cited intangible public good motivations. No one said they would pay taxes because they wanted to discuss Lexington baseball and only one person said having a minor league team would make him proud. Nine people supported higher taxes because they thought Lexington would be a better place to live if it had a baseball team. Thirty people, or 36.6 percent of those willing to pay higher taxes, thought it would create jobs and boost the Lexington economy. This, despite the many news articles and opinion pieces in the Lexington newspaper debunking the stimulus of sports teams, lends credence to the view that a substantial portion of stadium backers are simply misled and mistaken about the economic benefits of sports. The survey concluded by asking demographic and economic questions about the respondents and their households--how many people they normally lived with, their sex, their race, their birth year, how long they had lived in Kentucky, how much education they had completed, and their household s 1996 income before taxes. They were given a space on the back cover of the survey booklet to comment on any other concern they might have about the future of Lexington.

10 Empirical Model Using standard CVM methodology, the survey results can be used to determine whether people are willing to pay for a new arena or a new stadium. Furthermore, the willingness to pay (WTP) can be decomposed into use value and nonuse value components. In the present cases, the use value represents the WTP for game attendance at a new arena or baseball stadium. The nonuse value measures the WTP for the consumption of the public goods, such as fan loyalty, that might result from an arena or stadium. In order to derive such results for the two cases under consideration, WTP Tobit v regressions for the UK and baseball models are specified in the following way: (1) WTPUK = f(income, UKGAMES1, MOREGAME, PUBGOOD, D) (2) WTPBB = f(income, BBGAMES, D) where the TAXARENA and TAXSTAD variables are the amount requested from each respondent in the discrete choice WTP question, INCOME measures the ability to pay the increased taxes, UKGAMES1 is the number of UK basketball games attended during the past year, MOREGAME is the number of additional games attended with the new arena, PUBGOOD is a variable which represents the public good aspects of UK basketball, BBGAMES is the number of baseball games attended with a minor league team, and D is a vector of standard demographic variables. Willingness to pay should increase with income, if the goods associated with the arena and stadium are normal. The expected effects of the other demographic variables on WTP are ambiguous.

11 The numbers of games attended in the past and the future are expected to have a positive relationship to willingness to pay. For the basketball arena improvement, as the number of games attended in the past increases the expected benefit of an improved arena increases. The increased number of games attended after the arena improvement and stadium construction is expected to be positively related to willingness to pay. Each of these variables measures the use value of the quality change. The non-use value of changes is equal to willingness to pay after setting these variables equal to zero. The variable PUBGOOD is a scale variable measuring the motives for non-use values of an improved UK basketball arena. Dummy variables measuring the importance of activities related to UK basketball other than game attendance were summed. The activities included the importance to the respondents of reading and discussing UK basketball, the respondents' overall interest in UK basketball, and the contribution of UK basketball to the overall quality of life. vi Empirical Results We use complete case analysis dropping any observation with item nonresponse on any variable in Table 1. vii The sample size for both the basketball and baseball models is 190. In the basketball arena model, none of the demographic variables is a statistically significant predictor of willingness to pay. We drop these variables and conduct a likelihood ratio test for the joint effects of these variables. The likelihood ratio test indicates that the demographic variables are jointly insignificant (χ 2 =4.82[6 d.f.]). Therefore, we focus on our discussion on the more parsimonious reduced model in Table 2. None of the conclusions drawn from this model differs from those of the full model described by equation (1). viii

12 Income does not affect willingness to pay. ix Each game attended in the previous season adds $0.83 to WTP. Each additional game attended in future seasons due to the new arena increases WTP by $0.73. Both results are significant and suggest use values comprise an important component of WTP Non-use value motives, as measured by PUBGOOD, are also positive and statistically significant predictors of willingness to pay. Each additional non-use value motive that is included in the PUBGOOD scale leads to an increase in willingness to pay of $1.66. Results for the Tobit regression models of determinants of willingness to pay for the baseball stadium broadly parallel the results for the basketball arena. The demographic variables are insignificant predictors of willingness to pay. The likelihood ratio test indicates that the demographic variables are jointly insignificant (χ 2 =6.09[6 d.f.]). The results for the model without demographic variables appear in Table 3. The ability to pay variable (INCOME) is positively and significantly related to willingness to pay at the.10 level, indicating that the baseball stadium is a normal good. Willingness to pay increases with the number of baseball games that the respondent expects to attend. Mean willingness to pay, use value and non-use value, and 95% confidence intervals are estimated by evaluating the coefficients from the Tobit models at the mean of the independent variables. Table 4 presents the results for both the basketball and baseball cases. Non-use value estimates are found by setting the number of games attended in each model equal to zero. Use value is then estimated as the difference between the mean willingness to pay and non-use value calculated when games equal zero. Willingness to pay for the UK arena quality improvement is $4.64, which

13 decomposes into a non-use value of $2.77 and a use value of $1.86. Willingness to pay for the baseball stadium construction is $4.12, which decomposes into a non-use value of $2.87 and a use value of $1.24. For both projects, non-use value is a significant portion of willingness to pay and is greater than use-value. The use and non-use values are statistically different for the baseball stadium at the.05 level. The use values presented here are the average use values for both users and nonusers. Closer inspection reveals that nonusers, who will never attend a game, are unwilling to pay much. In the basketball case, the WTP estimate for nonusers is insignificantly different from zero and only about one dollar. Willingness to pay estimation for baseball nonusers was impossible since so few expressed a positive WTP. The nonuse values in these cases appear to accrue almost entirely to users. They are most likely to consume whatever enhanced public goods result from the new facilities. But the samples of nonusers are small, perhaps too small for parametric estimation of NUV. Nonparametric estimation of NUVs, or estimation with a survival model, will be pursued in further research. Policy Implications Previous studies have shown that publicly funded sports buildings fail to generate enough income to justify their construction. But all previous studies have failed to account for the public goods associated with fan loyalty. Can the WTP estimates derived here justify public subsidy of sports facilities? To determine whether WTP can justify the construction of a new facility, the annual individual WTP must be aggregated for the entire population. If aggregate WTP is interpreted as the annual flow of benefits generated by the facility, the present

14 discounted value of the stream of future benefits can be interpreted as the capital value of the benefits. If the capital value of the benefits exceeds the total cost of the stadium, then construction of the stadium is justified on efficiency grounds. But private investors will want to build the stadium only if the capitalized use value benefits exceed the total costs. If use values are less than the total costs, private investors would lose even if total benefits, use and nonuse, exceed total costs. The stadium will not be built without a subsidy, and efficiency would suffer. To get the annual aggregate stream of benefits for all of Lexington, we multiply the individual mean WTP estimates by two different measures of population: estimated 1995 total population and estimated number of households in The total population will result in the higher estimate of aggregate WTP. The lower estimated aggregate WTP results from multiplying mean individual WTP by the number of households. The true value of annual aggregate WTP probably lies somewhere in between. The aggregate annual WTPs, for the basketball arena, along with use values and nonuse values, appear in Table 5. For each measure of population, the tables present upper and lower bound figures. The upper bound figures are calculated by multiplying the relevant population figure by the estimated individual WTP. The lower bound figure is 51 percent of the upper bound. Fifty-one percent of the surveys were returned. If the 49 percent of the sample who failed to return the surveys have no interest in UK basketball, they may be assumed to have zero WTP. The upper bound aggregate WTP estimates for the basketball arena range from $445 thousand per year, calculated on the basis of households, to $1.12 million, based on

15 total population. The lower bound aggregate WTP estimates range from about $227 thousand to $570 thousand per year. Suppose the new arena is expected to have a 40-year useful life and assume a discount rate of 8 percent. x Under these assumptions, the annual WTP estimates in Table 5 imply a capital value of total use and nonuse benefits from a new arena of anywhere from $2.71 million to $13.33 million, depending on which aggregate WTP figure is used. The new arena is projected to cost $100 million or more. Clearly, if Lexington taxpayers were asked to pay for a new arena on their own, the costs would exceed the benefits. This should come as no great surprise. Most respondents believe the new arena would do nothing to enhance the prestige or competitiveness of the UK basketball program. The team would not leave town or cease to exist if it failed to get a new arena. The public goods produced by the team will continue to be produced. We might expect a much different result if the Wildcats staying in town depended on their getting a new arena. Upper and lower bound aggregate WTPs for the baseball stadium can be calculated in the same way as for the arena. The results appear in Table 6. The upper bound aggregate WTP estimates for the baseball stadium range from a low of $395 thousand per year, calculated on the basis of households, to a high of $993 thousand, based on total population. The lower bound aggregate WTP estimates range from about $202 thousand to $506 thousand per year. A stadium with a 40-year life and financed at 8 percent could be justified if it cost no more than $11.8 million using the upper bound WTP for the full population and $2.4 million using the lower bound WTP for total households.

16 In the public discussions of the proposed baseball stadium, total cost figures in the range of $10 million to $12 million were commonly mentioned. It is possible, given the most generous estimates of aggregate WTP, that the total benefits of a baseball stadium could equal or exceed the total costs. But most of the WTP comes from nonuse values. It is unlikely that a baseball stadium would pay for itself from the revenues generated by its use, and so it is unlikely that such a stadium would be built without a subsidy. In fact, Lexington has no baseball stadium nor have any private investors publicly announced plans to build one. However, the baseball results must be interpreted with care. The Tobit estimation of WTP contains no significant public good variables. Few nonusers expressed a positive WTP, yet nonuse value accounts for 69.7 percent of total WTP. How could that be? Thirty respondents, virtually all of them users, said they would be willing to pay higher taxes because of the economic stimulus provided by a stadium, despite the many news articles, op-ed pieces, and editorials appearing in the Lexington newspaper to the effect that a minor league stadium would have little effect on the Lexington economy. This suggests that much of the support for public financing of new sports facilities may be based on a sort of stadium illusion, in which people attribute all activity associated with a new sports building as a net gain to a community. If, as in the two cases examined here, the nonuse benefits accrue almost entirely to users, imposing a general tax to subsidize a stadium is probably not justified, even if total WTP exceeds the cost of the stadium. Alternative means of financing, such as personal seat licenses, paid by users, should be used instead.

17 The UK arena results shed light on the case of the New York Yankees, and any other cases that might share key elements. Like the Wildcats, the Yankees are one of the top teams in the league and they want a new stadium. Like the Wildcats, the Yankees are unlikely to move out of their home market, though like the National Football League Giants and Jets, they might move to New York s suburbs. Leaving the New York metropolitan area would deprive them of their huge local media market, an asset of such enormous value that a new stadium in a smaller, less affluent market could never compensate for. If New Yorkers believe a new stadium would neither keep the Yankees from leaving the New York metro area nor turn them into contenders, it is not likely that aggregate WTP for a new stadium could justify the $1 billion or more necessary to build a new stadium in Manhattan. The Lexington baseball results are instructive, too. Minor league sports probably produce few of the public goods produced by major league sports. In Lexington, this results in nonusers not wanting to pay anything for a new stadium. If the Lexington case is typical, few minor league stadiums merit public subsidy. Of course, the Lexington cases are merely suggestive, and a definitive answer to whether public goods justify public subsidy in any particular case can only be answered with a study of that case. Conclusions This paper has presented initial results of an application of the contingent valuation method to sports arenas and stadiums. In both cases examined here, non-use value represents well more than half of total willingness to pay for new sports facilities. The results confer plausibility on the claim that the value of intangible indirect benefits of

18 sports teams exceed the value of tangible benefits. But nonuse values in these cases accrue almost entirely to users. Future research should explore the willingness to pay, use and nonuse values of sports teams in cities that have a greater probability of losing an existing team. Future CVM studies should also examine the magnitude of stadium illusion misperceptions to allow for reliable estimates of the fan loyalty component of nonuse value. References Baade, Robert A. and Dye, Richard F "The Impact of Stadiums and Professional Sports on Metropolitan Area Development." Growth and Change. Spring, pp Feinstein, John "Pitino's Head was Uneasy Before He Wore the Crown." In "1997 Final Four," supplement to Wall Street Journal. March 28, pp. 6+. Fort, Rodney "The Stadium Mess." In Stee-rike Four! What's Wrong with the Business of Baseball?. Westport, CT: Praeger Publishers, pp Greene, William H., Econometric Analysis, 3 rd Edition, Prentice Hall, Upper Saddle River, NJ, Mitchell, R. C. and R. T. Carson Using Surveys to Value Public Goods: The Contingent Valuation Method. Washington, DC: Resources for the Future. Noll, Roger G. and Zimbalist, Andrew, eds Sports, Jobs, and Taxes: The Economic Impact of Sports Teams and Stadiums. Washington, DC: The Brookings Institution. Quirk, James and Fort, Rodney D Pay Dirt: The Business of Professional Team Sports. Princeton, NJ: Princeton University Press.

19 USA Today "Special Report: The Stadium Binge, Rich Owners Often Call the Tune in Stadium Dance." September 6.

20 Table 1. Data Summary N Mean Std Dev Minimum Maximum WTPUK WTPBB UKGAMES PUBGOOD MOREGAME TAXARENA BBGAMES TAXSTAD HOUSE SEX RACE AGE TENURE EDUC INCOME

21 Table 2. Determinants of Willingness to Pay for UK Arena Variable Coefficient S.E. B/S.E. Mean Marginal Effects Constant INCOME UKGAMES MOREGAME PUBGOOD Standard error Log-Likelihood

22 Table 3. Determinants of Willingness to Pay for Baseball Stadium Variable Coefficient S.E. B/S.E. Mean Marginal Effects Constant INCOME BBGAMES Standard error Log-Likelihood

23 Table 4. Willingness to Pay Estimates 95% Confidence Interval UK Arena Mean S.E. B/S.E. Lower Bound Upper Bound WTP $ $2.91 $6.36 NUV $ $1.30 $4.24 UV $ $1.08 $2.65 BB Stadium Mean S.E. B/S.E. Lower Bound Upper Bound WTP $ $2.78 $5.45 NUV $ $1.69 $4.06 UV $ $0.79 $1.70 TABLE 5. Estimates of WTP, UseValue, and Nonuse Value of New Arena Based on population upper bound lower bound PV 40 yr annupv 40 yr annu upper bound lower bound 8% amort rat 8% amort rat arena WTP arena NUV arena UV Based on households PV 40 yr annupv 40 yr annu 8% amort rat 8% amort rat arena WTP arena NUV arena UV estimated population was 240, estimated number of households was 95,958. TABLE 6. Estimates of WTP, Use Value, and Nonuse Value of Stadium Based on population upper bound lower bound PV 40 yr annupv 40 yr annu upper bound lower bound 8% amort rat 8% amort rat stadium WTP stadium NUV stadium UV Based on households PV 40 yr annupv 40 yr annu upper bound lower bound 8% amort rat 8% amort rat stadium WTP

24 stadium NUV stadium UV estimated population was 240, estimated number of households was 95,958.

25

26

27

28 uity te uity te uity te uity te

29

30 VARIBLE DEFINITIONS WTPUK = amount in dollars respondent i is willing to pay per year for a new basketball arena for the University of Kentucky. WTPBB = amount in dollars respondent i is willing to pay per year for a new minor league baseball stadium in Lexington, Kentucky. UKGAMES1 = number of UK basketball games attended by respondent i in Rupp Arena in season. PUBGOOD = the sum of four dummy variables measuring the consumption of public goods generated by the UK basketball team. The four dummy variables are: READ = 1 if person i reads about UK basketball several times a week or more, 0 otherwise. TALK = 1 if person i discusses UK basketball several times a week or more, 0 otherwise. INTEREST = 1 if person i expresses highest level of interest in UK basketball, 0 otherwise. QUALLIFE = 1 if person i says the quality of life would fall without UK basketball, 0 otherwise. MOREGAME = 1 if person i says he would attend more games than he has in the past if UK builds a new arena, 0 otherwise. TAXARENA = $1, $5, $10, or $25, depending on the amount requested from respondent i on the discrete choice willingness to pay for a new arena question.

31 BBGAMES = the number of baseball games person i says he would attend if a stadium was built. TAXSTAD = $1, $5, $10, or $25, depending on the amount requested from respondent i on the discrete choice willingness to pay for a new stadium question. HOUSE = the number of people normally living in the same household as person i. SEX = 1 if respondent i is male, 0 otherwise. RACE = 1 if respondent i is white, 0 otherwise. AGE = age in years of person i. TENURE = number of years person I has lived in Lexington. Interval data, with maximum value of 25 assigned to those responding more than 20. EDUC = number of years of formal education completed by person i. Interval data, with maximum value of 18 assigned to those who have completed graduate or professional school. INCOME = 1996 household income before taxes, in thousands of dollars, of person i.

32 Notes i The CVM uses survey questions to elicit people s preferences for public goods by finding out what they would be willing to pay for specified improvements in them. The method is aimed at eliciting their willingness to pay in dollar amounts. It circumvents the absence of markets for public goods by presenting consumers with hypothetical markets in which they have the opportunity to buy the good in question. Because the elicited willingness to pay values are contingent upon the particular hypothetical market described to the respondent, this approach came to be called the contingent valuation method. (Mitchell and Carson, 1989, pp. 2-3.) ii Pitino left UK for the NBA in 1997, but the University remains interested in building its own arena. It financed a feasibility study in 1997 of a new arena and in 1998 it extended its soon-to-expire lease at Rupp for six years, just long enough, some have noted, to plan, finance, and build a new arena. Results for surveys returned before and after Pitino left are not statistically different. iii We took several steps to improve the response rate. We pretested the survey to determine whether any unanticipated problems would occur by mailing the survey to a random sample of 40 Lexington households. In the main mailing, we offered people who responded within one week the chance to win $100 in a lottery drawing. One week after the main mailing, we sent postcards to all 500 addresses asking them to please complete the survey promptly, and thanking them if they had already done so. One month after the main mailing, duplicate surveys were mailed to all addresses that had not yet responded

33 to the original mailing. iv The survey instrument and details about the coding of the variables are available from the website: v The WTP dependent variables can be analyzed in several ways, including ordinary least squares. Ordinary least squares will probably be inferior to other methods, however, since the data (1) are censored at zero (there are no negative willingness to pay values) and (2) the willingness to pay values are intervals, not continuous data (Greene, 1997). Estimating the models with the interval data techniques would ignore the censoring. Estimating the models with the Tobit would require that the midpoints of the WTP intervals be used. While there are costs and benefits associated with both methods the Tobit model is chosen. For many issues, payment card willingness to pay questions generate many zero willingness to pay responses. These data are no different with 68% and 63% of the sample expressing zero willingness to pay for the UK and baseball contingent markets. Ignoring this large percentage of zero values would result in significant bias. vi PUBGOOD in this sample is almost uniformly distributed over the 0,4 interval. Considering the relatively small sample size, the scale is reliable according to Cronbach s alpha (α=.69). We also estimated models using the variable representing the contribution of UK basketball to the overall quality of life instead of PUBGOOD. Results are similar so we proceed with the PUBGOOD variable. An appendix containing these results is available upon request. vii For both equations (1) and (2), the simultaneous equations Tobit model (Greene, 1995, p. 608) was used to test for exogeneity since WTP and the number of games to attend are

34 joint decisions. Several models were used to explain each of the potentially endogenous variables, UKGAMES1, MOREGAME, and BBGAMES. In each case, the null hypothesis of exogeneity could not be rejected. Therefore we proceed with the assumption of exogeneity. viii The results for the full model, including demographic variables, are available on request. ix This result may be due to the low percentage of willingness to pay relative to income, measurement error of the independent and/or dependent variables, or that improvement in the basketball arena is an income-neutral good. x The period and the discount rate are meant to be illustrative only. Some stadiums have useful lives far longer than 40 years, while others are replaced in much shorter periods. As for the proper discount rate to use, it depends on many factors. As measured by attendance, use value from teams and stadiums depends heavily on the winning record of the team presumably so does the nonuse value. The proper discount rate to apply to future expected benefits should reflect the risk of bad seasons as well as good.