A Closer Look at Stadium Subsidies
Tuesday, April 29, 2008
Despite what many people believe, professional sports venues typically do not spur large-scale economic activity.
The 2008 Major League Baseball season will be the last one played in Yankee Stadium. After 85 years, the most storied venue in American sports will be torn down. Starting in 2009, the New York Yankees will play in the “new” Yankee Stadium, built right next to the old one. Across town in Queens, another New York baseball stadium, much less famous and hallowed, will also shut its doors after 2008. The close of Shea Stadium, home to the New York Mets since 1964, will make way for Citi Field. Both projects have been in the works for some time. While the new Yankee Stadium has been heavily subsidized by New York taxpayers, Citi Field is entirely a private endeavor—which, as modern sports stadiums go, makes it somewhat unusual.
Since 1990, construction of stadiums and arenas for professional sports franchises has occurred at an incredible pace. In that time, Major League Baseball (30 teams) has opened 19 new stadiums and has three more currently under construction. The National Football League (32 teams) has opened 17 new stadiums; done major renovations to four others; has three under construction; and has four more projects at various stages of planning and negotiations. The National Basketball Association (30 teams) has opened more than two-thirds of its 30 arenas since 1990, and at least three NBA franchises are actively seeking new arenas.
In most cases, state and local governments have been closely involved in the financing, design, construction, and management or ownership of professional sports facilities. Even Washington has played a role: the local and state bonds used to fund new stadiums and arenas typically are exempt from federal income tax. This has been the subject of hearings before the House Oversight Subcommittee on Domestic Policy, with some lawmakers questioning whether subsidizing stadiums for private gain is consistent with the goal of aiding “public” infrastructure projects. Hundreds of millions of tax dollars are at stake, so it is important for business leaders and elected officials to understand the costs and benefits of publicly financed stadiums.
Since 1990, construction of stadiums and arenas for professional sports franchises has occurred at an incredible pace.
Both must be seen in historical context. A 1926 article in a magazine called The Playground documented a stadium construction boom that began shortly after World War I. “Not only universities but cities and high schools and private agencies are also joining the stadium ranks and building large structures to accommodate the crowds who attend the athletic activities, festivals, pageants and other large community events,” it reported. The number of stadiums increased from 11 in 1917 to 70 when the article was written. In the five years from 1921 to 1926, 56 new stadiums were built, an average of more than 11 per year. Writing in 1957 in a magazine called Municipal Finance, economist Ralph Wulz suggested that the stadium boom was initiated in part by municipalities that “flocked to the fold in constructing huge memorials to war veterans and war casualties.” The subsequent growth in public ownership of such facilities was, Wulz suggested, fueled by Depression-era “make-work” programs. The stadium construction that occurred between 1917 and 1926 cost around $25.5 million—roughly $295.65 million in 2000 dollars.
The Playground article pointed out that the shape, size, and cost of stadiums were three of the most important factors to be considered. Shape mattered because the traditional ellipse, or Roman plan, made it more difficult to hold track events that required a straightaway, while the horseshoe structure did not have this problem. Planners also had to make sure the stadium could be expanded as the city grew or when “the university has graduated more enthusiastic alumni to add to those already crowding the bleachers for the big games.” Two additional concerns listed in 1926 were the adaptability of the structure and its management. Adaptability was important “in order that the stadium may have as broad a use as possible.”
Concerning cost, the 1926 Playground article suggested that $10 to $15 per seat was a reasonable expenditure. The stadium built at Ohio State University in 1922 cost $1.7 million and seated 63,056—a cost per seat of $26.96 in 1922 dollars, or $277.58 in 2000 dollars. To give some perspective, Brookings Institution economist Benjamin Okner reported in a 1974 volume entitled Government and the Sports Business (edited by Roger Noll) that Pittsburgh’s Three Rivers Stadium, which sat 59,594 and operated from 1970 to 2000, had per seat construction costs of about $2,964 in 2000 dollars. Paul Brown Stadium in Cincinnati, which seats 65,535 and opened in 2004, cost around $6,104 per seat in 2000 dollars. The Ohio State stadium was financed, in part, by a $1 million pledge from alumni, students, and “ardent supporters”; the remainder was covered by football game receipts. Three Rivers Stadium was financed using bonds issued by the city of Pittsburgh, and Paul Brown Stadium has been financed by a sales tax increase. The stadium at Ohio State was intended for use by the university’s football, baseball, track, indoor tennis, golf, wrestling, and fencing teams, and also for intramural athletics. Three Rivers Stadium was home to the Pirates of Major League Baseball and the Steelers of the NFL, while Paul Brown Stadium is used by the NFL’s Bengals.
In 1957, Wulz argued that one basis for public ownership of stadiums was that “private enterprise could not provide the service which the public demanded and at the same time realize an adequate profit upon its investment.” He also noted that the facilities did host events of a “civic nature and thus may be worthy of community-wide support.” Recognizing that city managers had been tasked with operating enterprises “which lost money or indeed went bankrupt under private management,” Wulz focused on how much the municipality should charge the private entities using the stadium or auditorium. Only after an analysis of the facility and its uses, he said, could public officials determine if “a tax subsidy is warranted.” What types of activities deserve a public subsidy? “Governmental activities and perhaps activities at which no admission is charged,” said Wulz. “At the same time,” he reckoned, “commercial type activities” should “pay the full cost of the services or facilities which are provided.”
Depending on how one measures the public share of stadium costs, it ranges from 58 percent to 63 percent after 2000.
Clearly, stadiums built with public funds have evolved over time. No longer are they built to honor the sacrifices of American soldiers. No longer are they built to be flexible venues capable of hosting a great variety of events. And no longer does the public sector determine the appropriate price to charge private enterprise for use of this publicly supplied resource. Today, sports stadiums are largely the private domain of for-profit businesses that the public sector subsidizes, often with special taxes.
In recent years, Judith Long, an urban planning expert at Harvard University, and Andrew Zimbalist, an economist at Smith College, have produced comprehensive discussions of stadium and arena subsidies. State and local governments assist pro sports franchises in myriad ways, including covering building and operating costs. To truly understand the extent of stadium subsidization, it is important to account for all the ways in which public money pays for the facilities. Public sources of information tend to focus on the public share of capital costs, which can produce misleading results regarding the extent of the public subsidy. For example, Zimbalist and Long show that for most of the stadiums and arenas built for professional sports franchises and in operation since 1990, “when net operating costs are included, the public share goes up.”
Depending on how one measures the public share of stadium costs, it ranges from 58 percent to 63 percent after 2000. The average public contribution to the total of capital and operating cost is between $149 million and $161 million in 1995-99, and between $249 and $280 million in 2000-06. Additionally, Long and Zimbalist note that public participation varies among stadiums, which suggests that it is highly sensitive to the bargaining skills and efforts of the municipal officials in different cities.
Over time, both the purpose and the real cost of public support for stadiums and arenas have changed. It may be that the subsidies state and local governments provide for stadium and arena construction and operation are justified by the community benefits those facilities provide. But the evidence says otherwise.
It is not quite correct to argue that local governments could use the tax revenues they spend on stadiums in “better” ways, such as on schools or health programs. Typically, the funding for stadiums does not come directly out of an existing government budget but rather from a new source of revenue, like special taxes on tickets or add-ons to the local sales tax. The municipality likely would not impose these taxes for any purpose other than subsidizing the stadium, so other governmental services are not necessarily being shortchanged. (Of course, the increased taxes do reduce the disposable income of local consumers, so the stadium subsidy does impose opportunity costs on citizens, despite having no such effect on the government’s budget.)
The most basic question about stadiums, arenas, and sports franchises is the extent to which they contribute to the vitality of the local economy. Supporters of publicly financed stadiums argue that the benefits are substantial, while opponents say they are small and highly concentrated among the wealthiest citizens. To buttress their case, supporters mostly use economic impact studies that predict how the local economy will be affected by the stadium, while opponents compare the economy before and after the facility is constructed. Supporters tend to imply that redistribution of economic activity from the suburbs or outlying areas of a city to the downtown is desirable, while opponents generally oppose this sort of redistribution and focus instead on job and income creation.
There is little evidence of large increases in income or employment associated with the introduction of professional sports or the construction of new stadiums.
The typical economic impact study gathers data on all aspects of spending related to a stadium, including the money spent to build it and the money spent by fans in connection with the stadium (including on tickets, at restaurants, and at hotels). The impact of this spending ripples outward into other areas of the economy through a multiplier. By linking spending to employment, the study then calculates how many jobs a stadium has created. It does not perform a cost-benefit analysis, which would address the opportunity costs of raising taxes to pay for a stadium and consider alternative uses of those funds.
Academic researchers have examined the prospective economic impact studies and found a variety of methodological errors in them, all of which raise doubts about the magnitude of the predicted spending and job increases. Other scholars use data from multiple years before and after stadium construction to measure the impact of the stadium. These ex post studies reject stadium subsidies as an effective tool for generating local economic development.
My own research, conducted with economist Brad Humphreys (who is now at the University of Alberta), has used perhaps the most extensive data, incorporating yearly observations on per capita personal income, employment, and wages in each of the metropolitan areas that was home to a professional football, basketball, or baseball team between 1969 and the late 1990s. Our analysis tried to determine the consequences of stadium construction and franchise relocations while controlling for other circumstances in the local economy. Scholars Robert Baade, Allen Sanderson, Victor Matheson, and others have taken slightly different approaches, but the results are fairly constant from one analysis to another. There is little evidence of large increases in income or employment associated with the introduction of professional sports or the construction of new stadiums.
Indeed, my work with Humphreys finds that the professional sports environment—which includes the presence of franchises in multiple sports, the arrival or departure of teams, and stadium construction—may actually reduce local incomes. For example, we found that the overall sports environment reduced per capita personal income, a finding that was new in the economic literature at the time we published it (1999). We also found that, in many local economies, wages and employment in the retail and services sectors have dropped because of professional sports.
There are several possible explanations for why development does not occur. First, consumer spending on sports may simply substitute for spending on other types of entertainment—and on other goods and services generally—so there is very little new income or employment generated. Sports fans that attend a game may reduce their visits to the movies or to restaurants to free up finances for game tickets and concessions. Patrons of local restaurants and bars who come to watch the games on television also are likely to cut back on their other entertainment spending.
Second, compared to the alternative goods and services that sports fans may purchase, spending related to stadium attendance has a relatively small multiplier effect. This is because spending at the stadium translates into salaries for wealthy athletes, many of whom live outside the city where they play. High-income individuals generally spend a smaller fraction of their income than low- and middle-income people—and much of the spending professional athletes do occurs in a different community than where they earned it. So the money paid to players does not circulate as widely or abundantly as it would were it paid to people with less wealth and more attachment to the city.
Third, whether the stadium subsidy comes from raising local taxes or from slashing public services—or from both—its effect is to reduce the net spending generated by the stadium project. Plus, imposing new taxes introduces new administrative costs and makes the economy less efficient. Consider the common practice of funding stadium and arena subsidies with new taxes on hotel occupancy and rental cars. One argument for such taxes at the local level is that they are paid by outside visitors, many of whom may be in town to see the sporting events. But the taxes would also be paid by traveling businessmen and conventioneers. When comparing cities to host an upcoming meeting, businesses and professional associations may select between otherwise comparable cities based on which one has the lower hotel and rental car taxes. In other words, the new taxes used to subsidize the stadium construction may ultimately reduce visits to the city by non-sports-related travelers.
My work with Humphreys finds that the professional sports environment may actually reduce local incomes.
Newspapers routinely run articles about the great business done by bars and restaurants in the neighborhood of a stadium. Yet such anecdotal evidence is unconvincing: it simply shows that the recipients of highly concentrated benefits are easier to find than those who suffer the extremely diffuse costs. The aggregate economic studies account for the distribution of gains and losses, and they find no proof that the gains outweigh the losses.
The principal criticism of research finding no net economic boon from stadiums is that downtown stadiums are likely to have larger benefits than suburban stadiums. Yet this analysis is heavily influenced by stadiums constructed in the late 1960s and 1970s, which were located predominantly in the suburbs. For example, Thomas Chema, former executive director of Cleveland’s Gateway Economic Development Corporation, says that the value of stadiums “as catalysts for economic development...depends upon where they are located and how they are integrated into a metropolitan area’s growth strategy.”
Economists Rob Baade (Lake Forest College), Mimi Nikolova (Lake Forest College), and Victor Matheson (College of the Holy Cross) provide stark visual evidence in support of this argument, comparing the impact of Chicago’s Wrigley Field and U.S. Cellular Field. The economic development possibilities near U.S. Cellular Field are obviously limited by the vast parking lot and multilane highway that surround the stadium. City and regional planners Arthur C. Nelson (Virginia Tech University) and Charles Santo (University of Memphis) each find that teams that play in the central business district of a city tend to be associated with an increase in the metropolitan area’s share of the regional income. Of course, this could simply be evidence of income redistribution, or of economic activity shifting from one area to another, rather than evidence of region-wide benefits.
Analysts such as Santo, along with Cleveland State University economists Ziona Austrian and Mark Rosentraub, suggest that the real question should be: Does a stadium help the redevelopment of an area that actually needs redevelopment? To them, a downtown area is deserving of help even if that help comes at the expense of the rest of the area. From this perspective, the studies that find little economic growth flowing from stadiums and sports franchises are not relevant. Instead, the mere possibility that a new stadium will aid urban redevelopment in a central city or downtown area is a sufficient rationale for the subsidies.
A second—and, to me, more telling—criticism of the existing research is that the data used are not precise enough to capture the true economic effects of stadiums. Economists such as Baade, Matheson, Robert Baumann (College of the Holy Cross), Marc Lavoie (University of Ottawa), and Gabriel Rodriguez (University of Ottawa) argue that sports comprise much too small a component of the local economy for the effects to be visible in aggregate data. In their research, Baade, Baumann, and Matheson use monthly sales tax data, while Lavoie and Rodriguez’s analysis utilizes monthly hotel occupancy rates in Canadian cities, to find a less aggregate measure of economic activity that may be linked to sporting events. But in the end, evidence from the monthly data does not support the notion that stadiums and sports franchises deliver sizable economic benefits.
It is clear that not all citizens in a community benefit equally from the presence of professional sports franchises in their city.
Despite the lack of evidence for widespread growth in income and employment, or even increased hotel occupancy, state and local government subsidies for stadiums, arenas, and professional sports may still be sound public policy. They certainly provide consumption and public-good benefits that are not measured by pure economic data. Long and Zimbalist have provided credible estimates of the costs of subsidies. If the consumption benefits derived from game attendance and the public-good benefits of having a franchise are greater than these costs, then the welfare of the community is enhanced by financing stadiums. But how big are the consumption and public-good benefits?
The consumption benefits are the most straightforward to estimate. They are no different than net benefits from consumption of any other good. Information on ticket prices and attendance allows us to determine total spending at the stadium. If ticket demand falls as the price of tickets rises, we can find the difference between the maximum amount people are willing to pay for tickets and the actual payments made. This difference, known as consumer surplus, represents the value of the consumption benefits from attendance at the stadium. Estimates of these benefits vary by sport and city, and they depend on how responsive attendance demand is to changes in ticket prices. However, the results suggest that public-good benefits of around $10 per person may be sufficient to justify stadium subsidies.
Interestingly, Humphreys and I found that the overall sports environment—which, as mentioned earlier, includes the presence of franchises in multiple sports, the arrival or departure of teams, and stadium construction—in a given area reduced per capita personal income by about $10. In other words, every man, woman, and child in the metropolitan area was poorer by $10 as a result of the sports environment. This suggests that the public-good benefits are worth just enough to pay for the subsidies. Sports economists and policy analysts are using a variety of methods to get more precise estimates of the public-good benefits. In the future, we should know better whether these benefits are sufficient—in combination with private consumption benefits—to cover the public financing of professional sports stadiums.
Of course, even if the benefits of stadiums and arenas cover the subsidies, the subsidies still may not be sound policy. First, there may be enormous variation in the distribution of the consumption and public-good benefits. It is clear that not all citizens in a community benefit equally from the presence of professional sports franchises in their city. Indeed, because the tax revenues used for the subsidies are often generated from lotteries and sales taxes whose burden falls disproportionately on the poor, while the consumption benefits go mostly to relatively wealthy sports fans, the net benefits are distributed regressively. Second, we should consider the net benefits to the community of alternative uses of the funds spent subsidizing sports facilities. Good policy means using the money where the net benefit is greatest, not simply where the net benefit is positive. That’s something state and local governments should keep in mind before pledging millions of dollars to fund the next new stadium project. And it’s something Congress should remember when evaluating the future of U.S. tax policy.
Dennis Coates is a professor of economics at the University of Maryland, Baltimore County.
Image by Shutterstock/Darren Wamboldt.