$19B for Stadiums — The Real Economics of Pro Sport (2026)

The core finding of this report can be stated plainly. Professional sport is a highly profitable business whose revenues come mainly from private media contracts, whose capital gains flow overwhelmingly to franchise owners, and whose venues are still routinely paid for in large part by the public ✓ Established [18]. Since 2000, state and local governments in the United States and Canada have committed $19 billion to new major league venues, roughly $330 million per facility [18]. The economic research on what that money buys is unusually consistent. A comprehensive 2023 survey by the economists John Charles Bradbury, Dennis Coates and Brad Humphreys concluded that it is accurate to describe the academic consensus as generally opposed to public underwriting of professional sports venues [17]. In June 2026 Coates told the House Ways and Means Committee that the answer to whether stadiums are effective tools of local development is short: no [19]. The gap between that evidence and continuing public spending is the subject of this report.

The largest single source of money in world sport is the NFL media package signed in March 2021. Its agreements with CBS, Fox, NBC, ESPN and Amazon run for 11 years through the 2033 season and are worth more than $110 billion, roughly $10 billion a year and more than double the value of the previous contracts ✓ Established [3]. CBS, Fox and NBC each pay an average of more than $2 billion a year, and ESPN about $2.7 billion a year for Monday Night Football [3]. The league kept an option to leave most of the contracts after the 2029 season, which gives it a chance to resell the rights at a higher price before the deals expire [3]. Because the league sells these rights collectively, the proceeds are split evenly. According to S&P Global Market Intelligence, the league’s shared revenue passed $14 billion for the latest season [1]. Divided equally, that is more than $430 million per club before a single ticket, hot dog or sponsorship is sold locally.

Basketball has followed the same path at a steeper angle. In July 2024 the National Basketball Association signed 11-year agreements with Disney, NBCUniversal and Amazon covering the 2025-26 to 2035-36 seasons, worth a reported $76 billion ✓ Established [4]. Disney will pay an average of about $2.6 billion a year, NBC about $2.5 billion and Amazon about $1.8 billion, an increase of roughly 2.5 times in annual fees to an average of nearly $7 billion [4]. The effect on the league’s books was immediate. Final accounting for 2025-26 put basketball-related income, the revenue pool that players and owners divide, at $11.68 billion, up from $10.25 billion the year before [15]. That is growth of 14% in a single season, driven almost entirely by a contract negotiated in a boardroom rather than by attendance or local spending. The pattern matters for what follows: the revenue that makes franchises valuable increasingly comes from national and global audiences, not from the city that hosts the team.

European football shows the same dependence on broadcasting, but also its limits. In December 2023 the Premier League sold its UK rights for 2025-26 to 2028-29 to Sky Sports and TNT Sports for a record £6.7 billion over four years, with up to 270 live matches a season [5]. The league described the result as a 4% increase in live rights value on the previous process, far below the leaps of the 2010s [5]. Deloitte’s Annual Review of Football Finance shows the consequence ✓ Established. Premier League clubs generated £6.8 billion in 2024-25, up 8%, but broadcast revenue rose only 2% to £3.4 billion [6]. Growth came instead from commercial income, up 13% to £2.4 billion, and from matchday revenue, which rose 15% to pass £1 billion for the first time [6]. When television money stops rising quickly, clubs turn to sponsors and to the people in the seats. Fans, in other words, absorb the slowdown.

Why do broadcasters pay these sums. Live sport is one of the few kinds of programming that large audiences still watch at the moment it airs, which makes it valuable to advertisers and to streaming services seeking subscribers. Supply is also controlled. Leagues sell rights collectively rather than club by club, and they set the number of games, teams and broadcast windows themselves [1] [3]. New supply is now emerging at the margins of the market. Deloitte projected that global revenue from women’s elite sport would reach at least $2.35 billion in 2025, up from $981 million in 2023, with basketball and football the largest contributors [44]. Commercial deals made up 54% of that projected total and broadcasting 25% [44]. Even so, women’s elite sport in its entirety earns about a tenth of what the NFL alone earns in a year [2] [44]. The money in professional sport is concentrated in a handful of leagues that control scarce, live and highly marketable inventory.

None of this revenue is controversial in itself. Broadcasters, sponsors and fans pay voluntarily for something they value, and leagues have become very good at selling it. The public question begins elsewhere: at the stadium, where the same businesses that sell $110 billion media contracts ask cities and states to pay for the buildings in which the games are played [3] [18]. It continues through the tax code, where federal tax-exempt bonds have helped finance most major venues built since 2000 [20]. And it ends with a claim that local leaders repeat at almost every stadium announcement, that the team and its venue will generate jobs, growth and tax revenue sufficient to justify the investment [19]. The sections that follow trace each step: how franchise values are built, how revenue is divided between owners and players, what taxpayers contribute, what the evidence says about local returns, how other countries organise the same business, and what policy could change.

The headline numbers are now remarkable even by the standards of financial markets. Forbes put the Dallas Cowboys at $17 billion in its 2026 list, the most valuable team in the league for the 20th consecutive year, followed by the Los Angeles Rams at $13.5 billion, the New York Giants at $12 billion and the New England Patriots at $10.6 billion ✓ Established [8]. Ten NFL clubs are now valued at $10 billion or more [8]. A year earlier the Cowboys were valued at $13 billion and the average team at $7.1 billion, which Forbes said was 25% above 2024 and 104% above 2021 [9]. For the first time in 2025 every team was worth at least $5 billion, with the Cincinnati Bengals setting the floor at $5.25 billion [9]. Taken together, the figures imply that the average NFL franchise has roughly tripled in value in five years. Few asset classes of comparable size have performed as well over the same period, and none with the same degree of revenue already contracted years in advance.

Revenue alone does not explain these prices. If the NFL’s $23 billion in revenue were split evenly, each club would bring in a little over $700 million a year [2]. An average value of $10.36 billion therefore equals about 14 times annual revenue [8]. Three features of the business support that multiple. The first is scarcity: there are 32 franchises, and the league decides whether and when to add more. The second is certainty: national media payments are contracted through the 2033 season, so a large share of future income is already known [3]. The third is the market for buyers. Sports teams have become trophy assets for very wealthy individuals and, increasingly, for institutional investors. These features turn a sports franchise into something closer to a monopoly licence than an ordinary company, and they explain why owners can sell at prices that bear little relation to the annual profit of the club.

Institutional money has added a new layer to the market. In August 2024 NFL owners voted for the first time to allow private equity funds to buy stakes of up to 10% in a club ✓ Established [11]. Only league-approved investors were admitted: Arctos Partners, Ares Management and Sixth Street, plus a consortium of Blackstone, Carlyle, CVC, Dynasty Equity and Ludis [11]. Each investment must be at least 3% of a club and must be held for at least six years [11]. The practical effect is liquidity. An owner can now sell a minority slice at a market price without selling control, and every such sale creates a new public reference point for what a team is worth. Valuations that were once based largely on a handful of whole-club sales are now reinforced by a steady stream of minority transactions. One of the approved vehicles, Ludis, was founded by the Hall of Fame running back Curtis Martin to bring more investors from racial and ethnic minorities into NFL ownership [11].

These valuations remain estimates, and estimates differ ⚖ Contested. In August 2026 Sportico valued the Cowboys at $15.5 billion [10]; a few weeks later Forbes put the same club at $17 billion [8]. A gap of $1.5 billion on a single franchise shows how much depends on method, from the treatment of stadium revenue to the multiple applied to future media income. Stadium revenue is the important link. When a club secures a new venue financed largely with public money, the income stream from that venue becomes part of what the club is worth. Bradbury, Coates and Humphreys document an example: the Atlanta Falcons received $200 million in bonds toward Mercedes-Benz Stadium, but the team retains perpetual access to dedicated hotel tax revenue after the bonds are repaid, a stream with a net present value of $700 million to the owner [18]. Public support of that kind does not appear in a single budget line, but it does appear in the sale price of the team.

European football shows that high values do not require high profits. Premier League clubs reported combined pre-tax losses of £948 million in 2024-25, up from £135 million the year before, largely because of spending on player transfers [7]. Wage costs reached a record £4.4 billion, and the average ratio of wages to revenue stayed at 65% [6]. Yet English clubs continue to change hands at high prices, and investors from the Gulf and the United States compete for them. Saudi Arabia illustrates a further variant. In June 2023 the Public Investment Fund took 75% ownership of four clubs, Al Ittihad, Al Ahli, Al Nassr and Al Hilal, and Saudi Pro League clubs then spent a record $957 million in the summer transfer window [43]. In both cases the buyer is often pursuing prestige, visibility or national strategy rather than dividends. Where the buyer is a state fund, the line between private investment and public subsidy disappears entirely.

Franchise values are therefore a measure of what owners can extract, not of what a community gains. A rising valuation tells us that media income is secure, that the number of teams is limited and that buyers with deep pockets are willing to pay for the privilege of ownership [8] [11]. It does not tell us whether the city that hosts the team is richer for it. That distinction is often blurred in public debate, where the value of a team is cited as if it were a measure of its importance to the local economy. Even the Bengals, the least valuable club in the 2025 list, were worth $5.25 billion, a figure that says nothing about the economy of Cincinnati [9]. The two are different things, and the evidence on the second question, examined in section five, is far less flattering than the Forbes list. Before reaching it, the report turns to the other parties with a claim on the revenue: the players who produce the product on the field.

In the major North American leagues the division of revenue is set in writing. The NFL’s 2020 agreement raised the players’ share to 47% in 2020 and a guaranteed 48% from 2021, with the ability to reach 48.5% through a media kicker as new television contracts took effect ✓ Established [12]. The agreement runs through the 2030 season, giving the league a decade of labour peace in the period when its media income more than doubled [12] [3]. The salary cap is an output of that formula rather than a wage budget. For 2025 the league set the cap at $279.2 million per club, up from $255.4 million in 2024, an increase of $23.8 million in a single year [13]. When revenue rises, the cap rises with it. In that sense players share directly in the growth of the media business. What they do not share in is the second source of owner wealth: the change in the value of the club itself.

Basketball uses a similar model with a slightly larger player share. The NBA’s agreement calls for players to receive 51% of basketball-related income in salaries and benefits, and the league withholds 10% of player salaries each year in escrow so that the split is achieved precisely ✓ Established [14] [15]. The salary cap for 2025-26 was set at $154.647 million [14]. With basketball-related income of $11.68 billion for that season, the players’ 51% share was worth about $5.96 billion [15]. The 2023 agreement also limited how fast the cap can rise in a single year, a response to the 2016 spike that followed the previous media deal [14]. The design reflects a hard lesson for both sides. A sudden jump in television money benefits whichever players happen to be free agents that summer, and leagues and unions now prefer to smooth the gains over several seasons rather than see them captured by a few.

Baseball, the only major North American league without a salary cap, is where the division of revenue is now openly contested ⚖ Contested. Its agreement with the players expires on 1 December 2026, and a lockout is widely expected [45] [16]. Owners have proposed a 50-50 split of defined revenue, a team spending cap of $245.3 million in 2027 and a payroll floor of $171.2 million, with an escrow mechanism to claw back salaries if the players’ share exceeds 50% [16]. The players’ union argues that the players’ share in 2026 is projected to be well above 50%, and that the proposal would have cost them more than half a billion dollars had it applied this season [16]. Bruce Meyer, the union’s interim executive director, has said that the union has never been broken and never will be [45]. The dispute turns on definitions. What counts as revenue, and what is deducted before it is shared, determines whether a 50% split is generous or a pay cut.

European football divides revenue in a very different way, because it has no collective cap and no fixed split. Clubs compete in an open market for players, and Premier League clubs spent an average of 65% of revenue on wages in 2024-25, a total of £4.4 billion [6]. On that measure players capture a larger share of operating income than their American counterparts. The price of that competition is chronic loss-making. The £948 million in combined pre-tax losses reported for 2024-25 means that many owners are subsidising their clubs from outside the football business [7]. This is the model in which owners pursue trophies, status or a future sale rather than an annual profit. It also explains why European clubs rarely receive American-style stadium subsidies: an owner who is already funding losses has little leverage to threaten a move, because in a promotion and relegation system the club cannot easily be transplanted to another city.

Fans are the third party in the revenue split, and they pay at several points. Team Marketing Report’s Fan Cost Index, which prices four average tickets, two beers, two soft drinks, four hot dogs and one parking space, rose 8.9% in the NFL in 2024 to $691.41, and the average ticket rose 12.2% to $136.38 [42]. In the Premier League matchday income rose 15% in 2024-25 to pass £1 billion for the first time [6]. Viewers who watch at home pay through subscriptions to the broadcasters and streamers whose rights fees fund the leagues [3] [4]. And many of the same people pay again as taxpayers when a new stadium is built [18]. A family that buys tickets, subscribes to a sports service and lives in a jurisdiction that funds a venue contributes to the team’s revenue three times. Only one of those payments is a voluntary market transaction in the usual sense.

The division between players and owners is negotiated between two organised parties with lawyers, economists and the threat of a strike or lockout. Taxpayers are not at that table. They enter the system through a separate negotiation, between a team and a city or state, in which the team typically holds the stronger hand because it can credibly threaten to move [18]. The asymmetry is important. Players have bargained their way to roughly half of league revenue [12] [14]. The public, which in many cities pays for a large share of the stadium, usually receives no share of revenue at all, and often forgoes property tax on the building as well [18]. NFL clubs shared more than $14 billion in national revenue last season, and none of it passed through a municipal budget [1]. The next section measures how large that public contribution has become, and how fast it is growing.

The public share of stadium costs has fallen in percentage terms while rising in absolute terms. Bradbury, Coates and Humphreys find that the median public share of construction costs declined from 70% in the 1990s and 2000s to about half in the 2010s, and that venues opened or planned in the 2020s have received roughly 40% of their funding from taxpayers ✓ Established [18]. Because construction costs have soared, a smaller share now means a larger cheque. The median public contribution for venues opened in the 2020s is $500 million [18]. The reported numbers also understate the true cost. Work by Judith Grant Long, cited in the same review, finds that unreported public contributions, such as land, infrastructure and forgone taxes, raise public obligations by 25% to 40% above the reported figures [18]. The authors warn that a new wave of stadium replacement is approaching, and that public outlays show a positive growth trajectory that could prove costly to government budgets in the coming decades [18].

Buffalo shows how a deal evolves from announcement to opening. In March 2022 New York State and Erie County agreed to provide $850 million toward a new stadium for the Bills, $600 million from the state and $250 million from the county, against a projected cost of $1.35 billion ✓ Established [21]. That was about 63% of the projected cost, and at the time it was the largest public commitment ever made to an NFL facility [21]. The NFL lent the club $200 million, and the owners, Terry and Kim Pegula, were required to at least match it [21]. The new Highmark Stadium opened on 23 June 2026 at a final cost of about $2.1 billion [22]. The club covered the overruns and took on additional debt as costs rose [23], so the public share of the final bill fell to roughly 40%. The public commitment, however, did not fall at all. It was fixed at $850 million from the start, plus the value of the land and the operating arrangements that come with it.

Nashville then set a new record. The Tennessee Titans’ domed stadium is projected to cost $2.1 billion and is due to open in 2027 [24]. The public share of $1.26 billion consists of $500 million in bonds from the state and $760 million in revenue bonds issued by the Metropolitan Sports Authority, the largest sum of public money ever committed to a professional sports venue in the United States ✓ Established [24]. The local obligation is often described as a fraction of the total and as self-financing, because it is to be repaid from future tax proceeds generated at the site [18]. Bradbury, Coates and Humphreys translate it into a more relatable measure: the $760 million local obligation is equivalent to roughly $2,600 per Metro Nashville household, and the $500 million state contribution adds about $190 per Tennessee household [18]. The stadium it replaces opened in 1999. A building less than three decades old, in other words, is being replaced with public money.

The Chiefs case shows how interstate competition raises the price. Kansas first offered STAR bonds covering up to 70% of stadium costs to both the Chiefs and the Kansas City Royals in 2024, and a bipartisan council of lawmakers later extended the deadline for the teams to accept until 31 December 2025 [27]. Missouri, the teams’ existing home, answered with a package that would pay for up to 50% of new or renovated stadiums [27]. Economists warned that the arithmetic did not work. STAR bonds are repaid by sales taxes collected in a defined district around the project, and much of that spending would simply be redirected from other local businesses [26]. Geoffrey Propheter of the University of Colorado Denver put it bluntly: you are not going to generate enough net revenue to cover one of the facilities, let alone two [26]. Kansas approved the Chiefs deal anyway [25]. The two states are part of the same metropolitan economy, so the region gains a stadium it already had while one state pays to move it a few miles.

Washington and Las Vegas complete the picture of the current wave. In September 2025 the District of Columbia Council gave final approval, by 11 votes to two, to a plan for the Washington Commanders to return to the site of RFK Stadium, with the team contributing $2.7 billion and the city about $1.1 billion for the stadium, housing, green space and a sports complex [28]. Mayor Muriel Bowser called it the largest economic development project in DC history, and the venue is due to open in 2030 [28]. In Nevada, Senate Bill 1 capped public financing for the Athletics’ new 30,000-seat ballpark on the Las Vegas Strip at $380 million of an estimated $1.5 billion, through county bonds, a special tax district and transferable tax credits [29]. The team is expected to begin play there in 2028 [29]. The specific instruments differ from state to state. The underlying transaction is the same: public money in exchange for the presence of a team.

The federal government contributes through the tax code. Interest on most state and local bonds is exempt from federal income tax, and teams and governments work hard to ensure that stadium bonds qualify [19]. A Brookings Institution study by Ted Gayer, Austin Drukker and Alexander Gold found that 36 of the 45 major league stadiums built or significantly renovated from 2000 to the time of their research used tax-exempt bonds, at a cost to the federal government of $3.7 billion, $3.2 billion in subsidy and $500 million in additional revenue losses to wealthy bondholders ✓ Established [20]. Bradbury, Coates and Humphreys estimate forgone federal revenue from 2000 to 2020 at $4.3 billion [18]. The bill is set to grow. Replacing the 32 major league venues that will be at least 30 years old and unrenovated in the past 15 years would, at the current median, cost $16 billion in public money, and another 34 venues that opened in the 2000s will reach 30 by 2040, implying a further $17 billion [18].

Research on the local economic effect of stadiums began in earnest in the late 1980s, when Robert Baade and Richard Dye found little evidence of any systematic relationship between building a stadium, or hosting a team, and metropolitan incomes or employment [19]. Coates and Humphreys extended that work in 1999, combining data from the NFL, Major League Baseball and the NBA, and found that the overall effect of the professional sports environment was a reduction in per capita personal income in the host metropolitan area ◈ Strong Evidence [19]. Their explanation is about where money goes. Spending at a stadium is mostly spending that residents would otherwise have made on other entertainment in the same city. Owners and players are wealthier than typical residents and save more of what they earn, and many players live outside the city for much of the year [19]. Each dollar spent at the stadium therefore produces fewer and smaller ripples through the local economy than the same dollar spent elsewhere [19].

Later studies looked for gains in the industries most likely to benefit. Coates and Humphreys examined earnings and employment by sector in 2003 and found no effect on employment or wages in most parts of the economy, including the hotels, bars and restaurants that tourism arguments emphasise ◈ Strong Evidence [19]. The only sector with higher wages was recreation, which includes the athletes themselves [19]. More recent work uses property data to detect localised effects. Gabriel Ahlfeldt and Wolfgang Maennig found that land values in Berlin were higher up to about 3,000 metres from three sports facilities, and a Columbus study by Feng and Humphreys found higher home values near two privately financed venues [19]. Coates draws the obvious conclusion: venues may raise property values or activity close to them, but the effect does not extend more than a couple of miles and certainly not across a metropolitan area [19]. Nearby owners gain; taxpayers elsewhere pay.

Mega-events, often used to justify new venues, follow the same pattern on a larger scale. Robert Baade and Victor Matheson found that the NFL’s estimates of the economic impact of a Super Bowl could be exaggerated by a factor of as much as ten, and that measured increases in local economic activity range from about $30 million to $120 million, not the hundreds of millions promoted [31]. A study by Heller and Stephenson estimated that the 2017 Super Bowl in Houston raised local hotel revenue by $44 million [18]. Houston’s 2% hotel tax on that extra revenue produced roughly $880,000, less than 0.3% of the $310 million in public money used to build NRG Stadium [18]. The Olympic Games are worse still. The Oxford Olympics Study 2024 by Bent Flyvbjerg and Alexander Budzier found that every Games since 1960 has run over budget, at an average of 159% in real terms, and estimated the cost of Paris 2024 at $8.7 billion, with an overrun of 115% ✓ Established [30].

The strongest argument for subsidies is not economic growth but civic value: the pleasure residents take in having a team, even if they never attend. Economists measure this with surveys that ask how much residents would pay to keep or attract a team. Bradbury, Coates and Humphreys compile the results of these studies for venues in Pittsburgh, Jacksonville, Portland, Indianapolis, Calgary and Edmonton, and find a consistent pattern [18]. These so-called non-use values average about 13% of total construction costs and 16% of public contributions ◈ Strong Evidence [18]. That is a real benefit, and it could justify some public funding. But it supports a contribution of roughly a sixth of what governments actually pay. On this evidence a city that funds 40% or 60% of a stadium is paying several times what its residents themselves say the team is worth to them as a civic amenity [18].

If the evidence is so clear, why do subsidies persist. Bradbury, Coates and Humphreys point to political economy rather than economics [18]. The benefits are concentrated among owners, players, construction firms and nearby landowners, while the costs are spread thinly across taxpayers. Teams can threaten relocation, and that threat works: the Chicago White Sox received $157 million in public assistance to replace Comiskey Park after threatening to leave [18]. Teams that were not seeking to move have also received large subsidies, including the Atlanta Falcons with $700 million, the New York Yankees with $293 million and the Texas Rangers with $500 million [18]. Mega-events are used as a reward. Six of the nine Super Bowls held from 2011 to 2019 were staged in stadiums that had opened within the previous four years [31]. Consultants’ impact studies, which count gross spending rather than net new activity, supply the numbers that politicians cite [31] [18].

The consensus is not entirely uncontested ⚖ Contested. In a 2025 commentary Bruce Johnson, Rodney Fort and Mark Rosentraub argued that the language of consensus overstates the case, and that governments should weigh costs and benefits project by project rather than rule out public support altogether [33]. Coates himself acknowledges disagreement over what implications to draw from the literature and whether it constitutes a consensus [19]. There is also evidence that the form of ownership matters. An older study cited in the 2024 review found that privately financed facilities were built at about 60% of the cost of public ones [18], which suggests that public money raises the price of the building as well as the bill to the taxpayer. The disagreement, however, concerns how much intangible benefit to credit and how to structure deals. It does not concern the central claim made by stadium advocates, that new venues generate net local growth. On that point the evidence is close to unanimous [17] [32].

England offers the clearest contrast. Tottenham Hotspur’s new stadium opened in April 2019 at a cost of about £1 billion, financed by the club with Goldman Sachs arranging £637 million of the funding ✓ Established [41]. Everton’s Hill Dickinson Stadium on the Liverpool waterfront opened in 2025 at a cost of about £800 million, supported by £350 million of long-term private financing, after 133 years at Goodison Park [40]. The move ended a search for a new home that had lasted more than a decade [40]. The difference is structural. In the English football pyramid a club that moved to another city would lose its league position, its history and its supporters, so the threat of relocation that drives American negotiations carries little weight. Clubs are bound to their towns, and owners who want a bigger stadium generally have to pay for it.

Germany goes further and limits who can control a club. Under the 50+1 rule, the members of a club must hold a majority of voting rights in its professional football company. In 2025 the Bundeskartellamt, the federal competition authority, concluded that the rule is fundamentally compatible with competition law because it preserves club identity and member participation, but required the league to apply it consistently [34]. The authority demanded changes to the arrangements around Bayer Leverkusen, VfL Wolfsburg, RB Leipzig and Hannover 96, where exemptions or inconsistent enforcement had weakened the rule [34]. The United Kingdom has taken a different route to the same goal. The Football Governance Act received Royal Assent on 21 July 2025, creating an Independent Football Regulator for the top five divisions of the men’s game, with powers to test owners, scrutinise club finances, block breakaway competitions and address the distribution of revenue ✓ Established [35].

Japan and Australia show other models again. In 2016 the Japan Sports Agency issued guidelines intended to turn stadiums and arenas from cost centres into profit centres, and in 2022 the government included sports venues in a programme of public-private concessions targeting ¥7 trillion through March 2032 [39]. The leading example is Es Con Field Hokkaido, a 35,000-seat ballpark that cost about ¥60 billion and is owned and operated by Nippon Ham, the food company that owns the Fighters [38]. It sits at the centre of a 32-hectare development of parks, shops, farms and restaurants [38]. In Australia, Tasmania’s entry into the Australian Football League in 2028 was made conditional on a new roofed stadium at Macquarie Point in Hobart, whose estimated cost has climbed to $945 million [37]. The state pledged $375 million, the federal government $240 million and the AFL $15 million, with about $300 million more to come from state borrowing [37]. The state’s own planning commission recommended that it should not be built [36].

The common thread is leverage. In a closed league with a fixed number of franchises, every city without a team is a potential bidder, and every team can threaten to leave ◈ Strong Evidence [18]. The relocation threat is credible because it has been carried out: four of the ten NBA, NHL and NFL relocations in the decade before 1996 were clearly linked to the tax-exempt financing of a new facility [19]. In an open European pyramid, a club’s value is tied to its place, its league position and its supporters, and moving it would destroy much of what makes it valuable. In Japan, corporate ownership changes the calculation again. Nippon Ham treats the Fighters and their ballpark as part of a broader business and has invested its own capital accordingly [38] [39]. These are not differences in the economics of stadiums. They are differences in who holds the bargaining power when a stadium is negotiated.

Outside the United States the public still pays in other ways. Host cities for the Olympic Games carry the largest risks of all, with every Games since 1960 over budget [30]. The Oxford study notes that for other types of capital investment 10% to 60% of projects typically come in on or under budget; for the Olympics the figure is zero [30]. Tasmania shows that an American-style demand can be imported wherever a league controls entry: the stadium is a condition of the licence, not a choice made on its merits [37] [36]. And where states own clubs outright, the distinction between subsidy and investment ceases to exist. Saudi Arabia’s Public Investment Fund owns 75% of four leading clubs, and the record $957 million they and their rivals spent on players in 2023 was a decision of national policy rather than a commercial bet [43]. The question in every country is not whether public money is involved in sport, but how much, on what terms, and with what share of the returns.

The international comparison suggests a simple conclusion. The large stadium subsidies common in North America are not a law of economics; they are a product of league structure and bargaining power ◈ Strong Evidence [18] [19]. Even within North America, 25 of 98 venues opened since 1990 were built with 5% or less in reported public funding [18]. Where teams cannot credibly threaten to leave, as in English football, owners have financed major stadiums themselves [41] [40]. Where governments have set clear rules on ownership, as in Germany and now the United Kingdom, they have done so without stopping clubs from competing at the highest level [34] [35]. Where a league controls entry and cities compete for the privilege of hosting, as in the NFL, the AFL or the Olympic movement, the public pays more [25] [37] [30]. Policy that wants to change the outcome has to change the bargaining position, not simply the size of the subsidy.

The federal attempt to stop tax-exempt financing of stadiums produced the opposite of its intended effect. The Tax Reform Act of 1986 prevents more than 10% of debt service on tax-exempt bonds from coming from stadium-generated revenue, which was taken to mean that ticket surcharges could not repay the debt [19]. Local governments responded by repaying it from other sources, and by 2001 more than 66% of the 99 major league facilities then operating used general revenues to cover stadium expenses [19]. In 1996 Senator Byron Dorgan complained that Congress thought it had shut down public financing of private stadiums in 1986, only to find that the only remaining healthy public housing is in sports stadiums for wealthy team owners [19]. Moynihan’s bill of the same year would have ended the exemption. In introducing it, he said: let the new stadiums be built, but not at the expense of the American taxpayer [19]. It did not pass, and neither has any successor [19].

At the state level the main dynamic is competition between jurisdictions. The Chiefs negotiation set two states within the same metropolitan area against each other, and the bid with the larger public share won [25] [27]. In Tasmania the government introduced legislation to fast-track approval of the stadium and bypass the usual planning commission process, after the costs had risen from $715 million to $945 million [37]. In each case the institutional design favours the team. Deadlines are set by the team’s lease, the alternative is presented as the loss of the franchise, and the evidence on economic impact arrives in the form of consultants’ reports commissioned by the parties that want the deal [18] [31]. Kansas’s own 2021 audit, which found that only three of 16 STAR bond projects had met their tourism goals, did not prevent the programme from being used for the largest project in its history [26]. The risks that follow are predictable, and the table below sets out the most significant, with an assessment of how likely and how costly each is on current evidence.

Better deals are possible, and the research identifies how they are structured. The first principle is to pay only for what the public receives. If non-use values average 13% of construction costs, a public contribution of that order is defensible and anything far above it is a transfer to the team [18]. The second is to count every cost. Property tax exemptions, land transfers and infrastructure should be included in the public share, because they are often larger than the headline grant [18]. The third is to share in the returns. The Minneapolis Metrodome was 90% publicly financed, yet its lease gave the public landlord such a large share of team revenue that it ranked among the better deals of its time [18]. And the fourth is to limit public money to infrastructure that serves wider purposes. The New England Patriots’ Gillette Stadium, opened in 2002, received no direct public construction funding, though the state spent $70 million on surrounding infrastructure [18].

Supporters of public funding make a case that deserves a fair hearing ⚖ Contested. The first argument is civic value: residents do place a measurable value on having a team, even if it is smaller than typical subsidies [18]. The second is urban regeneration. The District of Columbia presents its RFK deal as a development project that includes housing, green space and a sports complex, not just a stadium [28]. The third is the localised effect on property values, which is real within a mile or two of a venue [19]. The fourth is that the consensus language may be too sweeping, and that decisions should be made case by case [33]. The weight of evidence does not support the growth claims that usually accompany a deal, but these arguments do support a narrower and more honest framing. The comparison below summarises the strongest version of each side.

Beyond stadiums, three other questions in sports economics remain genuinely open. The first is how much franchises are really worth: Forbes and Sportico differ by $1.5 billion on the Cowboys alone [8] [10], and both rely on private financial data. The second is the division of revenue in baseball, where owners call their proposal an even split and the union calls it a pay cut of more than half a billion dollars [16]. The third is whether the media boom will continue ⚖ Contested. The NBA’s new contracts increased annual fees by about 2.5 times [4], but the Premier League’s latest domestic deal grew by only 4% [5], and its broadcast income rose just 2% in 2024-25 [6]. The NFL’s option to leave its contracts after the 2029 season will be the next major test of whether rights values keep climbing [3]. If they slow, the pressure to raise revenue from stadiums, tickets and the public purse will grow.

The evidence therefore narrows the policy debate rather than closing it. It is no longer credible to argue that stadiums pay for themselves through economic growth; the research on that question is about as settled as empirical economics gets [17] [32]. The real choices concern how much to pay for intangible benefits, how to account for hidden costs, whether federal taxpayers should keep underwriting local deals and how to stop jurisdictions bidding against each other [18] [19] [20]. Those are political questions, and they can be answered in different ways by different communities. But they should be answered on the basis of what the evidence shows, not on the basis of impact studies that count every dollar spent at a stadium as new money [31]. Coates noted that federal bills to end the exemption, including one introduced in 2025, have repeatedly failed to gain traction [19]. The final section draws these threads together.

Following the money produces a clear picture. Broadcasters and streamers pay roughly $10 billion a year to the NFL and nearly $7 billion a year to the NBA, and those payments are shared among clubs on terms set by the leagues [3] [4] [1]. Players receive about half of the resulting revenue through collective bargaining, a share that has been stable in football and basketball and is now under attack in baseball [12] [14] [16]. Owners receive the rest, together with the full benefit of rising franchise values, which have roughly tripled in five years in the NFL [8] [9]. Fans pay through tickets, concessions and subscriptions, with matchday costs rising sharply in both the NFL and the Premier League [42] [6]. And the public pays for a large share of the venues, through direct grants, revenue bonds, tax exemptions and forgone property taxes, at a scale that is set to rise in the coming replacement wave [18] [20].

Following the risk produces a different picture. The parties that capture most of the gains carry the least of the risk. Media income is contracted years in advance, player salaries are tied to revenue by formula and franchise values are supported by scarcity that the leagues control [3] [12] [8]. Taxpayers, by contrast, carry the risk of cost overruns, revenue bond shortfalls and venues that are retired after three decades while still standing [18] [26]. In Buffalo the club absorbed the overruns, but the public commitment remained fixed at $850 million regardless [21] [23]. In Kansas, the debt will be repaid from a sales tax district whose spending comes largely from other local businesses [26]. The result is an unusual arrangement in which the most profitable private actors in the transaction are insulated from risk, and the public, which receives the smallest measurable benefit, is exposed to most of it.

A better approach follows directly from the evidence. Public contributions should be sized to the measurable civic value of a team, which research puts at roughly 13% to 16% of costs, rather than to the team’s asking price [18]. Every element of public support, including land, infrastructure and property tax exemptions, should be counted and published before a vote [18]. Leases should give the public a share of the revenue that a publicly financed building generates, as some earlier deals did [18]. At the federal level, ending the use of tax-exempt bonds for professional sports venues would remove a subsidy that costs taxpayers across the country and benefits only the jurisdictions that win the bidding [20] [19]. And states could agree not to use public money to lure teams from neighbouring jurisdictions in the same metropolitan area, which produces costs without regional gains [25] [27].

Several developments in the next two years will show whether policy is moving. Major League Baseball’s labour agreement expires on 1 December 2026, and the outcome will determine whether the sport adopts a salary cap and how its revenue is shared [45] [16]. The Titans stadium is due to open in 2027, putting the largest public stadium commitment in US history into service [24]. The Chiefs project in Kansas will test whether STAR bonds can fund a venue on this scale [25] [26]. In Britain, the new Independent Football Regulator will begin to use its powers over club ownership and revenue distribution [35]. And the NFL’s option to leave its media contracts after the 2029 season will show whether the rights boom has further to run [3]. In Australia, the Tasmanian stadium will show what happens when a league makes a venue a condition of entry despite a negative planning assessment [36] [37].

Professional sport is a remarkable business, and much of its success is earned. Leagues have built products that hundreds of millions of people want to watch, and media companies pay accordingly [3] [4] [5]. Players have bargained for a fair share of what they produce [12] [14]. The problem lies in one specific link in the chain: the claim that public money for stadiums is an investment in local growth. On that claim the evidence of nearly four decades is consistent and clear [17] [19] [32]. Stadiums move spending around a region; they do not create much of it. Their benefits are real but modest and concentrated nearby, and their costs are large, partly hidden and spread across taxpayers who may never attend a game [18]. A city that wants a team may still decide to pay for one. It should do so knowing what it is buying, and at a price that reflects what the evidence says the team is worth.