Measuring Market Power in Professional Baseball,Basketball,Football,and Hockey

Abstract

Using Forbes magazine’s estimates of the current value and revenues of professional sports teams, we derive a long-run variant of the Lerner Index. We apply the strategy to professional teams in baseball, basketball, football, and hockey over the 2006–2019 period. All teams have positive and significant price-cost margins over the entire period. Analysis of variance shows that local market factors and past team performance have less impact on a team’s market power than do common league-wide effects. The strongest market power is in leagues with more aggressive revenue sharing policies. Price-cost margins are higher for professional teams in North American than for the most valuable European soccer teams, consistent with the stronger exemption from antitrust law in the United States and the weaker revenue sharing policies in Europe.

JEL Classifications: L43, L13, L83

Keywords

market power Lerner Index revenue sharing local markets price-cost margin antitrust

In 2018–2019, the four largest professional sports leagues generated almost US$37.4 billion in revenues with the National Hockey League (NHL) earning US$4.9 billion, the National Basketball Association (NBA) US$8 billion, Major League Baseball (MLB) US$9.9 billion, and the National Football League (NFL) US$14.5 billion. As of 2019, Forbes estimates that the teams collectively are worth US$222 billion. One reason for their substantial value is that in North America, teams benefit from legal monopoly status. While individual teams compete with one another in games, the leagues have been granted exemptions from antitrust laws and so they are allowed to restrict entry, limit the number of games, and share revenues. This level of cooperation on limited output and shared proceeds is legally binding, unlike the incentives of firms to renege on agreed pricing or production in cartels. In essence, each of the four leagues acts as a single firm with the teams treated legally as if they were subsidiaries. The league can decide where its franchises can be located, how often they will play, how many players they can employ, and how they can hire and compensate their athletes In contrast, European soccer teams can be dropped from their league if they lose too often and are in competition with teams in lower divisions seeking to win their way into the top leagues.

As monopolies, we would expect the North American professional leagues would price their product above marginal cost without risk of future lost profitability from new entry. But do the “subsidiary” individual teams have the same market power as the league or is there variation in market power across teams? If a team’s ability to price is based on its local market, its success on the field, or the loyalty of its fan base, then there may be variation in market power across the teams. On the contrary, if individual team profits are shared in common, as would be the case with pooled net revenues that were equally distributed across teams, the league market power would be the same as the individual team market power. As a result, the team’s market power will be divorced from its local market conditions and past competitive success depending on the extent of revenue sharing in the league.

This study presents a new variation on the Lerner Index computed from Forbes’ estimates of the revenues and present values of U.S. professional sports teams. Price-cost margins are commonly used as indicators of the share of the consumer price that is the mark-up over cost. We show that the ratio of the team’s market value to its annual revenue represents a long-run measure of the traditional Lerner Index. Variation in these indexes across teams within a league will show how much the league insulates member teams from their own local market conditions or their competitiveness. Variation in price-cost margins across leagues will show which leagues have the greatest market power. Variation across time will show how market power varies over the business cycle.

Our results show that all four leagues have the market power to set price above marginal cost. All the North American leagues gained market power in the recovery. Common macroeconomic and league factors across teams are the dominant factors explaining variation in market power over the 2006–2019 sample period, whereas team won–loss records, championships, and the size of the local market are less important for maintaining long-run price-cost margins. The highest price-cost margins are in the NBA and the NFL that have the strongest revenue sharing policies, whereas the smallest are in the NHL that has the weakest redistribution from richer to poorer firms.

Background

Professional sports leagues in the United States have market power because they can legally collude to limit entry. As a result, they can set the number of games, control broadcasting, restrict ticket sales, and restrict sale of licensed memorabilia. However, MLB is the only league that has a nonstatutory exemption from the Sherman Anti-Trust Act, thanks to a Supreme Court ruling that ruled baseball was not a business and therefore not subject to antitrust regulation (Federal Baseball Club vs. National League, 1922). The antitrust exemption was upheld in Toolson vs. New York Yankees, Inc. (1953). Because these rulings apply only to MLB, other professional leagues such as the NBA, NHL, and NFL are technically subject to antitrust law although there has been no definitive legal test. An argument favoring exemption from antitrust law is that these leagues are viewed as single entities with subsidiaries rather than competing businesses. As individual entities, the leagues are viewed as a single firm allocating production across its subsidiaries rather than as a group of colluding firms (Farzin, 2015).¹ The antitrust protection of the other leagues is sufficiently strong that Grow (2012) argues that even if the MLB exemption were to be reversed, its current practices are so similar to the other leagues that do not enjoy the added protection that monopoly power in MLB would be unaffected.²

Professional sports in the United States appear to have a stronger exemption from antitrust law than do professional leagues in Europe. Sports Governing Bodies (“SGBs”) such as the British Premiership, the German Bundesliga, or the Italian Serie A administer rules that could violate European antitrust laws if the rules are believed to limit economic competition. However, the European Commission and the European Court of Justice have been willing to grant that professional sports leagues regulate competition among their members for noneconomic reasons (Farzin, 2015). Nevertheless, all of the European soccer leagues have rules that force teams to place high enough in the league standings to retain their status in the top league. Teams that lose too many games are relegated to lower leagues while others are promoted from below. As a result, professional teams in Europe have less security against competition than is the case in the United States. These differences in rules regarding entry and exit may lower price-cost margins for European teams compared with teams in North America.

Method

Abba Lerner (1934) showed that profit maximizing firms facing imperfect competition will set price such that

L I = \frac{P - M C}{P} = | \frac{1}{E^{D}} |,

(1)

where the Lerner Index (LI) is shown to be the ratio of the difference between price (P) and marginal cost (MC) as a ratio of the price. That ratio, in turn, is equal to the inverse of the absolute value of the elasticity of firm demand. Without market power, p = MC, $E^{D} \to - \infty$ , and so LI = 0. The other extreme would have $L I = | E^{D} | = 1$ as would be the case if the team acted as a pure monopolist with no substitutes for its product and a marginal cost of generating revenue of 0. More generally, the elasticity of demand will be greater than 1 in absolute value as there are substitutes for the team’s services (competing entertainment events, other media offerings, alternative licensed clothing). As a result, the team’s Lerner Index will be greater than 0. The greater the Lerner Index, the greater the market power.

The Lerner Index is still valuable as a measure of market power, even if the firm does not maximize profit. A positive price-cost margin signals better than normal economic profit, even if the maximization condition in Equation 1 is violated so that the highest possible profit is not realized. In practice, as argued by Boland (1981), it is not possible to prove profit maximization, although it remains useful as a maintained hypothesis for motivating our empirical analysis.

The difficulty with applying Equation 1 to measure market power of professional sports teams is that we do not observe quantity or price. Professional sports teams are multiproduct firms producing a myriad of products, including tickets, media access, memorabilia, and advertising. Developing a market power measure will require aggregating across all of these products. A second complication is that, to a much greater extent than most firms, professional sports teams are bound by numerous extended contracts that bind them to workers, lease arrangements, and media and advertising partnerships. As a result, market power would be reflected more accurately in present value of price-cost margins rather than the short-run price-cost margin depicted in Equation 1.

Forbes magazine has been providing annual estimates of the present value, operating revenues, and total revenues of professional sports franchises. The Forbes’s measures are all estimates based on projections rather than direct access to firm financial information. The projections are based on various sources of information, including revenue sharing provisions, purchase offers, arena contracts, media experts, and team-provided information such as attendance, ticket sales, and salaries.

We can write the Forbes’s estimate of total revenue for team j in league k and year t as

T R_{j k t} = P_{j k t} . q_{j k t},

(2A)

where $P_{j k t}$ is the unobserved price and $q_{j k t}$ is the unobserved quantity produced by that team. Forbes also reports the “current value” of the franchise. We interpret that measure to be the present value of the stream of net earnings generated by the team,

P V_{j k t} = \frac{P_{j k t} . q_{j k t} - A C_{j k t} . q_{j k t}}{i},

(2B)

under the assumption that the current net earnings for year t are expected to continue in perpetuity. In Equation 2B, $A C_{j k t}$ is the average cost of providing sports services and i is the interest rate.³

We use the Forbes’s financial estimates to generate a long-run approximation of the Lerner Index for each franchise in each year. Multiply both sides of Equation 2B by an estimate of the interest rate to get $i . P V_{j k t} = P_{j k t} . q_{j k t} - A C_{j k t} . q_{j k t}$ . Then, divide both sides by Equation 2A to get a long-run measure of the price-cost margin for team j⁴:

L I_{j k t}^{L R} = i . (\frac{P V_{j k t}}{P_{j k t} . q_{j k t}}) = \frac{P_{j k t} . q_{j k t} - A C_{j k t} . q_{j k t}}{P_{j k t} . q_{j k t}} = \frac{P_{j k t} - A C_{j k t}}{P_{j k t}} .

(3)

In other words, we multiply an estimate of the interest rate to the ratio of the Forbes’s estimate of current franchise value relative to its current total revenue to generate an estimate of the Lerner Index. Because this measure incorporates the stream of future profits of the franchise, we view it as the long-run measure of the Lerner Index. This variation of the Lerner Index can be estimated despite not observing either the team’s price or quantity. This long-run measure will be more forward looking, incorporating future anticipated streams of revenue or future growth of demand for the team’s services which may not yet be captured by current revenue.

Revenue Sharing

Revenue sharing will lower the team’s dependence on its own market. Instead, the team’s market power will now reflect the extent to which the team’s revenue depends on shared rents across all teams versus its own efforts. In the limit, if all revenues and costs are equally shared by the league, there would be no variation in market power across teams. Instead, the Lerner Index would be identical across all teams. Consequently, an indicator of the strength of the league’s revenue sharing agreement will be if a team’s market power is independent of its local market or past competitive success. Evidence supporting that conjecture was presented by Berri et al. (2015) who showed that team performance was only weakly tied to team revenues. Winning 10% more games led to only 2.7% more revenue in baseball and basketball and less than 1% in football. Profit maximizing teams will lower salaries if the link between player salary and profit decreases, and as a result, profits for the league as a while rise. However, as Vrooman (2000, 2009) explains, revenue sharing can lead to higher profit overall if team owners want to maximize profit and not won–loss record.⁵

While all leagues use revenue sharing, the details differ. In the NFL, gate receipts are split 60% to the local team and 40% shared. Licensed deals (such as merchandizes & jerseys) and TV revenue are shared equally among all teams (Berri et al., 2015). There is less revenue sharing in the MLB. In the 2012–2016 Basic Agreement between the 30 Major League Clubs and the MLB (2012) Players’ Association, teams send 34% of each team’s net revenue to the league, and these proceeds are shared equally among the teams. That means that higher earning clubs subsidize the lower earning clubs.

Perhaps the most aggressive revenue sharing plan is the one initiated by the NBA starting with the 2013–2014 season. The plan calls for all teams to contribute roughly 50% of their annual revenue after expenses into a pool. Each team then receives an allocation equal to the league’s average team payroll for that season from the revenue pool (Lombardo, 2012).

Hockey has the least aggressive revenue sharing program. As laid out in the NHL Players’ Association (2012) collective bargaining agreement, the Redistribution Commitment is only 6% of the league-wide hockey revenues. The richest 10 teams contribute 50% of the total based on the gap between their revenues and the revenues of the 11th richest team. Added to this is 35% of the gate revenue from the Stanley Cup Playoffs. These funds are then reallocated with the poorest teams getting the largest shares from the shared pool.

Evaluating the relative degree of redistribution, clearly, the NHL has the least redistributive and the NBA and the NFL have the most redistributive policies. As the degree of redistribution through revenue sharing increases, the team’s dependence on its own revenues diminishes. Consequently, we would expect that team market power would be most tied to the local market conditions and team performance in the NHL. In the NBA and the NFL, the team’s market power will be tied more to the league market power rather than local market conditions.

Data

We use the Forbes’s estimates of MLB, NBA, NFL, and NHL team revenues, costs, and present values from 2006 through 2019. The MLB has 30 teams, the NBA has 30 teams, the NFL has 32 teams, and the NHL has 30 teams. Consequently, our data set includes 420 observations from baseball, basketball, and hockey and 448 observations from football. Our long-run measure, $L I_{j k t}^{L R}$ , defined by Equation 3, multiplies an assumed interest rate of 0.03 times the ratio of the Forbes’s estimate of the team’s total revenue relative to its current value. Our findings are not sensitive to the choice of alternative interest rates.

Figures 1 to 4 show how the cumulative distribution of team long-run market power changes over time. We present the distributions for 2006, 2011, and 2019. Our graphs illustrate the median of each distribution. Across all four leagues, there was a consistent upward trend in our long-run Lerner Index after 2011. From the first to last period, the median price-cost margin rises from 7% to 15% in MLB; 8.5% to 19% in the NBA; 14.5% to 19% in the NFL; and 7% to 9% in the NHL. Three of the leagues demonstrated little erosion of their market power during the Great Recession. Only the NFL had a decline in market power between 2006 and 2011, albeit from the highest price-cost margins among the four leagues at the time. Since 2011, the NBA has gained the most market power.

Figure 1.

Cumulative distributions of long-run price-cost margins for Major League Baseball, 2006 2011, 2019.

Figure 2.

Cumulative distributions of long-run price-cost margins for the National Basketball Association, 2006, 2011, 2019.

Figure 3.

Cumulative distributions of long-run price-cost margins for the National Football League, 2006, 2011, 2019.

Figure 4.

Cumulative distributions of long-run price-cost margins for the National Hockey League, 2006, 2011, 2019.

Explanatory Variables

We selected variables that were commonly used to explain variation in team value in prior studies.⁶ Table 1 presents the summary statistics for the city and team factors we use to explain variation in the Lerner Indexes. We divide our measures into three groups: attributes of the team, the city, and the league. Our aim is to assess the relative importance of these factors in shaping a team’s market power and to examine their relative importance under more or less egalitarian redistribution policies.

Table 1.

Summary Statistics.

Team	MLB	NBA	NFL	NHL
$L I_{j k t}^{L R}$ Lerner Index using present value	0.102	0.143	0.149	0.093
$L I_{j k t}^{L R}$ Lerner Index using present value	(0.039)	(0.062)	(0.031)	(0.034)
Championships	3.69	2.40	1.69	3.13
Championships	(5.21)	(4.14)	(1.88)	(5.04)
Winning percentage previous season	0.50	0.50	0.50	0.485
Winning percentage previous season	(0.68)	(0.153)	(0.19)	(0.103)
Google searches: Key word team name	19.42	16.49	21.25	13.86
Google searches: Key word team name	(11.12)	(11.61)	(11.12)	(7.61)
Metro market
ln(population)	15.37	15.17	15.03	15.14
ln(population)	(0.65)	(0.79)	(0.80)	(0.90)
ln(per capita income)	10.92	10.86	10.90	10.84
ln(per capita income)	(0.18)	(0.18)	(0.17)	(0.21)
League
Google searches: Key word team name	26.58	22.40	27.61	26.34
Google searches: Key word team name	(0.99)	(11.03)	(9.32)	(9.08)
Year fixed effects
City fixed effects
Team fixed effects
N	420	420	448	420
Years	2006–2019	2006–2019	2006–2019	2006–2019

Note. MLB = Major League Baseball; NBA = National Basketball Association; NFL = National Football League; NHL = National Hockey League.

Team traditions will be largely controlled by team fixed effects. We do control for the number of championships and the team’s winning percentage in the previous season.⁷ These data were compiled from available online league records. The strength of interest in each team was measured by an annual index of the number of Google searches on the team name, compiled in Google Trends.

The strength of the local market is based on local population and per capita income, and the traditional factors used to measure a local market’s ability to pull in consumer dollars.⁸ We control for macroeconomic shocks using annual dummy variables. The time period covered includes the Great Recession and subsequent recovery which should illustrate how each league fares in contractions and expansions. In our pooled league equations, we add controls for the city fixed effects to capture the remaining unmeasured persistent strengths in the local support for professional sports teams. We include a dummy variable for Canadian teams to correct for differences in currency valuations. For U.S. cities, the population and per capita income measures come from the Department of Commerce, Bureau of Economic Analysis. The Canadian data were obtained from Statistics Canada.

Our last set of measures is at the league level. These are largely captured by league dummy variables, but we also include annual indexes of Google searches on the league as opposed to the teams themselves. Consistently, interest in the league exceeds interest in individual teams, as indicated by relative search intensity. This will allow us to disentangle interest in the sport versus interest in the team.

Table 2 shows the average long-run price-cost margins for each of the 122 teams in our sample. Every team has a positive price-cost margin in every year of the sample period. In general, the largest price-cost margins are in the largest cities (New York, Los Angeles, Chicago), whereas the smallest margins are in the smallest markets (Kansas City, Columbus, Raleigh).

Table 2.

Estimated Long-Run Average Price-Cost Margin for U.S. Professional Sports Teams, 2006–2019.

Baseball		Basketball		Football		Hockey
Team	$L I_{j k t}^{L R}$	Team	$L I_{j k t}^{L R}$	Team	$L I_{j k t}^{L R}$	Team	$L I_{j k t}^{L R}$
New York Yankees	0.152	Boston Celtics	0.197	New York Giants	0.165	Toronto Maples Leafs	0.133
Los Angeles Dodgers	0.139	Chicago Bulls	0.192	Chicago Bears	0.164	New York Rangers	0.120
Chicago Cubs	0.129	Los Angeles Clippers	0.186	Dallas Cowboys	0.163	Montreal Canadiens	0.112
Boston Red Sox	0.129	Los Angeles Lakers	0.179	Minnesota Vikings	0.162	Chicago Blackhawks	0.105
New York Mets	0.127	New York Knicks	0.174	Washington Redskins	0.161	Boston Bruins	0.104
San Francisco Giants	0.115	Golden State Warriors	0.168	New England Patriots	0.161	Philadelphia Flyers	0.103
St. Louis Cardinals	0.110	Brooklyn Nets	0.168	Miami Dolphins	0.161	Vancouver Canucks	0.102
Los Angeles Angels	0.107	Dallas Mavericks	0.159	Philadelphia Eagles	0.160	Detroit Red Wings	0.101
Philadelphia Phillies	0.106	Houston Rockets	0.146	Indianapolis Colts	0.159	Washington Capitals	0.091
Texas Rangers	0.104	Miami Heat	0.144	Pittsburgh Steelers	0.157	Los Angeles Kings	0.091
Washington Nationals	0.104	Sacramento Kings	0.142	Houston Texans	0.157	Dallas Stars	0.090
Chicago White Sox	0.104	Washington Wizards	0.137	New York Jets	0.157	Colorado Avalanche	0.089
Seattle Mariners	0.101	Phoenix Suns	0.136	San Francisco 49ers	0.154	Pittsburgh Penguins	0.087
Atlanta Braves	0.100	San Antonio Spurs	0.136	Seattle Seahawks	0.153	Edmonton Oilers	0.085
Houston Astros	0.100	Orlando Magic	0.134	Baltimore Ravens	0.152	San Jose Sharks	0.084
Baltimore Orioles	0.098	Philadelphia 76ers	0.133	Denver Broncos	0.151	New York Islanders	0.083
San Diego Padres	0.096	Portland Trailblazers	0.133	Green Bay Packers	0.150	Calgary Flames	0.081
Arizona Diamondbacks	0.094	Oklahoma City Thunder	0.132	Oakland Raiders	0.143	Anaheim Ducks	0.080
Toronto Blue Jays	0.093	Denver Nuggets	0.130	Carolina Panthers	0.143	Minnesota Wild	0.079
Pittsburgh Pirates	0.091	Indiana Pacers	0.130	New Orleans Saints	0.141	Winnipeg Jets	0.078
Minnesota Twins	0.090	Utah Jazz	0.129	Kansas City Chiefs	0.141	Ottawa Senators	0.078
Miami Marlins	0.090	Toronto Raptors	0.128	San Diego Chargers	0.141	New Jersey Devils	0.076
Detroit Tigers	0.090	Milwaukee Bucks	0.127	Tennessee Titans	0.141	Buffalo Sabers	0.071
Cleveland Indians	0.088	Atlanta Hawks	0.125	Atlanta Falcons	0.141	Florida Panthers	0.068
Colorado Rockies	0.087	Charlotte Hornets	0.122	St Louis Rams	0.140	Nashville Predators	0.068
Cincinnati Reds	0.086	Memphis Grizzlies	0.122	Tampa Bay Buccaneers	0.140	Arizona Coyotes	0.067
Oakland Athletics	0.085	Minnesota Timberwolves	0.120	Arizona Cardinals	0.140	Tampa Bay Lightning	0.067
Milwaukee Brewers	0.085	Detroit Pistons	0.119	Cleveland Browns	0.137	Carolina Hurricanes	0.067
Kansas City Royals	0.084	Cleveland Cavs	0.118	Cincinnati Bengals	0.133	St Louis Blues	0.066
Tampa Bay Rays	0.082	New Orleans Pelicans	0.116	Detroit Lions	0.132	Columbus Blue Jackets	0.065
				Jacksonville Jaguars	0.130
				Buffalo Bills	0.123
Average	0.102	Average	0.143	Average	0.149	Average	0.093

Note. Long-run measures are computed using Equation 3, $L I_{j k}^{L R} = i . (P V_{j k t} / (P_{j k t} . q_{j k t}))$ .

Analysis of Variance (ANOVA)

We employ an ANOVA to identify how much of the variation in price-cost margins are attributable to the fixed attributes of the team (its city, its competitive tradition, its tie to its fan base); how much is due to common changes across all teams (macroeconomic shocks, league-wide effects); and how much is due to remaining time-varying factors specific to the team. The analysis is set up as follows:

L I_{j k t}^{L R} = \sum_{j = 1}^{J} α_{k j} C_{j} + \sum_{t = 1}^{T} β_{t k} T_{t} + ε_{i j t},

(4)

where the dependent variable is the team’s long-run market power measure at time t, $C_{j}$ is a dummy variable designating team j, $T_{t}$ is a dummy variable indicating year t, and $α_{k j}$ and $β_{t j}$ are coefficients measuring the team-specific and year-specific effects. We can then assess the share of the variance that is due to variation across teams that does not change over time (between effect); variation across time that is common to all teams due to league-wide factors or macroeconomic shocks (within effect); and to other idiosyncratic factors, $ε_{i j t}$ .

The results from the ANOVA are in Table 3. The dominant source of variation in baseball, basketball, and football is common year and league effects. Permanent team effects only explain 13% to 20% of the variation in market power in these three leagues. When we add time-varying variables that should capture changes in year-to-year local demand for professional sports, the added explanatory power is extremely small and only statistically significant for professional basketball.

Table 3.

Variance Decomposition of Lerner Index Measures in Common Within Effects, Team Effects, and Unexplained Idiosyncratic Effects.

Source of variation	Baseball	Basketball	Football	Hockey
Within: Year and league effects	0.72	0.71	0.67	0.47
Between: Team effects	0.20	0.14	0.13	0.33
Total explained	0.92	0.85	0.80	0.81
Idiosyncratic	0.08	0.15	0.20	0.19
Added heterogeneity controls	0.00	0.02*	0.00	0.00
N	420	420	448	420

Note. Additional controls include annual measures of log per capita local income, the log of local population, the team’s winning percentage, and an index of Google searches for the team, all in the year preceding the measured Lerner Index.

*Significance at the 0.05 level.

Professional hockey is very different. Year and league effects only explain 47% of the variation in team market power. Team fixed factors explain 33% the variation, whereas time-varying team effects explain an additional 19%. While the remaining time-varying factors are not the controls that we included in the regression, our results suggest that the local economy and community are more important to NHL teams, consistent with the league’s relatively weak revenue sharing program.⁹ In the leagues with the stronger revenue sharing policies, common factors due to league and macroeconomic shocks dominate in explaining team market power.

Pooled League Analysis

To examine the importance of the league in setting a team’s market power, we pool the data from all four leagues, and then examine how various factors differ in importance across the leagues. To make the strategy clear,

L I_{j k m t}^{L R} = \sum_{k = 1}^{4} γ_{k} L_{k} + \sum_{k = 1}^{4} \sum_{i = 1}^{I} θ_{i k} L_{k} X_{j k t} + \sum_{k = 1}^{4} γ_{k} L_{k} G_{k t} + ε_{m} + ε_{t} + ε_{m k j t} .

(5)

The coefficients $γ_{k}$ will provide the league effects conditional on the remaining controls, $X_{j k t}$ . These will include local log population and log per capita income, the team’s total championships, and the team’s Google searches. We also include a measure of the strength of demand for the league, indexed by the number of Google searches on the league. We also include controls for fixed municipal ( $ε_{m}$ ) and time $(ε_{t})$ effects.

The estimation is reported in Table 4. In all regressions, hockey is used as the reference sport. The first column constrains all fixed effects and covariates to zero coefficients, and so we only estimate the four $γ_{k}$ terms. These effectively recreate the unconstrained mean values of $L I_{j k t}^{L R}$ in Table 1. The NHL teams have a mean price-cost margin of 0.093, whereas MLB averages 0.102 (0.093 + 0.009); the NBA averages 0.143; and the NFL averages 0.149. Coefficients in the remaining columns should be interpreted similarly, where the coefficients on the noninteracted term represent the value for the NHL and the coefficients on the interacted terms represent the difference between that league’s value and the NHL.

Table 4.

Pooled League Regressions Identifying Factors That Affect Team Long-Run Lerner Indexes, 2006–2019.

	(A)	(B)	(C)
MLB	0.009*	0.01*	−0.578*
MLB	(1.94)	(2.24)	(2.62)
NBA	0.05*	0.054*	−1.114*
NBA	(8.97)	(10.17)	(3.80)
NFL	0.056*	0.061*	−0.267
NFL	(13.6)	(14.9)	(1.16)
ln(population)^a			−0.02
ln(population)^a			(0.52)
ln(per capita income)^a			0.035
ln(per capita income)^a			(1.09)
Championships^a			0.004*
Championships^a			(6.58)
Google League Search			0.0007*
Google League Search			(2.71)
Google Team Search^a			−0.0006*
Google Team Search^a			(2.41)
Interactions
ln(population)^a
MLB			0.008
MLB			(1.51)
NBA			0.013*
NBA			(2.27)
NFL			0.002
NFL			(0.47)
ln(per capita income)^a
MLB			0.043*
MLB			(2.20)
NBA			0.09*
NBA			(3.27)
NFL			0.028
NFL			(1.30)
Championships^a
MLB			−0.003*
MLB			(3.70)
NBA			−0.002*
NBA			(2.81)
NFL			−0.002
NFL			(1.90)
Google Team Searches^a
MLB			0.0004
MLB			(1.53)
NBA			0.001*
NBA			(2.80)
NFL			<0.0001
NFL			(0.31)
Constant	0.093*	0.052*	−0.023
Constant	(26.3)	(8.45)	(0.03)
Year fixed effects		Yes	Yes
City fixed effects		Yes	Yes
R²	.24	.72	.78
N	1,708	1,708	1,708

Note. All estimation corrected for clustering at the individual team level. T-statistics reported in parentheses. MLB = Major League Baseball; NBA = National Basketball Association; NFL = National Football League.

Joint significance at the .05 level that the four terms including the factor.

Significance at the .05 level.

In Column B, we add in the controls for the fixed effects, $ε_{m} and ε_{t}$ . The constant term which represents the value of $L I_{j k t}^{L R}$ conditional on controls for city and time fixed effects drops to 0.052, and so 44% of the price-cost margin in hockey is attributable to the city and year effects. The interacted terms hardly changed, and so the numerical decrease in estimated price-cost margins is similar across all four leagues. Additional investigation showed that the year effects were the largest contributor to the decrease in estimated price-cost margins, and so common league and macroeconomic effects raise the price-cost margins on average by about 0.04 across all leagues.

The year dummies show a general increase in price-cost margins of 0.07 over the sample period, a rise of 57% relative to the mean. All of the gains occurred during the recovery from the Great Recession. The pattern of results suggests that market power in professional sports rises procyclically.

In Column C, we include the full specification outlined in Equation 5. All the joint exclusion tests on the individual controls, $X_{j k t}$ , were rejected,¹⁰ meaning that the effects of these factors differ by league. Comparing across the four parameters attached to each factor, local population only raises market power for teams in the NBA and does not affect price-cost margins in any other league. Local per capita incomes affect market power in MLB and the NBA. When we combine the effects of income and population, the two factors commonly used to define the size of the local market for retail sales, the impact on the long-run Lerner Index is only statistically significant in the NBA. In the other leagues, if there is an effect on market power, it is shared across all teams.

Teams with more championships have additional market power in Hockey. Having more championships does not affect team market power in any of the other leagues that have more aggressive revenue sharing of ticket receipts. Greater interest in the individual team, as indicated by more Google searches, has negligible or even negative effects on market power. In contrast, more searches on the league significantly raise the market power for all the teams in the league.

The general conclusion from Table 4 is that team won–loss success and local market strength are only marginally important to the team’s price-cost margin. It may seem strange that so little of the variation in firm market power is explained by the strength of the local market or past team wins and losses when these variables were so important in explaining the value of European soccer teams (Scelles et al., 2016) or North American professional team’s values (Alexander & Kern, 2004; Scelles et al., 2013). However, the present value of the team does not, by itself, imply market power. Teams in small markets may have lower value and still have market power to price above marginal cost in their limited market. Alexander (2001) found that baseball teams in some of the smallest markets (Cincinnati, Kansas City, Milwaukee, Minnesota) had relatively few substitutes for the entertainment dollar, and so they could charge relatively high ticket prices. Nevertheless, because revenues are partially redistributed to other teams in the league, it is the demand for the league more than the demand for the team that drives market power for the teams in the league, consistent with our finding in Table 3 that common factors are more important than team factors in explaining the variation in team long-run Lerner Indexes.

We would expect the supply of substitutes for sports entertainment to affect local team market power. We can assess the market power of professional sports in the city by comparing the magnitudes of the estimated city fixed effects, $ε_{m}$ . The estimates in Column B incorporate the effects of population, income, fan interest, and winning tradition. We refer to these as the unconditional measures of city market power. The estimates from the Column C specification are net of these effects, and so we call them the conditional estimates. We report the largest and smallest city long-run Lerner Indexes in Table 5. The complete listing is provided in the Appendix. The conditional estimates should reflect cities that, for unobservable reasons, are willing to pay a premium over cost for the services of the team. The cities with the strongest market power include both large and small market teams, and so Toronto and Vancouver have among the largest price-cost margins. These cities also rank high in the unconditional estimates. There are more surprising results at the other tail. Despite their rabid fan base, Green Bay has the lowest conditional price margin. Montreal ranks highest in unconditional price margin, but after controlling for victories, income, and population, it ranks among the lowest, a pattern roughly matched by San Francisco. That many of these teams are so financially successful is testimony to the market power of their league and its willingness to redistribute from the most profitable markets, making up for the weaknesses in their locational market power.

Table 5.

Largest and Smallest Contributions of a Municipality to the Market Power of Its Professional Teams, 2006–2019.

Cities with the most professional sports market power
Market power measure
City	Conditional		Unconditional
City	Measure	Rank	Measure	Rank
Toronto	0.0294	1	0.0209	7
Los Angeles	0.0269	2	0.0294	5
Chicago	0.0239	3	0.0339	3
Orlando	0.0200	4	−0.0002	32
Vancouver	0.0185	5	0.0282	6
Cities with the least professional sports market power
Market power measure
City	Conditional		Unconditional
City	Measure	Rank	Measure	Rank
Green Bay	−0.0489	49	0.0084	17
Montreal	−0.0476	48	0.0419	1
Jacksonville	−0.0422	47	−0.0123	47
Raleigh	−0.0400	46	−0.0051	39
San Francisco	−0.0394	45	0.0135	10

Note. The unconditional measure is based on estimates of $ε_{m}$ in Column B of Table 4, controlling only for year fixed effects. The conditional measure is based on estimates of $ε_{m}$ in Column C of Table 4 which includes controls for log population, log per capita income, Google search intensity, and the number of championships.

Comparing Market Power in North American and European Professional Sports

To illustrate the relative market power of the various professional leagues, Figure 5 presents the cumulative distribution of our long-run Lerner Indexes for the four North American Sports Leagues and comparable measures for the 30 most successful European soccer teams. All estimates are for 2015, the latest year for which we had consistent Forbes’s data for Europe and North America. All four North American leagues have price-cost margin distributions that lie to the right of the European teams.

Figure 5.

Cumulative distributions of long-run price-cost margins for five professional sports, 2015.

The higher price-cost margins for the MLB, NBA, NFL, and NHL are consistent with their presumed greater protection against antitrust laws. North American teams are protected from new entrants and do not face the possibility of forced exit through relegation.¹¹ But the European leagues have also not used revenue sharing beyond shared television revenues.¹² Our findings are consistent with the conclusion that both antitrust exemptions and revenue sharing would lead to higher long-run Lerner Indexes in the North American sports leagues.

Conclusion

Using estimates of the value of professional sports teams and of annual revenues provided by Forbes magazine, we generate measures of long-run Lerner indexes for baseball, basketball, football, and hockey teams over the 2006–2019 period. The estimates show that all teams and leagues have market power that rises with economic expansion. Market power is attributable more to common factors affecting all teams within a league and less to individual team athletic success or the size of its local market. The largest Lerner indexes are in the NBA and the NFL that have the most aggressive revenue sharing, whereas the smallest are in the NHL that has the least comprehensive revenue redistribution policy. North American teams have more market power than European soccer teams, which we attribute to the ability of North American leagues to restrict entry due to their exemption from antitrust laws and to the weaker revenue sharing in Europe.

Footnotes

Appendix

Largest and Smallest Contribution of a Municipality to the Market Power of its Professional Teams, 2006–2019.

City	Conditional		Unconditional
City	Measure	Rank	Measure	Rank
Atlanta	Reference	15	Reference	31
Baltimore	−0.0175	29	0.0086	15
Boston	−0.0062	20	0.0383	2
Buffalo	−0.0324	43	−0.0127	48
Calgary	−0.0225	34	0.0054	21
Charlotte	−0.0218	33	−0.0056	40
Chicago	0.0239	3	0.0339	3
Cincinnati	−0.0250	36	−0.0072	43
Cleveland	−0.0153	27	−0.0083	44
Columbus	−0.0302	41	−0.0105	46
Dallas–Fort Worth	0.0163	7	0.0185	8
Denver	−0.0167	28	0.0042	23
Detroit	−0.0104	22	0.0008	29
Edmonton	−0.0294	40	0.0100	14
Green Bay	−0.0489	49	0.0084	17
Houston	0.0043	10	0.0119	12
Indianapolis	−0.0055	18	0.0064	20
Jacksonville	−0.0422	47	−0.0123	47
Kansas City	−0.0278	38	−0.0039	36
Los Angeles	0.0269	2	0.0294	5
Memphis	−0.0042	17	−0.0138	49
Miami	0.0042	11	0.0051	22
Milwaukee	−0.0215	32	−0.0070	42
Minneapolis–Saint Paul	−0.0149	26	0.0024	28
Montreal	−0.0476	48	0.0419	1
Nashville	−0.0326	44	−0.0041	37
New Orleans	−0.0290	39	−0.0095	45
New York City	0.0181	6	0.0304	4
Oklahoma City	0.0001	14	−0.0027	35
Orlando	0.0200	4	−0.0002	32
Ottawa	−0.0230	35	0.0025	26
Philadelphia	0.0023	12	0.0150	9
Phoenix	0.0058	9	−0.0013	34
Pittsburgh	−0.0189	31	0.0085	16
Portland	−0.0063	21	−0.0007	33
Raleigh	−0.0400	46	−0.0051	39
Sacramento	−0.0026	16	0.0077	18
Salt Lake City	0.0006	13	−0.0049	38
San Antonio	0.0133	8	0.0028	25
San Diego	−0.0057	19	0.0024	27
San Francisco	−0.0387	45	0.0135	10
San Jose	−0.0303	42	0.0070	19
Seattle	−0.0144	25	0.0110	13
St. Louis	−0.0181	30	0.0031	24
Tampa Bay	−0.0136	24	−0.0060	41
Toronto	0.0294	1	0.0209	7
Vancouver	0.0185	5	0.0282	6
Washington, D.C.	−0.0135	23	0.0131	11
Winnipeg	−0.0259	37	0.0007	30

Note. The unconditional measure is based on estimates of $ε_{m}$ in Column B of Table 4, controlling only for year fixed effects. The conditional measure is based on estimates of $ε_{m}$ in Column C of Table 4 which includes controls for log population, log per capita income, Google search intensity, and the number of championships.

Acknowledgements

We are grateful to the editor and referees for excellent advice in improving the article. We thank Katherine Lacy for her help mentoring the undergraduate research seminar. David Boswell and Zachary Graves helped with data collection.

Authors’ Note

Jing Ru Tan is now affiliated with Fitbit, Inc., San Francisco, CA, USA.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Peter F. Orazem

Notes

Author Biographies

Gerald T. Healy III is an alumnus of Iowa State University. He majored in economics and minored in general business. He hopes to pursue a higher degree in economics in the future.

Jing Ru Tan is from Penang, Malaysia. He graduated from Iowa State University with a BS in economics. He works as an analyst with Fitbit in San Francisco with a focus on managing Fitbit’s enterprise resource planning system.

Peter F. Orazem is university professor of economics at Iowa State University where he has taught since 1982. He is director of the program for the Study of Midwest Markets and Entrepreneurship. His specializations are in labor economics, economic development, and applied econometrics.

References

Alexander

D. L.

(2001). Major League Baseball: Monopoly pricing and profit-maximizing behavior. Journal of Sports Economics, 2(4), 341–355.

Alexander

D. L.

Kern

(2004). The economic determinants of professional sports franchise values. Journal of Sports Economics, 5(1), 51–66.

Berri

D. J.

Leeds

M. A.

von Allmen

(2015). Salary determination in the presence of fixed revenues. International Journal of Sport Finance, 10(1), 5–25.

Boland

L. A.

(1981). On the futility of criticizing the neoclassical maximization hypothesis. The American Economic Review, 71(5), 1031–1036.

Brook

S. L.

Fenn

A. J.

(2008). Market power in the National Football League. International Journal of Sport Finance, 3(4), 239–244.

Che

X. G.

Humphreys

B. R.

(2015). Competition between sports leagues: Theory and evidence on rival league formation in North America. Review of Industrial Organization, 46(2), 127–143.

Darling

Tubene

(1996). Determining the population threshold of minor trade centers: A benchmark study of non-metropolitan cities in Kansas. Review of Agricultural Economics, 18(1), 95–101.

Farzin

(2015). On the antitrust exemption for professional sports in the United States and Europe. Jeffrey S. Moorad Sports Law Journal, 22(1), 75–108.

Federal Baseball Club v. National League, 259 U.S. 200. (1922). Justia. Retrieved July 4, 2020, from https://supreme.justia.com/cases/federal/us/259/200/

10.

Gale

H. F.

(1996). Retail sales pull factors in U.S. counties. The Review of Regional Studies, 26(2), 177–195.

11.

Grow

(2012). In defense of baseball’s antitrust exemption. American Business Law Journal, 49(2), 211–273.

12.

Gruidl

J. J.

Andrianacos

(1994). Determinants of rural retail trade: A case study of Illinois. The Review of Regional Studies, 24(1), 103–118.

13.

Lerner

A. P.

(1934). The concept of monopoly and the measurement of monopoly power. Review of Economic Studies, 1(3), 157–175.

14.

Lombardo

(2012, January 23). Inside NBA’s revenue sharing. Street & Smith’s Sports Business Journal. https://www.sportsbusinessdaily.com/Journal/Issues/2012/01/23/Leagues-and-Governing-Bodies/NBA-revenue.aspx

15.

Major League Baseball. (2012). 2012–2016 basic agreement. https://ipmall.law.unh.edu/sites/default/files/hosted_resources/SportsEntLaw_Institute/2012MLB_MLBPA_CBA.pdf

16.

National Hockey League Players’ Association. (2012). Collective bargaining agreement between National Hockey League and the National Hockey League Players’ Association, September 16, 2012—September 15, 2022. http://www.chlclassaction.com/wp-content/uploads/2016/07/1B.-Selected-pages-from-the-Collective-Bargaining-Agreement-between-the-NHL-and-its-players.pdf

17.

Noll

R. G.

(2003). The organization of sports leagues. Oxford Review of Economic Policy, 19(4), 530–551.

18.

Scelles

Helleu

Durand

Bonnal

(2013). Determinants of professional sports firm values in the United States and Europe: A comparison between sports over the period 2004–2011. International Journal of Sport Finance, 8(4), 280–293.

19.

Scelles

Helleu

Durand

Bonnal

(2016). Professional sports firm values: Bringing new determinants to the foreground? A study of European soccer, 2005–2013. Journal of Sports Economics, 17(7), 688–715.

20.

Toolson vs. New York Yankees, Inc., 346 U.S. 356. (1953). Justia. Retrieved July 4, 2020, from https://supreme.justia.com/cases/federal/us/346/356/

21.

Vrooman

(2000). The economics of American sports leagues. Scottish Journal of Political Economy, 47(4), 364–398.

22.

Vrooman

(2009). Theory of the perfect game: Competitive balance in monopoly sports leagues. Review of Industrial Organization, 34(1), 5–44.