MLB teams often proclaim that attendance affects team payroll. In recent years, fans have heard everything, from Rays’ owner Stuart Sternberg saying lower attendance will decrease payroll, to Pirates team President Frank Coonelly indicating that the team’s increased attendance in 2011 would add additional financial “flexibility.” But is this relationship actually true? In this article, I analyze the extent to which attendance affects payroll.
To analyze this relationship, I compare changes in a team’s attendance to changes in a team’s payroll using data from 2009-2013. To control for year-to-year league-wide fluctuations in attendance and payroll, I compute the average attendance and payroll figures for each season and divide each team’s figures by that average. (For example, if the average payroll in 2010 were $50M and a team had a payroll of $75M, the value used would be 75/50 or 1.5.)
To focus on the crux of the relationship, I compute the yearly change in each team’s payroll and attendance figure. I do this to create a fixed-effects relationship. The ensuing analysis is looking first at how each team fluctuates individually before aggregating the results from each team into the larger MLB picture. In other words, I am not comparing the Orioles to the Mets, but rather I am only comparing the Orioles to the Orioles and so on.
Additionally when making these comparisons, I employ a one-year lag on the payroll figure. A team’s payroll is generally determined before the season begins, so if this relationship holds, the effect of an attendance increase should be seen in the following season’s payroll. This is the relationship MLB management have implied as well.
Each MLB team has four observations in my data (even though the data span five years) because I am looking at changes in values.
To begin, here is a graph depicting the general correlation among MLB teams:
The correlation is 0.2821, which suggests there may be some relationship. Here is the league-wide aggregate of this data, which further demonstrates some direct-correlation:
To get a better picture, I disaggregate the data by team.
Below is a ranking of the absolute value of the difference between a change in attendance and the subsequent change in payroll. Each team’s figure is an average of each team’s team-season values from 2009 to 2013. In short, this is a ranking of how closely teams seem to follow this proposed relationship from 2009 to 2013.
The Marlins certainly stand out (I will discuss their situation shortly). Otherwise, it is clear that there is a range in variance.
Digging deeper, the main emphasis of the proposed relationship is that a large change in attendance will lead to a large change in payroll. As the data demonstrate, many teams have stable attendance and payroll figures (such as Atlanta, Boston, Washington and so on).
Thus, to really focus on the crux of this relationship and evaluate it fairly, I analyze the same set of data only for team-season observations where the absolute value of a team’s change in attendance was at least 0.15 according to my index. I focus on this set of team-season observations because the heart of the issue revolves around the following question: Will a large attendance spike or fall directly impact payroll the following season? In order to analyze this, I should focus on teams that actually had attendance spikes or falls.
The number of teams that experienced this high level of attendance fluctuation is 16 (out of 120 observations, or about 13% of the observations). If there is a relationship, then it should still be apparent in this smaller, more refined set of data.
The list of these team-season occurrences is below, ranked in descending order of the magnitude of their attendance changes.
|Team||Year (of Attendance Change)||Change in Payroll (Following Season)||Change in Attendance||Absolute Value of Difference|
|Los Angeles (N)||2011||-0.15||-0.27||0.12|
|New York (N)||2010||-0.17||-0.21||0.04|
|Los Angeles (N)||2013||-0.05||0.17||0.22|
Among these teams, Miami stands out. In 2012, management drastically increased team payroll as the team was coming into its new stadium, Marlins Park. However, the previous season saw no increase in attendance, suggesting that other factors can affect spending. The following season (2012), the team saw a large spike in attendance because of the new stadium and higher payroll. However, in the 2013 season, the team unloaded its top players despite 2012’s attendance increase.
In short, the Miami example demonstrates how many factors can influence both attendance and payroll, and that payroll can have an impact on attendance (an issue of endogeneity, or simply the relationship reversed). Still, fans commonly hear that attendance will impact payroll, so lets look at the data closer. Perhaps Miami is just an exception.
To get a better idea of how well correlated these two variables are for this smaller set of teams, I aggregate the data again.
*Test 2 includes team-season observations which had at least a .15 change in attendance.
The relationship appears to weaken, but a correlation still exists (.1654 to be exact). While the correlation here is substantially less than in the larger data sample, a relationship still appears somewhat viable.
So, when management says attendance will impact payroll, are they being truthful? My analysis suggests that management may be implying a correct relationship. At the very least, the data from 2009 to 2013 do not discount the possibility that attendance can impact payroll.
Still, I would argue that attendance is likely one of many factors playing a role. Thus, it may have a small effect on the margin or under certain circumstances. But as the Marlins’ example demonstrates, many other factors impact payroll as well. Thus, any claims by team management about the relationship between attendance and payroll should be analyzed in a case-by-case basis, perhaps by looking at how closely the team under its current management has followed this relationship in the past.
Image Courtesy of SI.com
Georgetown University Class of 2017
Connect with Griffin on Linkedin: https://www.linkedin.com/pub/griffin-cohen/7a/b54/727
Follow GSABR on Twitter: @GtownSports
Like GSABR on Facebook