Are South Florida Fans at Fault?

Soon after the Florida Marlins announced that the club would explore relocation and hold a “fire sale” to reduce payroll, South Florida sportswriters and broadcasters identified the responsible party: the fans. Never mind the team’s inability to obtain public funding for a new ballpark; the ungrateful fans deserve the blame for this Fishy fiscal fizzle. Although owner Jeffrey Loria exonerated them, some Marlins fans—seemingly eager to jump aboard another Miami bandwagon—confessed culpability for this baseball banana republic. In response to a newspaper’s online poll that asked “who is most to blame for the Marlins’ fire sale,” 36% of the 8,429 participants fingered the “fans who don’t show up,” while only 19% blamed elected officials for not helping the team get a new stadium.

On the surface, South Florida does indeed look like a lousy market for Major League Baseball. The Marlins regularly rank near the bottom in home attendance despite two World Series championships during the past nine seasons. Last season’s attendance of just over 1.8 million fans was substantially below the MLB average of nearly 2.5 million. But attendance is simply a counting stat, and we can analyze the data to explore a little deeper before writing off South Florida as a hopeless baseball backwater.

Fan support is just one variable in the attendance demand function. Teams operate in disparate market environments and with unique bundles of quality-price attributes. For simplicity, suppose teams A and B are identical in every respect including incidence and intensity rates for local baseball fans, except for a population of 4 million in market A and 2 million in market B. It’s reasonable to expect that team A will generate greater attendance, all other variables being equal.

Sports economists build regression models to analyze the effects of these variables on the demand to attend sports events. The models typically explain around 60% of the variance in attendance among teams. Because a model can incorporate only observable inputs such as ticket and concession prices, market income and population, stadium age, and team winning percentage, the unexplained 40% includes intangibles like fan loyalty and marketing effects (plus random error). If we fit a model that does a good job of predicting an average team’s attendance and then compare predicted vs. actual attendance, we could interpret the difference as a measure of a fan loyalty and support for a team.

Several studies have confirmed the link between winning and attendance in baseball. A particular team’s sensitivity to this effect can vary greatly, though. When an MLB club performs efficiently at the gate, predicted season-long average attendance will be close to the actual attendance number. A team that enjoys intense fan support—think of markets like Boston and St. Louis —is super-efficient at converting its price-quality-market attributes into paying customers. Fans remain loyal and keep paying to watch baseball games regardless of the team’s performance. At the opposite end of the spectrum, Tampa Bay’s inefficiency means that winning 21% more games in 2005 to reach the .500 mark would have increased attendance only by about 8%.

To quantify this notion of attendance efficiency, I use a conversion ratio scale where 1.0 suggests neutral fan support (attribute inputs = attendance output). Most MLB teams produce conversion ratios between 0.8 and 1.2, and these values tend to stay consistent over the 2000-2005 seasons included in this analysis.

The following table compares per game attendance predicted by the demand model versus actual attendance for Florida’s past six seasons:

             Predicted Avg    Observed Avg                 Conversion
 Season       Attendance       Attendance    Difference       Ratio
    2000          22,903          15,229         -7674         0.66
    2001          25,282          15,765         -9517         0.62
    2002          23,236          10,038        -13198         0.43
    2003          24,921          16,290         -8631         0.65
    2004          28,025          22,091         -5944         0.79
    2005          28,065          22,792         -5273         0.81

The Loria regime took control of the Marlins in 2002, when discouraged fans purchased fewer than 2,000 season tickets. Over the next three seasons fans gradually returned to Pro Player/Dolphins Stadium, and by 2005 actual attendance was within around 5,000 fans of the number modeled from the team’s price-quality-market attributes. Conversion ratio, an approximation of intangible fan loyalty, has consistently trended upward to a nearly respectable 0.81. Given the team’s combination of performance, market, and stadium inputs—many of which are beyond management’s control—the Marlins’ plan to rebuild and reenergize South Florida baseball fans appears to be working. Not by quantum leaps, as the team might wish, but by perhaps more realistic incremental progress.

So exactly how bad a market for big league baseball is Miami? Using the 2005 conversion ratio and predicted-observed difference statistics as proxies for fan loyalty, South Florida fans may not be as blasé as hypothesized. The Marlins’ attendance efficiency was similar to that of the 2005 World Series winner and the host market for the 2005 All Star Game, and considerably better than that of the cross-state AL franchise:

          Pred.-Obs
          Attendance   Conversion
Team      Difference      Ratio
FLA         -5,273        0.81
CHA         -6,414        0.82
DET         -5,656        0.82
MIN         -4,813        0.84
OAK         -5,307        0.83
TB          -8,437        0.62

The “significant market correction” being made by the Marlins to bring payroll in line with revenue makes business sense. The degree of fan support for Oakland seems similar to that for Florida, forcing general manager Larry Beinfest to play his own version of Moneyball. That’s not necessarily a bad thing. And don’t be surprised if the effects on future attendance from the great “fire sale” of 2005 are nothing like those after 1997.