Root, Root, Root, for the Home Team
What's the best sports town in America? Ask a couple of baseball-loving economists, and the answer becomes clear.

By John Kwoka

True, economics isn't usually known for its lighter side.

I may be thoroughly absorbed by my research into electricity deregulation, antitrust policy, and drug advertising. But I understand not everyone is equally interested in regional transmission organizations, data envelopment analysis, and advertising/sales ratios.

Occasionally, though, even economists delve into areas that, whatever their importance, offer a bit of pure fun. For instance, this question: What is the best sports town in the United States?

You may think this is an issue best discussed in the company of your friends, a widescreen TV, and your favorite beverage. I can report, however, that economics is able to cast a lot of light on the question, maybe even resolve it-not that any economics answer would ever end the fierce debate.

Pinpointing the top sports town is something that PhD student Kamen Madjarov and I have been working on since 2003, when Northeastern held a conference, organized by law professor Roger Abrams, to honor the hundredth anniversary of the first World Series.

That inaugural Series, which the Boston team won, is of historical significance to Northeastern because Boston's home games were played at the Huntington Avenue Grounds, on land that's now part of campus. The pitcher's mound was located about where the bronze statue of Cy Young stands today, near Churchill Hall's entrance.

To contribute to the World Series conference, I offered to apply some economics to the question of Boston's sports reputation. Then I watched as the project gathered momentum and eventually got a bit out of hand.

I'll explain all that later. First, a little background.

Everyone talks about what a terrific sports town Boston is, of course, but testing this assertion through the standard economic techniques of modeling and empirical analysis required several steps.

Step one was figuring out what, exactly, constitutes a great sports town. Clearly, it's not just having a good professional team. Many cities have fine pro teams that aren't supported by fans, which sometimes causes these teams to relocate to more hospitable towns.

This suggests the key to what makes a sports town great: loyal fans. Loyal fans come to games even when their team's performance and competitiveness start to falter. Fair-weather fans-who show up when a team is winning but drift off to other interests during the lean times-do not a great sports town make.

If loyal fans are the starting point, step two was translating the notion of team loyalty into a measurable variable that would allow us to compare cities. We looked at baseball for three reasons: Most debates about sports towns involve baseball; it's the sport that has the most data; and, obviously, it was the focus of the 2003 conference.

We measured loyalty by how well home attendance for each of the thirty Major League Baseball teams held up during the second half of the season. This is when more and more teams start to fall behind their division leaders, when making the playoffs may seem more and more unlikely (something long-standing Red Sox fans have become used to). This is also when fair-weather fans start to find other activities.

More technically (we are economists, after all), we estimated the attendance elasticity-the rate of attendance drop-off-during the last half of each team's home season. A low attendance elasticity implies a team's fans are loyal.

For step three, we needed data-lots and lots of data. For the purposes of the conference, we looked at 2002 attendance records for every home game played by every major league team. Since small stadiums like Fenway Park frequently sell out, the data were adjusted to account for stadium capacity.

When we correlated each team's attendance figures at every home game with its numbers of games-behind, we found that, in 2002, Red Sox fans really did seem to show up in greater proportions, even as their team fell farther and farther behind in the standings.

By our loyalty measure, the Red Sox seemed to be playing in the best sports town in America.

We presented this result at the end of the conference, a fitting conclusion to a commemoration of Boston's baseball history.

But our result only whetted Kamen's appetite. He was determined to test our finding further, and so, over the following months, he extended the data set to cover many more years. This allowed us to better capture patterns of fan behavior over time as well as between cities. We ended up with data on every major league game played between 1974 and 2003, a total of 62,336 games.

That's right: 62,336.

As before, the data set for each of these games included the home team's games-behind figure. This time, though, we also measured a variety of other factors that could affect attendance.

For instance, was it a day or a night game? A weekday or a weekend game? A doubleheader? A game against a same-division opponent? An interleague game? We noted the size and age of each stadium and whether the city had an NBA, an NFL, or an NHL franchise.

Finally, Kamen researched the temperature and rainfall recorded by the weather-reporting station nearest a baseball stadium for every one of the 62,336 days when games were played.

I believe it was at this point I realized our modest little exercise had gotten completely out of hand.

The rest of our work was actually pretty straightforward. We statistically estimated a demand function for baseball attendance and examined the separate impact of each of the factors.

We found that, more than any other variable, games-behind matters. It's the factor that most affects attendance. If a team falls too far behind its division leader in the latter part of a season, interest wanes, and attendance declines. All fans in all cities like to see a winner.

Interestingly, a visitor's games-behind also matters, and not the way you might expect. Fans actually prefer a stronger visiting team. They want to watch the home team win, of course, but they also want to think a game is going to be somewhat competitive, not a near-certain victory (games at Fenway Park when the New York Yankees come to town may be an exception to this rule).

Our results gave us some fascinating insights into other factors. Weekday games have lower attendance-12.3 percent lower, to be specific. Doubleheaders have 6.3 percent greater attendance. Divisional games are associated with slightly lower attendance; interleague play raises attendance by about 4.5 percent.

Every year a ballpark ages reduces attendance by about one third of 1 percent, meaning a ten-year-old venue will attract about 3.4 percent fewer fans. Small wonder new parks are in such vogue-although whether they are worth their cost is an entirely different question.

Kamen's efforts to measure weather conditions were rewarded. The data showed that, other things being equal, fans definitely don't like cool or rainy weather.

A city's alternative sports attractions have a big impact, too. The presence of either an NBA or an NFL franchise lowers baseball attendance by nearly 2 percent. An NHL franchise makes no real difference, probably because, at least until recently, hockey and baseball seasons didn't overlap.

Overall, though, our reams of research proved that games-behind is the factor with the largest impact on attendance.

Now, to the big question: Which major league team has attendance that holds up best when the team does not play well?

Our statistical model answered that question precisely, by calculating the elasticities of each team's home attendance with respect to its games-behind over the entire period of 1974 to 2003. We looked for the team with the smallest elasticity. That's the team with the most loyal fans.

And the answer is—drum roll, please—the Boston Red Sox! For every additional game behind that Boston was, relative to its division leader, 0.2 percent fewer fans showed up. In other words, if the Red Sox were ten games out, they filled 2 percent fewer seats. That's only about seven hundred fewer fans, a pretty small drop-off.

Before you think we cooked this result, I should say we were actually surprised by it. Boston's claim to best sports-town status has always seemed more based in repeated assertion than in hard evidence. We thought we might have arrived at a statistical fluke.

So we checked to see which team our model showed as having the second most loyal fans during this period. Sure enough, just as baseball aficionados would guess, it's the Chicago Cubs. Finding this, we were convinced we were measuring the real thing.

In fact, we are forced to admit that, if you look at just the past ten years, Cubs fans may have been slightly more loyal than Red Sox fans. But Sox fans definitely get the prize if you consider the entire thirty-year period.

So there you have it: Boston really is the best sports town in the United States.

Lots of people already believed this. We're pleased we've been able to "prove" it with economics. I know that won't stop many of you from arguing your opinion with friends at local debating venues. Still others, preferring the fieldwork approach, will go to games at Fenway Park as often as possible.

As an economist and a fan, I personally endorse all these approaches.

John Kwoka is the Neal F. Finnegan Distinguished Professor of Economics. When not dabbling in sports economics, he teaches and does research into industry economics, antitrust, and regulation.

  Iliustrations by Mark Steele