Examining Potential MLB Expansion Cities, Part 1

To commemorate Rob Manfred’s taking over as Major League Baseball’s 10th commissioner, we ran a series of articles giving Manfred suggestions on ways to improve the game during his tenure. One of these suggestions came from Jeff Zimmerman, who called for an increase in the number of teams from 30 to 36.

Zimmerman made some convincing arguments in favor of a need for expansion. He observed that expansion has stopped happening, with the number of teams staying steady while the pool of players to draw from has continued to grow. Additionally, he notes that league-wide run scoring increased the last four times MLB expanded. Adding more teams might help cure baseball’s declining run scoring.

If Manfred and Co. do in fact decide to increase the number of teams, they’ll need to find cities to support these new franchises. Recent history suggests this is no easy task. Major League Baseball added four clubs since 1993, and two of them — the Miami Marlins and Tampa Bay Rays — have struggled mightily to get butts in the seats. In fact, these two teams have brought up the rear in attendance numbers ever since the Expos left Montreal.

The Marlins’ low attendance figures can be blamed — at least in part — on the penny-pinching ways of Jeffrey Loria, who has a habit of alienating fans by holding bi-decade fire sales. But what about the Rays? They’ve been in the playoff hunt just about every year since 2008, and have a .552 winning percentage over this period — fourth best in baseball. Still, their attendance figures have been among the game’s worst. In the last three years, they’ve ranked 30th, 30th and 29th. The dilapidated state of Tropicana Field surely plays a role in inhibiting the Rays’ attendance figures, but that can’t be all that’s going on.

There are surely other factors that differentiate a city like Miami from, say, Milwaukee, which draws nearly twice as many fans despite having roughly one quarter as many people living in its greater metro region. But what else comes into play?

To get at this question, I built a regression model considering recent seasons for teams  from Metropolitan Statistical Areas (MSAs) that have exactly one team. I set average attendance for home games as my dependent variable, and considered a broad swath of variables for my independents. Some had to do with the team itself, like recent win-loss records and new stadiums. Unsurprisingly, having won a lot of games recently and opening a new stadium both result in an attendance bump. But including these variables was more about zeroing in on the less-obvious factors that drive attendance: Those that deal not with the team, but with the team’s home city.

Before I go any further, I’d like to make one technical note. When I use the word “city” here, I’m actually referring to that city’s greater Metropolitan Statistical Area, and not the city proper. For example, “New York City” refers to not just the five boroughs, but also much of the surrounding area, including Long Island and part of New Jersey.

In building my model, I did some out-of-sample testing to determine which variables to include in my final iteration. In other words, I built a few potential models using 60 percent of my data, and tested them on the remaining 40 percent to ensure I wasn’t overfitting the data. Somewhat surprisingly, a city’s population size did not make the final cut in my model. I’m thinking this has something to do with the selection bias in my sample: All large cities have a major league team, but perhaps many of the small cities with relatively poor attendance-drawing abilities aren’t in the sample. Population size is obviously a factor, but it’s just not accounted for in my model.

When it was all said and done, and the regression-induced dust had settled, I came away with 23 variables. These variables returned a respectable .76 R^2 coefficient. I’m not going to bore you by trying to explain every one of these variables, but will instead provide some key takeaways as to what characteristics make a city well-suited to draw fans to a big league ballpark. Here’s what I managed to uncover, ranked roughly in order of importance:

1) Cities with higher incomes and lower levels of poverty draw more fans

This one is pretty straightforward. Well, the explanation is, at least. My model includes five income-related variables of differing importance, but the bottom line is that high-earning cities do a better job of generating attendance. This isn’t surprising. If residents have higher incomes, they have more money to spend on goods and services, including buying tickets to baseball games. Simple as that.

2)  Cities with a higher percentage of black and/or Hispanics/Latinos draw fewer fans

Before I dive into this one, a bit of a technical clarification is in order. (Last one, I promise.) The U.S. Census Bureau considers race and ethnicity to be two different things, meaning census takers are asked to identify both race (based on physical appearance) and ethnicity (based on ancestry). This means that a person can identify as white and Hispanic or black and Hispanic, etc.

Now, on to the significance of a city’s racial and ethnic demographic. It’s been well documented that black players are underrepresented in the majors. So it’s not at all surprising that they’re underrepresented in baseball’s fan base as well, especially relative to the other major sports.

Blacks make up just a sliver of baseball’s viewing audience, so it’s easy to see why cities with a larger share of blacks wouldn’t draw quite as well. Cities with large Hispanic populations face a similar struggle, although the effect isn’t as large.

3) After accounting for income, cities with more college-educated individuals draw fewer fans.

It’s important to remember that my model also includes variables for income, which often goes hand-in-hand with education. So what this is really saying is: Given two cities with the same income demographics, the one with fewer college grads will draw more fans. This effect is fairly substantial, but it’s hard to pinpoint exactly what’s behind it. Maybe college-educated people have a broader array of cultural interests, making them more likely to spend their incomes on things other than baseball games? Sounds reasonable, but that’s only a guess.

4) Cities with a higher percentage of males draw more fans

This one makes sense too. While there are plenty of female baseball fans, they’re far outnumbered by their male counterparts: Males accounted for 70 percent of baseball’s viewership in 2013 according to Nielsen. The average man is much more likely to be a baseball fan than the average woman, so it’s only logical that teams find more paying customers in male-heavy cities.

 5) Warmer cities tend to draw more fans

Cold weather sucks. Especially when you have to sit in it for three to four hours at a time. This likely explains why cities with higher average temperatures tend to draw better than their colder counterparts.

6) Cities with an NFL team draw fewer fans, but the hit isn’t as big if the city also has an NBA team

Although I hate to admit it, football is far and away the most popular sport in America. So if there’s a NFL team in town, it will inevitably draw fans away from baseball. The second piece of this one is a little less straightforward, but I think it might have something to do with the type of city that’s able to host multiple professional sports teams. If a city has what it takes to support both an NFL and an NBA team, it’s likely well-suited to support sports teams in general, which seems to outweigh any hit caused by having an NBA team in town.

7) Cities with older populations tend to draw more fans

Baseball fans tend to be old. According to Nielsen, half of baseball’s TV viewership comes from people 55 or older, which is a significantly larger slice of the pie than in other major sports. And having a large population that is 65 or older substantially drives down attendance.

Given this factoid, it’s no surprise that cities with older populations do a better job of getting fans into the ballpark.

However, having too many people who are too old also seems to be a bad thing. This can probably be explained by the fact that younger people are more active, and therefore more likely to go to games instead of watching from the couch. No doubt, this is one of the culprits for the attendance problems faced by the teams hailing from Florida — the retirement capital of the U.S.

Now that we’ve established our parameters and takeaways, tomorrow we’ll dive in and examine possible expansion cities.

References and Resources

• U.S. Census: American Community Survey, 2013 1-year estimates; 2009-2013 5-year-estimates
• Bureau of Economic Analysis (BEA)
• Weatherbase.com
• Baseball-Reference.com
• Nielsen: Year in Sports Media Report, 2013
• World Bank
• Brookings Institute: Global Metro Monitor