What the geography of the All-Star game says about American cities and metros.
This weekend, the National Basketball Association tips off its annual All-Star game. The game's starters are chosen by fans and the rest of the team is chosen by coaches.
What is the geography of NBA stardom? In which cities do most superstars play? The short answer is that, like almost everything else, the distribution of NBA superstars is spiky: most of them end up in a handful of big cities.
That geography has become even spikier in recent years.
Of the starters on this year’s team, four of the five on the West team are from Los Angeles – Kobe Bryant and Andrew Bynum from the Lakers and the Clippers’ Blake Griffith and Chris Paul. Only one, Kevin Durant, hails from a small-market metro, Oklahoma City. Of the East’s starters, Carmelo Anthony is from New York and Derrick Rose is from Chicago. Two (Dwayne Wade and LeBron James) are from Miami and Dwight Howard is from Orlando (though he is rumored to be leaving for New York, Los Angeles or another bigger city when his contract expires later this year). The reserve players for the West hail mainly from small markets, while those for the East play mostly in big cities (with one exception, Roy Hibbert from the Indiana Pacers).
The short term is one thing, but to get at longer-term trends, I turned to MPI alumnus and UCLA doctoral student Patrick Adler as well as some published studies. Adler’s data covers the locations of NBA All-Stars since the league’s 1976 merger with the old American Basketball Association (or ABA).
Los Angeles takes first place with 74 All-Star appearances, followed closely by Boston with 64. Phoenix is third with 52, and Detroit and New York are tied for fourth with 48. Philadelphia has 47 (see the map above).
There's a close association between metro size and All-Star appearances (a correlation of .58) according to Adler’s analysis, which makes intuitive sense. The scatter-graph above plots All-Star appearances against population size. Cities above the line have relatively more All-Stars for their population size, while those below it have less. Los Angeles and Boston have substantially more All-Stars than their size would suggest. New York is right on the line; Chicago slightly below it. (If numbers on the graph for population look strange, it's because they are logged).
Big cities have an even bigger edge in true superstars, players who are All-Stars year after year. Since 1976, players making repeat appearances have taken nine of 10 All-Star spots. Sixty percent of spots have gone to the 65 players who have played in five or more All-Star games. And the dozen players who have played in 10 or more All-Star games account for roughly 30 percent of all slots. Los Angeles leads in the number of repeat All-Star slots with 69, followed by Boston with 61. Detroit has 48, Phoenix 47, Philadelphia 46, and New York and San Antonio 43.
Simple arithmetic explains some of the reasons that large metros produce more All-Stars. Los Angeles, for example, has two teams to choose from. Bigger metros also have a larger population of potential All-Star voters.
Some large metros have teams with very long histories; they’ve played more seasons than the newer teams in smaller metros. To control for this, Adler divided the total number of All-Star appearances per metro by the number of seasons played. When he did this, we found that the big city advantage disappears in the long run. Of cities with currently active teams, Boston leads with 1.78 All-Stars per season, followed by Phoenix with 1.44, Denver with 1.39, Detroit with 1.36 and Philadelphia and San Antonio with 1.31 each. Los Angeles has 1.16. But Chicago (with .97) and New York (.67) are much further down the list. Large metros do no better than small metros on this metric. There is no statistical association whatsoever between population size and the number of metro All-Stars per seasons played.
But the short run tells a completely different story. When Adler confined his analysis to just this past decade, he found a significant and positive association between population size and metro All-Star seasons (the correlation is .46).
What happened over this period to alter the geography of basketball superstars so substantially? There are many reasons – bigger salaries, bigger endorsement deals, free agency, the transformation of the NBA into a major entertainment business under David Stern.
It also coincides with the rise of the superstar city. The 1960s, '70s and '80s were the age of urban flight and the ascendant suburbs. Many date the absolute nadir of the big city as 1975, with the onset of the New York City fiscal crisis. But as we all know, economic and demographic shifts over the past two decades have powered the movement of talent and business activity back into the heart of their urban cores. This is the era Edward Glaeser dubs the "triumph of the city."
Simply put, superstar cities fuel superstar careers. When it comes to the All-Star game, big-city players benefit from the simple fact that they have more local fans to vote them onto the team. And of course, big cities have more money for salaries, more corporate customers, larger media markets, and more lucrative endorsement opportunities. The key to sustaining a superstar career is playing for a winning franchise surrounded by other All-Star players. This is much more likely to happen in an elite franchise in a big city.
All of this is compounded by the increased mobility not just of individual players but of groups of two or even three players. Since championships reinforce superstar careers, superstars band together to plot their moves to certain franchises and locations.
As spiky as the geography of NBA superstardom may be, it’s not as spiky as what you find in some other fields. According to my own research, New York and Los Angeles outperform Chicago and other cities by a vast margin in entertainment, arts, and culture occupations.
Even with free agency, NBA basketball is not a completely free market, as the recent lockout proved; smaller franchises do have some recourse when it comes to holding onto talent. Also, superstar management, coaches and histories can still trump the advantages of size for some relatively smaller market teams. San Antonio was a dominant power and source of superstar players (Robinson, Duncan, Parker, and Gonobli) during the early and mid 2000s. Oklahoma City is sending two players to the All-Star game this year – megastar Kevin Durant and Russel Westbrook, and it has one of the league's best records. Its success is premised on the ability to evaluate external talent and recognize internal assets, as it builds a solid team around Durant. Of course the question remains of whether he'll stick around when his current contract is up.
That said, the trend is toward more, not less, spikiness. Even as globalism expands the boundaries of our world, fundamental economic forces accelerate the clustering of economic activity and the geographic concentration of talent. The pool of cities that can attract and retain basketball superstars grows increasingly smaller as they themselves grow ever-larger.
Top image: Reuters/Gary Hershorn