« Markets in everything, American Indian edition | Main | Optimal insults »
Thinking about Sports and Economics
I spent last Saturday at a very interesting conference on Sports Statistics, run by the Sports Stats section of the American Statistical Association. It was a fun day, involving academics, sports journalists, and those Moneyball-inspired quants working for various sports teams.
But at some point I asked myself: Why do economists work on sports?
- Sports provide unique opportunities to test economic theories. Cribbing from a New York Times article, this is the Thaler defense:
"'My justification for doing this is that it's the one really high-stakes activity where you get to watch all of the decisions,'' Thaler said. ''If Bill Gates invited me to watch all of his decisions, I'd talk more about that.''
- Sports shapes broader national debates. Sports is a microcosm of our broader society and our national narrative on the important issues, from drugs, to race, to cheating, to sexual harrassment often play out on our sports pages. In honor of a particularly compelling example, let's call this the Jackie Robinson defense.
- Professional sports are an important part of the economy. I call this the Dog defense, not as a dyslexo-religious statement, but simply because dogs raise an important question: aren't pets a bigger part of the economy than professional athletics? If so, why are there so many papers on professional sports and so few on the economics of dogs?
- Sports participation is an important activity. It seems important to learn whether sports make us happier, healthier or more productive. For instance, it is important to learn, say, what the broader effects of Title IX were. Under this view, research on sports is part of the human capital agenda, leading me to call this the Gary Becker defense.
- Sports provides a useful teaching metaphor. Many of those teaching Sabermetrics-inspired courses argue that sports provides a useful vehicle for teaching something far more important - basic quantitative reasoning. When I teach my class on behavioral economics, I do so by analyzing anomalies in sports betting markets.
- Doing research on sports is fun. It was no mistake that the conference I attended was on a Saturday. Many of the academics in attendance were giving up leisure, not more important work. But for some, sports provides a chance to mix work with leisure; of course, if non of the above arguments holds, then it is just a chance to mix leisure with leisure.
Let me now translate this into advice, because I often hear from students wanting to write a thesis on sports. My first response is always: Don't. Too often, we find our sporting heroes more interesting than other people do. (Yes, I have been guilty of breaking this rule.)
But if you must work on sports, make sure you have a defense to this charge. I find the Thaler and Becker defenses most compelling, because they speak to the broader economic issues or yield policy implications. The Jackie Robinson defense is also important, but not applicable often enough. The Dog defense is often raised, but rarely compelling; neither pets, nor professional sports, are really a big part of the economy (estimates to the contrary usually turn out to be more applicable to the Becker defense).
Posted by Justin Wolfers on October 4, 2007 at 02:23 PM in Economics, Sports | Permalink
Comments
One thing to be careful about with sports is that it is a zero-sum game. It is a bad metaphor for the economy or the stock market in that respect.
Posted by: Arnold Kling at Oct 4, 2007 3:04:33 PM
I recently told a friend, fomerly of the Fed and currently with a large investment bank, that trading sports outcomes helped traders (who are familiar with sports) learn how to better trade and manage risk, thereby adding greater accuracy to the predictiveness of markets beyond sports. He countered that I should feel so strongly about poker and blackjack, but I demurred, proposing that an introductory probability course would be enough to move the student beyond the lessons of those games.
Posted by: caveat bettor at Oct 4, 2007 3:21:56 PM
The elephant in the room here is that many sports produce a vast quantity of analyzable data in a reasonably convenient form.
Posted by: Kieran at Oct 4, 2007 3:49:56 PM
Speaking of Jackie Robinson, are there important phenomena first seen in sports later seen elsewhere? Or important tools first used to study sports, later used elsewhere?
This is the sports as model system defense.
Posted by: Seth Roberts at Oct 4, 2007 3:59:13 PM
Sports are a (the) civilized proxy for war. Rooting for one's team is akin to praying for victory on the battlefield.
Nothing new about this. The Duke of Wellington claimed that the "Battle of Waterloo was won on the playing fields of Eton".
Wsr, even indirectly, is more important than pets.
Posted by: Peter Schaeffer at Oct 4, 2007 4:12:25 PM
To #5 I would add that many people, including me, teach courses on the economics of professional a/o college sports. Such a course typically blends IO, Public Choice and Finance, Labor, and Law and Regulation (i.e., Antitrust) topics. Students seem to enjoy these classes and I'd like to think that sports is the bait that may get them to learn (and remember) a concept or two.
Posted by: Aaron Aardvark at Oct 4, 2007 5:00:03 PM
"Speaking of Jackie Robinson, are there important phenomena first seen in sports later seen elsewhere?"
Yes, you might find interesting my 1996 National Review article "How Jackie Robinson Desegregated America" that quantitatively shows how Gary Becker's theory of discrimination played how just as predicted in major league baseball from 1947 into the 1960s, with important influences on the nation as a whole in its attitudes toward blacks. Here's an excerpt:
"Baseball's long struggle over race can yield some surprising perspectives on our national predicament. The Robinson epic is generally lumped in with the 1954 Brown decision against segregated public schools and the 1964 Civil Rights Act outlawing job discrimination. Yet two crucial differences stand out. 1) The integration of organized baseball preceded the civil-rights revolution, and in reality baseball helped make later reforms politically feasible by giving white Americans black heroes with whom to identify. 2) Government had almost nothing to do with this triumph of the competitive market. Baseball owners finally realized that the more they cared about the color of people's money, the less they could afford to care about the color of their skin.
"It's ironic that the hallowed civil-rights revolution owed so much to something as seemingly trivial as pro sports. Yet, without this business of producing heroes for public consumption, whites might never have cared enough about blacks to be bothered by racial injustice. It's not the most noble trait of human nature, but we tend to be more outraged by minor slights to winners (note the endlessly recounted tales of the indignities Robinson endured) than by mass atrocities against downtrodden losers.
"That competitive markets make irrational bigotry expensive -- not impossible, but costly -- was first formally demonstrated in 1957 by University of Chicago economist Gary Becker (the 1992 Nobel Laureate), and in the four decades since has barely gained a toehold in conventional thinking."
http://www.isteve.com/JackieRobinson.htm
Posted by: Steve Sailer at Oct 4, 2007 5:07:09 PM
My 1997 National Review article, "Track and Battlefield," took up the question of whether women would ever become physically capable of infantry combat by quantitatively analyzing the conventional wisdom that women were rapidly closing the gap with men in Olympic running events. We showed that the gender gap in track times had actually grown significantly from 1988 to 1996:
"Despite all the hype about 1996 being the "Women's Olympics," in the Atlanta Games' central events -- the footraces -- female medalists performed worse relative to male medalists than in any Olympics since 1972. In the 1988 Games the gender gap for medalists was 10.9%, but it grew to 12.2% in 1996. Even stranger is the trend in absolute times. Track fans expect slow but steady progress; thus, nobody is surprised that male medalists became 0.5% faster from the 1988 to the 1996 Olympics. Remarkably, though, women medalists became 0.6% slower over the same period."
We then showed that this was due (A) to the fall of the Berlin Wall, which had destroyed the East Germany steroid-cheating process; and (B) to the introduction of tougher steroid testing after 1988's 100m dash winner Ben Johnson had gotten caught and Florence Griffith-Joyner had gotten away with setting ridiculous records.
http://www.isteve.com/gendrgap.htm
Posted by: Steve Sailer at Oct 4, 2007 5:23:44 PM
My 1997 National Review article, "Track and Battlefield," took up the question of whether women would ever become physically capable of infantry combat by quantitatively analyzing the conventional wisdom that women were rapidly closing the gap with men in Olympic running events. We showed that the gender gap in track times had actually grown significantly from 1988 to 1996:
"Despite all the hype about 1996 being the "Women's Olympics," in the Atlanta Games' central events -- the footraces -- female medalists performed worse relative to male medalists than in any Olympics since 1972. In the 1988 Games the gender gap for medalists was 10.9%, but it grew to 12.2% in 1996. Even stranger is the trend in absolute times. Track fans expect slow but steady progress; thus, nobody is surprised that male medalists became 0.5% faster from the 1988 to the 1996 Olympics. Remarkably, though, women medalists became 0.6% slower over the same period."
We then showed that this was due (A) to the fall of the Berlin Wall, which had destroyed the East Germany steroid-cheating process; and (B) to the introduction of tougher steroid testing after 1988's 100m dash winner Ben Johnson had gotten caught and Florence Griffith-Joyner had gotten away with setting ridiculous records.
http://www.isteve.com/gendrgap.htm
Posted by: Steve Sailer at Oct 4, 2007 5:24:40 PM
As always, sorry about the double postings. I have no idea why this happens so often on this website.
Posted by: Steve Sailer at Oct 4, 2007 5:35:44 PM
There's another good reason I don't see mentioned here - the sheer amount of data. Most professional sports (baseball especially) provide enormous numbers of observations of events. Like fruit fly generations or stock price fluctuations, they allow data points to gather in sufficient quantity as to gain the nuance and force underlying what Bill James called the 'power of language.' I think that this is a fundamental component of the attraction of sports analysis for statistically-minded researchers, and one aspect which will keep the flow of interesting research coming for some time.
Posted by: Aaron Bornstein at Oct 5, 2007 11:11:14 AM
