Line of Inquiry: Matthew J. Notowidigdo on Rent Sharing in College Sports

In our research, we were interested in studying rent sharing in college sports. What share of the revenue generated by football and men’s basketball actually shows up as compensation for the players in those sports? And we estimate about 7 percent of the revenue that’s generated shows up as compensation. And that compensation is things like tuition, room and board, a slight stipend for meals and other living expenses. But our best guess is it’s about 7 percent of the total revenue.

And you could benchmark that by comparing it to the professional sports leagues, like the NBA or the NFL. In those leagues, the players, through collective bargaining, have negotiated about 50 percent of the revenue flowing to them as compensation. And so that’s the gap that’s really the initial motivation for our paper, which is: All of this extra revenue that’s being generated by these sports, how does it flow through the athletic department and who are the ultimate beneficiaries of that extra revenue?

The data that we collected comes from athletic department reports that list the amount of revenue generated by each sport within an athletic department, as well as how much is spent on each of those sports. And so I think one contribution of our paper is just putting this data together systematically for all of the schools in the so-called Power Five athletic conferences for as many years that we were able to get access to that data for. But the key that we have is the amount that’s spent on every single sport and the amount of revenue generated by every single sport in each of these schools.

So what we see immediately just from looking at that data is that football and men’s basketball are the so-called profitable revenue-generating sports. For almost every school and every year, the amount of revenue they generate is substantially more than their average expenses. By comparison, for all of the other men’s sports and all women’s sports, on average the expenses exceed revenue. So these sports actually are not profitable. They’re so-called nonrevenue-generating sports because they only can cover their costs if there’s some transfers within the athletic department or perhaps from the university to the athletic department, so that these sports are able to make ends meet.

The thing we were expecting to see in the data is that the women’s sports and other men’s sports were going to be the recipients of that extra revenue. And that’s what we found. What surprised us a little bit more was that we also saw increases in coaches’ salaries, and not just the salaries of the football and men’s basketball coaches. We also saw increases in salaries of all of the other coaches. And I think that’s very interesting because it suggests that since you can’t pay college players, that all of this extra revenue that’s being generated has to go somewhere. And we find that a lot of it shows up as increased salaries of all of these other coaches.

We also see increased spending within the athletic department. I think most colorfully that shows up as increased spending on athletic facilities. And in the paper, we have some examples of what this looks like in practice. It could be things like miniature golf courses, lazy rivers. And I think that also makes sense intuitively if you think about it. Again, if you can’t compensate the players, you can’t pay them salaries, but you’re still competing for talent, one way of doing that is to just offer increasingly lavish athletic facilities.

So what we were able to do in the paper is also collect new data on all of the student athletes, playing all of these sports at, again, the schools in the so-called Power Five athletic conferences. We could then take those student athletes, we found their hometown, we found what high school they went to, and we could also merge that to their demographic characteristics, as well as the kind of neighborhood that they grew up in. What we found didn’t really surprise us as lifelong college-football fans. What we found is that most of the student athletes playing football and men’s basketball are Black. And most of the student athletes at all of the other sports on average are much more likely to be white. If you look at the neighborhood characteristics, we find that the student athletes playing football and men’s basketball are more likely to grow up in relatively low-income neighborhoods, whereas the student athletes for all of the other sports, except for football and men’s basketball, are more likely to grow up in relatively high-income neighborhoods.

Again, this didn’t surprise us, but this was the first time that we’d been able to get this kind of data, again, by going to the level of the student athlete. And I think if you combine it with our results on the transfers across sports within the athletic department, then I think there’s kind of an uncomfortable reality of the current system of college sports, which is that the revenue is being generated by student athletes that are more likely to be Black. And it’s being transferred to sports played by student athletes that are more likely to be white. And I think in this current moment of racial reckoning, that’s an important thing to keep in mind.

The way that we set up this exercise is to try to answer a very specific hypothetical question, which is: If the amount of revenue generated by football and men’s basketball changed—like it went up unexpectedly or went down unexpectedly—how would the athletic department respond to that? And I think, you know, that’s a somewhat narrow question, but it’s one we thought we could answer very accurately with the data that we had collected. And what we found is that if the athletic department unexpectedly gets some additional revenue from football or men’s basketball, it reinvests some of it directly in those two sports. It spends more on those sports in following years. It also spends more on all of the other sports, the other men’s sports and women’s sports. It also spends more on coaches’ salaries and spends more on athletic facilities that benefit all of the student athletes within the athletic department. And so that’s what we think we see very clearly in our data. And I think that’s useful for answering some but not all of the kind of important policy questions that are happening right now as we debate whether and how much to pay college athletes.

So take as one hypothetical, suppose you were able to pay football and men’s basketball athletes a small amount above what they’re currently allowed to be paid. Well, that essentially is taking money out of the athletic department that otherwise would have gone to these other sports. So our prediction is that the spending on the other sports would fall and the salaries for all of the other coaches would fall. And lastly, we’d expect to see slightly less investment in these lavish athletic facilities. And so that’s our prediction that comes from the estimates that we provide in our paper. And you have to be willing to extrapolate from our current setting to hypothetical scenarios where you change rules about how college athletes are able to be compensated.

I think the pandemic provides some perhaps unexpected support of a lot of what I’ve been talking about, which is that what happened during the coronavirus pandemic was that college football bowl games were canceled. March Madness was canceled one year. These are opportunities for football and men’s basketball teams to play additional games and generate substantial revenue for their athletic departments. So what our paper would say is that if there’s an unexpected decrease in revenue as a result of a lost bowl game or March Madness being canceled, we’d expect that to cause financial difficulties within the athletic department that would ultimately reduce coaches’ salaries, and reduce spending on other sports. And as we show in the paper, that’s exactly what we see happening anecdotally, as many colleges during the coronavirus pandemic cut back on spending on their other sports and insisted that coaches take temporary pay cuts to deal with the financial fallout from the pandemic.

It seems like a pretty wild extrapolation, but what we do go through in the paper are some stylized exercises, where we ask the question: Suppose the players, through collective bargaining, were able to negotiate the same percentage of total revenue as professional athletes are able to do in professional football and men’s basketball. And what we find are some quite large salaries on average being paid to the players. It could run as high as several hundred thousand dollars for football players and men’s basketball players, again, in the universities in these so-called Power Five athletic conferences. So this isn’t college football entirely. It’s only in these schools that run particularly profitable college football programs. I think you want to take these specific numbers with a grain of salt, but I think they show the amount of money on the line. These players, if they could engage in collective bargaining, may be able to negotiate much higher compensation than they’re currently receiving as a result of the rules of the NCAA.