Yesterday, I tweeted this out:
It’s gotten a certain amount of traction (Twitter tells me around 20,000 people have seen that Tweet) and so I began to be concerned that I was being irresponsibly provocative. So let me explain how I got that number.
The Indy Star reported that the NCAA made $769.4 million off March Madness in 2013 ($681M from CBS for TV rights, $82.3 in ticket sales, and $6.1M in ancillary revenue streams). So, partly for fun, partly out of curiosity, but also out of the conviction that the labor of performing athletes is the primary driver of these revenues, I divided the total revenues by the total player minutes to come up with a figure of $30,224.91 generated per player minute of the 2013 NCAA D1 Men’s Basketball Tournament. Based on that figure, my five students together “generated,” with their 1017 minutes of basketball over 6 games, $30,738,733.50.
Incidentally, some other facts I discovered in the process: 674 players saw action during the tournament, sharing a grand total of 25,254 minutes of playing time. Of course, common sense would tell you, since half the teams are eliminated after the first round, that the players on teams making deeper runs are going to have a higher share of the minutes. What I found, though, was pretty amazing
- Just 14 players used 10 % of the total minutes.
- Just 36 players used 25 % of the total minutes.
- Less than one tenth of players used more than half of the total minutes.
- Just one third of the players used 75 % of the total minutes.
- Half the players used 85 % of the total minutes.
People go to NCAA games and watch them on TV to see basketball players play basketball. In a very real sense, minutes of basketball played creates revenue dollars for the NCAA, a fact not lost on the NCAA which has increased the number of minutes played by expanding the tournament. If so, then imbalances in the distribution of overall minutes matter. They matter no matter what. But they matter doubly, I would argue, when we consider that black players are disproportionately represented among the group of players using most of the minutes and so generating most of the interest and dollars (11 out of those top 14).
But minutes and NCAA revenues are just one way to frame the story.
Even someone who values the performance of these athletes as much as I do, who knows that if it weren’t for their hard-work and talent there would be no March Madness, must also admit that arenas, coaches, training and all other manner of capital investments (laid out before and during the tournament) also contribute to the madness and so to the revenues. It also reduces player labor to a single quantity: minutes, which isn’t the worst way to do it, especially from the NCAA standpoitn. Though I recognize the NCAA will make a bit more or less money depending on who makes a deep run in the tournament, it essentially makes its money regardless of who wins.
But there’s a reasonable argument to made that the productivity of a player, measured in terms of contributions to wins (which generate revenues) matters more, at least at the level of individual institutions. So while I think my $31M figure illuminates, albeit roughly, the correlation between minutes of player labor and revenues, I wouldn’t necessarily go the mat with an economist arguing that it’s the best way to measure exploitation.
So, to begin get a more precise sense of the exploitation of those five students of mine during the 2012-2013 season, I’m borrowing a page from Dave Berri, who in 2014 wrote a useful primer on the economic exploitation of college athletes for Time magazine. Let me walk you through that. Berri defines exploitation as “paying a worker less in wages than their economic contribution to the firm.” In terms of college athletes, exploitation occurs when the value of the scholarship, housing, and any stipend the athlete receives in exchange for competing is lower than the amount of revenue the athlete generates for the school. So, though I’m not mathematically adept, I believe we can turn this definition into relatively simple formula (as Berri goes on to do).
Exploitation (E) = Scholarship Value (SV)/Revenues (R) x 100%
Following Berri, I begin by getting the basketball revenue figures reported by Michigan to the Department of Education and posted on the latter’s “Equity in Athletics Data Analysis Cutting Tool” and discover that Michigan basketball reported $13,636,966 that year. Let’s just call it $13.6M.
2012-2013 University of Michigan Men’s Basketball Revenues = $13.6M
According to Berri, “Currently the NCAA restricts the payment of athletes to essentially the cost of attending the institution. But in a typical labor market, the payment to workers is unrestricted.” So the question is, what would Michigan have to pay its basketball players in an unrestricted market?
To get at this, we follow Berri in adopting the revenue sharing proportions used in the comparatively unrestricted major professional sports leagues in the United States, including the NBA, where the collective bargaining deal (because, you know, NBA players, unlike “student-atho-letes“, have a union) stipulates roughly a 50/50 revenue split between owners and players. (Berri notes that the labor market for professional athletes in the US is, in fact, still restricted and that the proportion of revenues they’d receive would likely be higher in a truly unrestricted market, but whatever.)
So, if the 15 players on Michigan’s roster were to receive 50% of the 2012-13 revenues they’d be splitting $6.8M, which works out to $400K apiece.
2012-2013 University of Michigan Average Men’s Basketball Player Revenue Share (In Unrestricted Market) = $400,000
The University of Michigan estimated the cost of attendance for out of state freshman and sophomores living on campus for 2012-2013 at $51,976. Let’s be generous and call that $52K.
2012-2013 University of Michigan Cost of Attendance = $52,000
Now let’s plug these values back into the exploitation definition/formula Berri gave us before. E (UM) = 52K (COA/player)/ 400K (1/2 hoops revenue/player) x 100 %. Do the math and I come up with Michigan players getting about 13 % of what they generate. Or, to put this another way that makes more sense to me: Michigan netted about $348K in profits off each player on the team in 2012-2013.
Average University of Michigan Profit Per 2012-2013 Player = $348,000
Here I hit a wrinkle that Berri does not account for (and I welcome anybody reading this to correct my efforts to do so). The University told the Department of Education that it cost about $7.5M to operate the men’s basketball program in 2012-13. I don’t have a breakdown of those costs (though I assume scholarships are figured in there and so I’m probably double-counting that expense). But just for fun, let’s subtract that from revenues. Doing so ($13.6M-$7.5M) gives us $6.1M in profits. Now let’s split that 50/50 with our players, leaving them $3.05M to split 15 ways. They’d each get a bit over $200K, which is still 75% more than the school’s own COA figure. In other words, by the most generous calculation I can come up with, the school made nearly $150,000 off each and every member of the 2012-2013 men’s basketball team.
Adjusted Average University of Michigan Profit Per 2012-2013 Player = $148,000
This gives us the average rate of exploitation. But, Berri, recognizing that pro franchises don’t pay all players the same amount but rather pay them to win games, applies a further calculation to factor in an approximation of each player’s contribution to the team’s wins. He takes the total revenues divided by the team’s wins to get at the value of a win, and then multiplies these by each player’s “win share” (or contribution to total wins, calculated through this complex formula, but also available here) to get at what he consider to be a more realistic and so equitable estimate of each player’s share of the revenues based on their actual productivity on the court.
[Caveat: I’m not really on board, philosophically with the individualistic, laissez-fair economic principles driving these calculations so you shouldn’t take this to mean that I argue that these numbers alone should dictate solutions to the problem of college athlete exploitation. But I think these numbers should be the starting point, after which we need to factor in other things that have value, even though that value isn’t reflected by an unrestricted market.]
Let’s go back to the five students I started with, who also happen to be the five players on the 2012-13 roster who led the team in win shares: Trey, Glenn, Nik, Tim, and Mitch. And don’t forget, if I were using Berri’s values, which do not subtract expenses from revenues, these figures would all be about twice as high. Here’s how that turns out:
That’s annually. In other words, when productivity is taken into account using win shares, we find that the University of Michigan made $1.3M off its $52,000 investment in Trey Burke.
Now, for each of Michigan’s 31 wins during that season, the numbers look like this:
Okay, now, let me also go back to my other starting point: March Madness, where Michigan got five of their 31 wins before losing to Louisville in the title game. How much did my students contribute to those wins? How much revenue did those five wins generate? How much did the UM pay for the players’ services in those five games? And how much profit did UM make off each of those players? To be really precise, I’d have to calculate the WinShares for each player for the five March Madness victories and I don’t have time. But to give an estimate, we just have to multiply the per win figures above by 5 (the number of March Madness wins).
Lastly, I want to relate all this to minutes. Basically, I want to know how much the UM made per minute that each of my students was on the floor during March Madness. So I’m going to take the total UM profit for each player for March Madness (the right hand column above) and divide it by each player’s total minutes.
So that’s the bottom line for me. The University of Michigan reaped just under $1,000.00 off of every minute of Trey Burke’s performance during the 2013 NCAA Men’s Basketball Tournament.
I want to say that I recognize I am neither an economist nor a statistician, and that both are real scholarly disciplines that people take years to master, just as I spent years mastering the skills involved in cultural interpretation. So perhaps I have something wrong here. If so I welcome corrections. I have not intended to mislead, but simply to find my way through a thicket of ideologies and numbers to get a sense of what the school I work for is doing in its contractual relationship with the students whose educational well-being and, in some sense, overall growth, I am entrusted to protect.
Lastly, a word on the term “adventure” in my title. I take it from Ian Hacking’s remarkable book The Emergence of Probability, in which, at the point in question, he describes four different kinds of “experiment” in early modern Europe. One of these is “the adventure,” which he describes as follows and in the spirit of which I have conducted my own little experiment: