How Joe Stiglitz Explains Texas A&M Football

Applying economic reasoning to disparate topics

One of my pet interests is college football. I’m not entirely in control of that. Both my parents went to USC, so I was born and raised a Trojan. The NCAA, which regulates college sports, recently changed the rules to allow players to collect compensation for their name, image, and likeness. I think the intent was to allow players to sell endorsements, but since boosters of programs could pay players for very little endorsing indeed, players are now paid. 

The rollout of NIL across programs was uneven across programs. Some were fast to spend millions, while others, like my Trojans, were slow and uncoordinated. Texas A&M was one of the first programs to lean heavily into it, spending somewhere around $30 million to sign the number one recruiting class in 2022. Yet, two and a half seasons on, the results have been frankly bad. The team as a whole went 5-7 in 2022, and 7-6 in 2023. As of 2024, almost half the class is no longer with the team. What went wrong? I think the answer can be found in a classic economic paper, “Credit Rationing in Markets with Imperfect Information.”

Stiglitz and Weiss (1981) are concerned with credit markets under imperfect information, and in particular, why credit might be rationed. After all, it is a basic tenet of economics that price can be raised until demand equals supply. If there are more customers than can be served at a given interest rate, why wouldn’t banks simply raise their rates? Their key insight is that different borrowers have different levels of risk, and this riskiness is not evenly distributed. As the price is raised, safer borrowers will drop out, and the bank is faced with a different set of customers than they would have had. It isn’t possible for them to reach the efficient amount of capital lent out and still be profit maximizing, even with the constraint that they price at marginal cost.

So let’s apply this to Texas A&M. Schools offer two things: money up front, and development to get to the NFL. Recruits have private information about their desires, and how likely they are to reach the NFL. Those who are unlikely to reach the NFL, likely (because physical gifts are easy to measure) because they’re lazy bums, will prefer money now to money later. By increasing the price they were willing to pay for recruits, they selected for worse players on the margins they couldn’t measure. This is why putting money into paying players may not be optimal. Instead, you want to put your money into building training facilities and hiring coaches, because those select for a better average quality. If you are going to spend money on players, you should want to spend them on graduate transfers, because they have had the time to reveal their type. 

This needn’t mean that one couldn’t improve outcomes by paying money, nor should we assume that the global maximum is the same as the low-spending local maximum. It would certainly be possible for someone to spend preposterous amounts of money and get the best players, even with the sorting. The trouble is that they would have to exceed the net present value of the future income streams, if they have no other way of discerning players types.