Success on the athletic fields has no (meaningful) connection with alumni donations at schools like Williams. Consider this recent comment:
If Williams was the worst athletic school in/laughingstock of the NESCAC, do you think alums would be more or less likely to donate (number and magnitude)?
It would have no effect. Why is this so hard for everyone to understand? Partly, it is the natural human delusion that what you do matters to other people. Alas, most of the time, other people don’t care about you or the success of your team. It is not uncommon for a Williams athlete (or athletic supporter) to think that, because he cares a great deal about whether or not the men’s basketball team is 20-5 versus 5-20 that everyone else must care to. Turns out that they don’t. Read the literature (pdf).
One of our key findings is that the impact of athletic success on donations differs for men and women. When a male graduate’s former team wins its conference championship, his donations for general purposes increase by about 7% and his donations to the athletic program increase by about the same percentage. Football and basketball records generally have small and statistically insignificant effects; in some specifications, a winning basketball season reduces donations. For women there is no statistically discernible effect of a former team’s success on current giving; as is the case for men, the impacts of football and basketball, while statistically significant in some specifications, are not important in magnitude. Another novel result is that for males, varsity athletes whose teams were successful when they were undergraduates subsequently make larger donations to the athletic program. For example, if a male alumnus’s team won its conference championship during his senior year, his subsequent giving to the athletic program is about 8% a year higher, ceteris
One might quibble that this data is only for Princeton and that Williams might be different. (And, indeed, looking at the data for Williams would make for an amazing senior thesis.) One might question the quality of the paper. (The tables are pathetic.) One might suspect that they tried very hard to find any effect whatsoever. But the main lessons are obvious.
1) Women don’t care about athletic success, even on their own teams! (I, for one, don’t find this surprising.) A female Eph’s alumni donations are not affected by how her team did while she was at Williams, nor are they affected by her team’s subsequent performance.
2) Men don’t care about the athletic success of the teams that they don’t play on. (Truthfully, there is some chance that I have misread the paper on that one. Corrections welcome.) Again, hardly surprising. I don’t think that I have ever had a conversation with a Williams alum about the current success of a team that he did not play on, other than football and basketball.
3) There is no evidence that people care about the success of high profile sports like men’s basketball and football.
4) Men only care a tiny bit about the success of the teams they played on. 7% is, in the large scheme of things, close to a zero result. Letting in 5 applicants from billionaire families in just one year would dwarf any increased giving from this athletic effect. And, again, I doubt the size of that effect since I know that the researchers spent a lot of time looking for it.
5) What is the obvious cheat? A failure to report the results of the pooled regression. I bet that if you just look at the results for all students at once, that there is no effect. They had to split things up by gender to get any statistically significant coefficients. Moreover, they fail to discuss (?) the obvious problem. The effect is not a result of their team winning the Ivy League (which is their main explanatory variable) but of their team being mentioned in alumni periodicals (which, presumably, happens more to teams that win the league then to teams that don’t).
Our mininformed commentator wrote:
Williams will continue to place admissions emphasis on athletic excellence not because it cares, necessarily, about wins v. losses, but because it recognizes and appreciates that a non negligible percentage of alums care.
Sort of depends on your definition of “non negligible,” doesn’t it? The vast, vast majority of Ephs alumni (at least beyond the 5 year mark, i.e., after everyone they knew and/or played with has graduated) have zero connection to Eph athletics. They never see a game. They could not name a single player. They have no idea about any team’s win-loss record. Maybe — maybe! — there is a small effect for men who played sports but, even there, I am doubtful.
To the extent that any of this research (and this paper is broadly consistent with other articles in the literature) matters for Williams, the correct decision is to increase the athletics program in terms of participation. The authors write:
That said, there is no reason to believe that former athletes at such institutions fail to develop an affinity for their own teams—our results on the importance of own-team championships could very well generalize. To the extent that this is true and universities care about turning their undergraduates into future donors, it would seem that universities should nurture broad varsity athletic programs.
If there is enough interest to field a JV lacrosse team or freshmen soccer team, then do so. That may increase alumni donations enough to justify the cost. But don’t delude yourself into thinking that lowering admissions standards for star athletes in order to win more games in order to generate more donations works in any meaningful way. It doesn’t.