Consider the abstract from “Athletics and Alumni Giving Evidence From a Highly Selective Liberal Arts College” pdf by Jessica Holmes, James Meditz and Paul Sommers. (Holmes and Sommers are Middlebury professors. Meditiz was their student.)

Using data on annual giving (between 1990 and 2004) for more than 22,000 active alumni from a highly selective liberal arts college, the authors employ a probit framework to analyze the likelihood of giving and a tobit framework to analyze the determinants of alumni generosity. Both the micro-level analysis and the statistical methodology allow the authors to test for differential impacts (by gender, age, or undergraduate involvement) of sports participation or a winning season on the propensity to give as well as on the generosity of alumni contributions. The results indicate that athletes are more likely to give and that they are more generous than their nonathlete counterparts, especially younger alumni who participated in one of the college’s historically most successful high-profile sports. A winning season in this particular sports program also leads to greater alumni giving and more generous gifts.

Keep in mind that we all agree that athletes tend to give more than non-athletes. This study confirms that, but it is nothing new. Now, there is an interesting discussion to be had about why that might be so, yet that discussion is not particularly relevant to Williams policy going forward. The central dispute is:

Does providing athletes with significant admissions advantages generate greater donations?

There are two main mechanisms by which such an effect might occur — assuming that more admissions advantages lead to better athletes lead to more wins/championships.

First (and this is covered by the Meer and Rosen (2008)), athletes on a given team might give more when either their team does better now or their team did better while they were at school. Meer and Rosen (2008) show clearly (in my view) that there is no such effect, or that the effect is too small to matter.

Second (covered in Holmes et al. (2008) here), all alumni, including both athletes and non-athletes might give more when specific teams (here football and hockey) do well today. Holmes et al. do not look at success in any other sports besides football and hockey, so, obviously, we need to be careful about generalizing to sports that have essentially no fans other than the parents of current athletes. Moreover, Holmes et al. find mixed results:

Football success decreases alumni donations!

Interestingly, whether one uses league title or winning percentage, football success translates into lower propensities to give; in years in which the football team wins a title, alumni are 7% less likely to give, and a 10-point increase in the winning percentage is associated with a 1% reduction in probability of giving.

Note how they leave this fact out of the abstract. That result alone should cause all the tips boosters at EphBlog to take a step back and re-evaluate. If the single sport that is the highest profile (most athletes, most fans) has a negative correlation between success and donations than we ought to rethink everything. And that is all the more true since football requires, by far, the largest number of significant admissions preferences.

It is true, on the other hand, that hockey success is correlated with alumni giving. Yet we will leave the details of that result to another day.

UPDATE: Just to be clear, this post is somewhat teasing. The are so many flaws with the analysis (see here) that it is highly, highly doubtful that football success decreases alumni donations. After all, have you ever met a Williams alum who a) Followed the football team closely enough to know their win/loss record and b) Gave less money, less often when the record was good? No. It is absurd.

Athletics success, whether current or past, whether in high profile sports like football/hockey or low profile sports, has no connection to alumni generosity.

Print  •  Email