The Admissions Game

There’s a new paper by Raj Chetty, David Deming, and John Friedman on Ivy-Plus admissions that has been getting a lot of attention.

The authors had access to anonymized admissions data linked to tax records and standardized test scores. This allowed them to examine how the likelihood of admission varies with parental income at any given point of the score distribution. The following figure shows that conditional on scores, those at the very top of the income distribution secure admission at significantly higher rates (see here for the source of this figure, and here for the original version by the authors):

Undergraduate admissions at Ivy-Plus institutions is need-blind, and those making acceptance decisions do not have access to information on the parental incomes of applicants. So the pattern observed in the figure must arise from the use of selection criteria that value characteristics abundant among the very affluent. What are these criteria? The authors point to preferential admission for legacies, recruitment of athletes, and the value placed on certain non-academic credentials such as extracurriculars and leadership traits.

Some of the discussion prompted by the release of this paper appears to suggest that the admissions criteria adopted by these colleges are designed to result in the over-representation of students from the highest reaches of the income distribution. It certainly appears that recruitment for fencing or sailing advantages a very thin slice of the population. But I think this has the causality essentially backwards. If colleges were to start recruiting students proficient at ten-pin bowling or darts, you would quickly find these activities proliferating in private high schools and among the economic elite. Similarly, if colleges sought proficiency in Malyalam or Mende, it would not be difficult for those with ample resources to adjust.

Back in 1813, Thomas Jefferson and John Adams exchanged a series of letters on what would later come to be called meritocracy. Jefferson argued for a robust system of public education so that a “natural aristocracy” based on virtue and talents could be empowered, rather than an “artificial aristocracy founded on wealth and birth.” Adams was skeptical that one could so easily displace an entrenched elite:

Aristocracy, like Waterfowl, dives for Ages and then rises with brighter Plumage. It is a subtle Venom that diffuses itself unseen, over Oceans and Continents, and tryumphs over time… it is a Phoenix that rises again out of its own Ashes.

The findings in the paper appear to vindicate Adams.

How, then, might one move towards a fairer system? Abandoning legacy admissions and making recruited athletes go through the same selection process as other applicants (as MIT does) would certainly help. But an exclusive focus on academic credentials would create its own difficulties. Elite colleges don’t just want people who will ace their classes, they want future leaders in all manner of fields—scientists and engineers, founders of companies, celebrated authors, distinguished jurists, elected officials, civil rights pioneers, and so on. Academic records are too flat a criterion to identify potential across so broad a range of human endeavor. Furthermore, even academic credentials are resource dependent, and their proper interpretation must take this into account.

Recent work by Sandra Black, Jeffrey Denning, and Jesse Rothstein has examined the effects of the Texas ten percent policy, which guarantees admission to any state university to applicants graduating in the top decile of their high school class. The authors find that the gains to those who would not otherwise have been admitted are substantial, while the losses to those displaced by the policy are negligible. The implication is that there may be value to considering both absolute and relative performance as criteria for admission, in a manner that is difficult (though not impossible) to game.

Another possibility is the use of lotteries. Elite institutions get many more exceptionally well-qualified applicants than they could possibly accommodate, and at some point admissions officers are surely left wondering whether they should just toss a coin. Literally doing so may not be such a bad idea, and those rejected would take it much less personally. Robert Frank has pointed out that even if the role of chance in determining performance is small on the whole, it looms extremely large for those at the very high end. Making the role of good fortune explicit would result in a more honest and transparent system.

---

For those interested in a closer look at issues related to meritocracy, this forthcoming paper looks at the history of the concept, identifies conditions under which monotonic and group-blind selection criteria may not maximize performance, and surveys some relevant empirical evidence (including the Texas paper mentioned above, as well as related work by Zachary Bleemer and a fascinating study by Ursina Schaede and Ville Mankki that was discussed in an earlier post).