Joanna Jacobs passes along Barry Schwartz’s recommendation that elite universities use a lottery for admissions (I couldn’t find the whole article):

There is probably a right answer to the questions “Whom should we admit?” or “Which college should I select?” But we won’t know until after the fact. Chance factors (roommate assignment, romantic successes or failures, or which English professor evaluates your first papers) might have a bigger effect on success and satisfaction than the tiny differences among applicants (or schools) within the range of acceptability. So once a set of “good enough” students or “good enough” schools has been identified, it probably doesn’t matter much which one you choose; or if it does matter, there is no way to know in advance what the right choice is.

College admissions is a crap shoot, and anyone who tells you otherwise is lying. Let colleges admit the all-stars, reject the losers, and show the people in the middle just how much randomness is involved. As long as students are using the Common Application to apply to multiple schools, though, the application becomes even more like a lotto ticket. Is this applicant applying to Cornell to hedge in case they don’t get into Columbia, or do they really want to go there?

The solution? Use a bidding market. Give each student who fills out the common application 1000 points and allow them to allocate them among schools they apply to. Weight students in the lottery based on the number of points they bid, so that students who bid more get more of a chance. People will still get into college - being in the “middle” group of Harvard applicants still makes you in the top group of many other great schools.

Bidding systems solve another problem as well. A friend of mine used to work in an admissions office, and she said they would look at other schools a candidate applied to and reject those who seemed highly qualified and applied to top tier schools because the office thought they weren’t likely to accept the admissions offer in the first place. Restricting the number of bidding points lets universities measure not only a student’s binary desire to attend signaled by applying (yes/no), but also the strength of their desire to attend the school (number of points bid).

Most people overestimate the role of going to a good college on life outcomes. They also overestimate how good admissions offices are at picking which people get in and which don’t. Exposing just how random it is, as Schwartz notes, will expose just how much “luck” is involved. Does that mean that a Harvard student and a Community College student are on the same intellectual level? Doubtful, but it will show the Princeton student that they could just as easily be at Duke, if only the lottery had been different.

Categories: Education, Incentive Centered Design, Information Economics, Information Markets, Matching Mechanisms, The Academy