The Fairness Of Preferential Admissions To Colleges

A re-posting of a comment I wrote back in 2010 on opportunity and outcome fairness.

I do not think three people could agree on what "fairness" means.

In golf and horseracing, there is handicapping, which is an attempt to remove beginning advantages to give all an equal chance of winning.

In the US, philosophically, we want to believe that all groups are capable of the same results, i.e. the same percentage of doctors, lawyers, police officers, firefighters, millionaires, successful entrepreneurs, CEOs, college graduates, homeowners, etc.

If one believes handicapping is "fair" to undo socio-economic or competitive disadvantages, then group outcomes should be statistically equal. For these believers, unequal outcomes at the group level indicate ineffective handicapping. It indicates a need to improve the "fairness" of the handicapping process to produce equal outcomes.

If one believes, "equal treatment" [equal opportunity] does not allow for handicapping, does not include school admission preferences, affirmative job action, remedial help, extra time on tests, etc., then equal outcomes will not occur because family socio-economic status predicts many of the outcomes and groups sort by socio-economic status in the US.

Different people can disagree whether "fairness" of equal treatment includes "handicapping".

[Subgroup Versus Whole Group Preferential Treatment And Outcomes]

There is also often a failure to distinguish the characteristics of the subgroup at the hiring or acceptance level from the characteristics of the whole group, and we often focus on subgroup "fairness" in place of group "fairness".

For example, suppose two groups, A and B, are very similar, except the top ten percent of Group A want to become lawyers and the top ten percent of Group B want to become doctors. What is a fair outcome at the top medical schools and top law schools?

By measurements of skills, test scores, grades, etc, we should expect more Group A's to be admitted to top law schools than Group B's. Additionally, we should expect more Group B's to be admitted to top medical schools than Group A's. Should medical and law schools allow the disparate outcome or should they equalize their admission rates for Groups A and B?

Different religions, race, country of origin, socio-economic status etc. groups rank careers and other aspirations differently and those rankings within the groups will sort each group differently. If one group ranks firefighting as a career over retail store management and the other group ranks retail store management higher, is it fair to want each job category to contain the same percentage of each group? Are the gate keeping tests and hiring processes that result in different outcomes "unfair" when group career preferences differ? Is it "unfair" if different career choices have different total life earnings effects?