While equal rights could be viewed as an issue of the past (at least in most developed countries), equal opportunity is still an open issue. In a free market wages and hiring decisions are not up to the law and thus equal rights are not enough to guarantee equal opportunity. While wages for equal work can be achieved with transparency and public pressure, it is much more difficult to quantify hiring discrimination. For this reason many people argue for a quota to actually affect these decisions with written law, so that injustices could be tried in court.

Violent Agreement over the Premise

While quotas seem contentious, both, proponents as well as opponents argue with the same premise: “hiring should be determined by merit, not demographics”. But proponents of the quota point to data showing demographics seemingly influencing hiring decisions right now, and want to correct them with a quota. While opponents argue that these decisions are due to individual preferences for certain types of jobs which might be correlated with demographics but are ultimately every individuals free choice. They might back that up with enrollment statistics correlated with demographics as well, where colleges and schools are subject to non-discrimination laws due to their public nature.

I do not really want to go to deep into the question which side is right, since that would at the very least require a deep dive into the data and might not even be answerable with the data.

So I am not going to. Why do I have any business writing about this topic then? Well, I guess it is in the title.

Blind Resume Quota

While it is possible to redact personal details from a resume - making it blind, it is basically impossible to do so in person (be it interviews or probation periods). So redacting personal details alone does not solve the problem. But now you have a better base distribution to compare hiring decisions to.

The fundamental assumption of a blanket quota is, that the chance that someone is qualified for a job is proportional to the size of their demographic. The idea is now, to use “passing the blind resume review” as a proxy for merit and demand that the chance that someone is hired given that they passed the blind resume review is independent of their demographic affiliation.

Quotas could be dynamically adjusted to this benchmark (with some margin for statistical error determined by a confidence interval).

The underlying assumption of this quota is of course, that there is no difference in the ability to write resumes which get accepted conditioned on merit. To me this seems like a much weaker assumption than a blanket “everyone has the same preferences” assumption.

Alternatives

I am not really married to the idea of blind resumes. But the idea to use a proxy for merit to detect actual discrimination would likely make quotas much less controversial.

To make an absurd comparison: While it would not make much sense to assert that the share of physicians in engineering jobs should be equal to their share of the population, it is much less likely that physicians will be underrepresented in engineering once one conditions on the number of accepted blind resumes.

Such a proxy should therefore be well suited to rule out preferences as a source. This ultimately means that people who had objections to blanket quotas might be perfectly fine with such a smart quota. At least if the proxy is decent.