How do you make fair admissions decisions?


It’s that time of the academic year again: The semester has just started and every class is immediately overwhelmed by student demand. On top of it, coordination among the teaching staff could have been better. Classes cluster on Mondays and Wednesdays, limiting our students’ flexibility when building their schedule. Put some strategic behavior on the part of your students into the mix and you’ll easily end up with fifty or more applications for a twenty-five to thirty students class. It’s a mess. So, how do you decide who gets to participate without stepping onto everyone’s toes? Alternatively, how do you democratize access to scarce resources?

Hard criteria are few and far between

Students who endure hardships1 are usually admitted, but their numbers are small. Also, and at the risk of kicking in an open door, what counts as a hardship? Some students must take my class because their study regulations demand two seminars from the same area to conclude their electives module. They have no alternatives as long as my class is the only one offered in that module. Does TINA count as a hardship?

What about good old “first come, first serve”? In fact, several students argued for their admission based on the fact that they registered on five past twelve a.m. right when admissions started. Fun fact: Our campus management software does not tell me who signed up first. More importantly, technology can be tricky. An occasional glitch in your internet connection may easily move you from first to the last position on the list. In my opinion, “First come, first serve” is neither fair – So, you stayed up late? – nor robust.

What if students were to rank their preferences? At Potsdam University, students may express priorities when they sign up for class. In principle, lecturers should admit priority students first and distribute any remaining seats among everyone else. Again, there is a serious drawback: How do you select within each group? For instance, 49 students registered for my class – every single one of them prioritized it. So, I am back to square one.

Are fair procedures the answer?

Apparently, you can’t just look at students to make the call. Maybe there are procedures which lead to fair results? Cues, auctions, and lotteries are on the table.

Some colleagues rely on cueing. They first admit students late in their studies. Any remaining seats are then filled with earlier semesters. Justifications of this method boil down to urgency: Older students must transition into the labor market as soon as possible. But why should an eighth-semester student feel more pressure than a seventh-semester student? If you are unable to justify this distinction, then you won’t be able to justify any other. Moreover, I believe cueing perpetuates the problem. If all students have to wait their turn, then all of them will eventually be eighth at some point.

Auctions would be preferable. After all, students who are most motivated to participate in my class should place the highest bids. In the ideal world, I would have the honors to teach a group of self-selected geeks. That is, if – and, unfortunately, only if – I could elicit sincere offers. I could require students to bid from their own resources and -rightly – face the wrath of the university’s ethics committee. Alternatively, I provide them with benefits to burn and – rightly – face my wife’s wrath. In short, without institutional backing auctions may seem ideal but they are not a realistic option.

That leaves casting lots. Lotteries have been around for ages. In ancient Greece, for instance, many public officials were chosen by lot based on a firm belief in the equality of all citizens. Lots make decisions without consideration of person or intent. In that regard, they are uniquely fair. On the downside, students are not all equal. After all, hardships are all about significant inequalities. Lotteries discard such circumstantial information.

Moving forward

This is the upshot so far: Fair admissions cannot be based on a single criterion or procedure. Rather, some combination of criteria and procedures is required to avoid discrimination against any group of students. Here is how I tried to solve the problem.

  1. I decided to accept 35 students. That is 10 more than originally planned and puts me at 140 percent course load.
  2. To fill those 35 seats I set up a lottery. It assigns a higher probability to older students. After all, I do subscribe to the argument that older students experience enormous social and economic pressures. However, given my ignorance about their course of studies and given the mounting pressure from younger cohorts, age does not justify deterministic admission.2
  3. Students who claimed exceptional circumstances, e.g., small children, were additionally admitted.3

In the end, 39 students were admitted to class. That’s quite a lot, but still manageable based on my past experience. Attrition will do its work and eventually reduce class size to about 30.


Every faculty member fights the admissions battle on her own, no solution is ideal. By combining different admissions criteria and processes I strove to democratize admissions as much as reasonable. The result is unavoidably imperfect. However, in light of my own limited ressources, I believe it constitutes a compromise that will be endurable for everyone.

  1. E.g., pregnancy or small children, chronic diseases, care-dependent relatives, physical and mental challenges.
  2. Each semester is divided by the summed sequence from 1 to the highest, non-hardship semester (8).
  3. Our chair sidestepped the study regulations complication by offering an additional class. Unfortunately, students were not too enthusiastic about that class.

Leave a Reply

Your email address will not be published. Required fields are marked *