graduate school admission

The school admissions talk continues! Today, I had lunch with a friend, and once again, our conversation veered towards graduate school admissions. She hails from a different university and is currently a first-year master’s student. When girls get together, gossip flows, and this time, she shared her admission experience.

XXXL Xiao Long Bao. Strongly recommend it.

Universities are categorized into tiers: the renowned 985 and 211, the Double First-Class Colleges, and those without any specific title. These rankings are hierarchical. Generally, it’s believed that a university with the “985” title trumps a “211” university without it. Next come the Double First-Class Colleges, which lack the 985 and 211 titles, and finally, the “no-title” colleges. When applying to graduate schools, students often aim to climb the title hierarchy, attempting to move from a 211 to a 985 institution.

Universities have the liberty to set their own admission criteria for graduate programs. They strategize by organizing summer camps simultaneously with competing universities or by discriminating based on the student’s undergraduate degree’s title—whether it’s from a 985 or, perhaps, a no-title institution.

According to my friend, who comes from a no-title institution but gained admission to a 211 university, it’s common knowledge that different universities have distinctive admission styles that vary in terms of discrimination. Furthermore, these styles tend to be consistent and can be discerned by examining past admission data or relying on anecdotes shared among students.

In essence, universities commit to certain admission strategies, and these strategies are known to applicants. For instance, during our conversation over a delightful meal, we discussed how a 211 university in Shanghai is famous for not discriminating against students from no-title universities, while an almost equivalent 211 university in Beijing is notorious for rejecting students without 985 backgrounds, regardless of other qualifications. We playfully criticized the latter institution for its strategy, which seems unfair and shortsighted—establishing a poor reputation, possibly admitting lower-ranked 985 students (well, because it’s a 211), potentially attracting opportunistic applicants, and ultimately missing out on potentially outstanding students from non-985 universities.

Why do they do this? Perhaps the admission committee is risk-averse, or they are simply unwilling to sift through numerous resumes from no-title students. But the more intriguing question is whether schools truly benefit from discriminating among applicants with various titles. Without actual comparison statistics, we cannot draw definitive conclusions. I would argue that the answer is quite the opposite.

Consider two 211 universities, $\mathcal S$ and $\mathcal C$, competing to admit students from a pool of applicants from 985 or 211 universities. The students’ types and distribution across different universities are as follows:

Both $\mathcal S$ and $\mathcal C$ aim to admit 30 students to their summer programs. Assume that students can attend only one summer program. Before admission, their types are unknown, but after selection, their types become clear. Type A students are eager to secure a spot in a summer program, while Type B students are indifferent since they know they’ll rank lower than Type A students once their types are revealed. Universities prefer having as many Type A students as possible. However, there are only 54 Type A students in total, so the two universities will have to compete for them.

In the absence of any signals, the universities would simply admit all applicants from 985 universities and expect to have 15 Type A students each. But what if, for instance, university $\mathcal S$ strongly signals that it won’t discriminate based on titles? In that case, all 24 Type A students may opt to apply to university $\mathcal S$, shifting the ratio from 0.4 to 0.57—surpassing that of 985 students. University $\mathcal C$ would face a less favorable pool of 211 applicants, whereas university $\mathcal S$ already has a decent pool of 211 students to select from.

It’s a simplified model, a product of a post-lunch, drowsy brainstorming session. Nevertheless, the underlying idea is that lower-tier universities may have a smaller proportion of outstanding students but a significantly larger population. By strategically signaling a reduced level of discrimination, these universities can aggregate higher numbers because the admission process essentially involves probabilities and gambling on the quality of students based on partially revealed information. Students are strategic and sensitive to signals and policy changes. It’s possible that when some universities do not act rationally (such as discriminating against certain student groups for various reasons), other universities may reap the benefits of being more astute.