College Admissions Are Broken And In Need Of Innovation

It happens every spring. The admissions office emails arrive and everywhere students find themselves celebrating, crestfallen, or mystified as the conversation inevitably turns to how broken the admissions process is. A quick Google search shows the number of articles on the topic has ballooned from roughly two a year in the 1990s and 2000s to 30 a year in the early 2010s, to over 300 a year in the 2020s. There is widespread discontent, whether due to scandals or lawsuits or the increasingly arbitrary feel of decisions. The college admissions process is ripe for innovation.

The Traditional Application Process Is Poorly Designed

Applying to college mimics a poorly designed course. Students are asked to complete and submit all of the required work (i.e., the application) simultaneously, and after it is evaluated, they are told only if they passed or failed; no meaningful feedback is given, no opportunities for course correction are presented, and much time is wasted as students spend countless hours on applications that colleges reject in minutes.

The Signal-To-Noise Ratio Is Poor

Signal-to-noise is a problem that has been getting steadily worse. With innovations such as the common application and other electronic submission systems, it becomes easier, if not less expensive, to apply to more colleges. Consequently, the average number of applications students submit steadily increases while the size of admissions staffs remains more or less constant. The result is a significant decrease in time spent reading applications. Potential solutions to this problem, such as machine evaluation of student applications as part of a presorting process, remain elusive while raising ethical concerns.

Colleges And Students Are Not Good At Selecting

Overall college retention rates are 82%, meaning almost 1 in 5 students will not return for a second year, and only 64% will complete their degrees within six years. If admissions is a process of selecting students for fit and likelihood of completion, there is room for improvement.

Data shows that an institution’s ACT score profile (median ACT, or even median of the bottom half of ACTs) is a better predictor of 6-year graduation rates than the selectivity of its admissions process. Given that ACT scores are factored into admissions, this raises questions about whether colleges are looking at the right things during their holistic admissions processes.

While colleges may do a poor job selecting students, there is little evidence to suggest that students are doing a better job choosing colleges. The tools students have available, such as the ranking services, have problems that have been well discussed elsewhere.

Where The Value Of College Lies Is Unclear

While the value of college might be seen as the knowledge students gain, Nobel Laureate Kenneth Arrow described an alternative in 1973. Arrow argued the value of college was not in the development of human capital but as a sorting signal for future employers. College admission sent one signal, and successful completion sent another.

At the same time, the emergence of job-specific boot camps such as App Academy show the decoupling of college from practical knowledge is already underway. Between free online courses and career-focused alternatives, the signaling value may be the most important aspect of higher education.

Innovation Is Possible, But Its Impact May Be Limited

The most significant admissions innovation was the publication of the deferred preference algorithm 60 years ago. While it has not transformed undergraduate admissions, it has played a role in how spots are allocated for elite high schools in Boston and New York, has been the driving force behind how medical residencies are assigned, and led to a Nobel Prize in economics.

For this process to work well, however, students and colleges must produce informed preference lists. If these lists are generated from poor signals the result will likely be no better than what is currently happening.

Other promising innovation avenues exist. Two platforms — Greenlight Credentials and Concourse — give students more control over their information, effectively creating a market where colleges make competing bids for students. Argus promises a new college applications approach using various tools to collect deeper insight into student capability and interest, providing a better signal for schools to make decisions about which students to pursue. Assuming these tools improve information flow and quality, they may lead to improved efficiencies and help colleges focus on students who are interested in them, hopefully decreasing false positives while also improving ultimate outcomes.

Whether these or other innovations can provide a better college admissions approach remains to be seen. If nothing better emerges, we risk exponential growth of the dissatisfaction with the status quo.