Changes to Computer Science major are needed, but current execution is questionable

In recent years, the University of Michigan has been facing a problem common to many universities: too many students are enrolling in computer science courses.

To alleviate this, the University began by trying to hire new computer science (CS) professors, with limited success. Though the Department of Computer Science and Engineering (CSE) does have some new professors slated to begin teaching next year, Electrical Engineering and Computer Science course waitlists remain complicated, and faculty-to-student ratios remain low. 

It’s easy to understand why convincing established computer scientists to enter academia is so difficult, especially when the starting salary of programmers is so high. With such a high opportunity cost of entering academia, many would-be professors are choosing to stay in the professional world, leading to a national shortage of computer science professors.

Computer science is now the most popular major at the University, representing 11% of all undergrads. With more and more students enrolling in CS courses, CSE announced that new U-M students will need to apply to the major during their senior year of high school through the Common Application, making the entry process more similar to the Ross School of Business’s application method. This would help limit the class sizes in CSE courses, but it would also make the program far more competitive and less accessible to most U-M students. Though there would be an option to apply into the major as an “Enrolled Discoverer,” for those who find a love of CS once already at Michigan, the University has confirmed that these seats for Enrolled Discoverers would only account for a minority of CS spots. If a first-year student is admitted to the University but rejected from the CS program, they are disallowed from applying as an “Enrolled Discoverer,” an arbitrary limitation with seemingly limited utility for either thinning class sizes or choosing the best CS majors.

While there are reasons to critique this new University policy, the changes will have concrete benefits for the program. The University has long been praised for being one of the few top-ranked universities with an open CS major, but this hasn’t always worked to its advantage. Schools with closed CS majors, like the University of Illinois Urbana-Champaign (UIUC) , have been able to keep class sizes small and offer more concentrations and electives. U-M students, on the other hand, must fiercely compete for spots in upper-level CS courses, and no concentrations are offered.

Though James Earl Jones routinely proclaims that “we are the best university in the world” at every home U-M football game, the differences between the computer science programs at UIUC and the University of Michigan tell a different story. Because of the more robust CS curriculum at Illinois — made possible through smaller class sizes — UIUC routinely ranks in the top 5 computer science programs in the country, beating Michigan while having a notably higher acceptance rate. 

Restricting the Computer Science major would lower class sizes and allow more specialization within the major. However, though changes to the Computer Science program as proposed might increase the University’s prestige at a national level and facilitate the management of class sizes, there are potential downsides to this shift. 

While the application will be open to any student interested in completing a bachelor’s degree in Computer Science, there is an undeniable gap in the quality of high school education students across the country and at an international level receive, highlighting the socioeconomic gap between these groups of students and subsequently giving some students an advantage over others. Students who went to high schools that offered computer science courses will have the upper hand at the expense of those who never had access to the same resources. Even though the proportion of high schools that offer Computer Science courses has increased, disparities still persist. For example, Latino students are 1.4 times less likely to enroll in Computer Science classes than white and Asian students, thus reinforcing the status quo in an already fairly racially homogenous discipline. 

This, not even to mention the core question: how do we choose good computer scientists? High school applicants offer severely limited information to an undergraduate admissions committee. Getting an 800 on the reading portion of your SAT might impress an admissions officer, but what does it say about an applicant’s ability to build a neural net 10 years down the line? Specialization is good, and necessary in our modern economy, but we are sorely remiss if we intend to forbid students from studying whole disciplines at the age of 18.  

Moreover, limiting admission to students who are entirely sure they are going to major in computer science will limit the intellectual diversity of the major and, consequently, the types of computer scientists the University is sending out into the world. It is important to admit students who have interdisciplinary interests, which students who are still unsure that they want to study computer science are more likely to have. Knowledge in the humanities and the social sciences is necessary for jobs in the technology industry, as exemplified by demands for tech-savvy humanities professionals in Silicon Valley. That is not to say that every computer scientist needs to recite Hamlet in their free time, but rather that different, and incredibly valuable, outlooks are possible when a student draws from more places than just their core discipline. 

The restrictive application process may also run the risk of scaring students out of applying due to the rigor and high expectations of admissions. Many students might not apply out of fear that, because they have not already had a measure of CS education, they will be disqualified from studying computer science at the University for the rest of their undergraduate career. Remember, students rejected from the CS major when applying may not reapply as Enrolled Discoverers. On top of courses that already discourage students from continuing to pursue a specific degree program (weeder classes), the intimidating threat of not getting into the program initially, and therefore precluded from applying as an Enrolled Discoverer, forces prospective CS majors with less-than-stellar programming backgrounds into a cruel gamble.

In order to combat the challenges facing the increasingly understaffed program while maintaining the University’s commitment to equity, it is vital that the CS department explores alternative solutions in both the short and long terms. Since many classes are currently capped by the number of seats available in person, CS majors often find registering for classes incredibly challenging. Over the course of the pandemic, however, the CS department successfully experimented with remote options for courses, which often streamlined the core operations of classes by freeing resources to help with office hours and grading. In offering remote sections of each core course in the major, the CS department could allow more students to enroll in classes with minimal learning loss due to the existence of in-person supporting resources.

Though a closed major remains necessary until hiring can catch up, the CS department would also be well-advised to model its admissions process after the Ford School of Public Policy instead of Ross. While Ross admits a majority of its class as direct-admits out of high school and opens a limited number of highly competitive seats to students already at Michigan, Ford only allows students to apply at the end of their sophomore year. 

Among students, there’s some debate over the efficacy of each model. John Sader, an Engineering freshman and prospective CS major, told us that because CS, like Ross, is selecting for certain skills upon admission, “CS is different enough (from) other engineering disciplines,” for a Ross-type model to be appropriate. Kevin Ji, an Engineering junior in the CS program, on the other hand, argued that a Ford-like model is better since it would give students “a year or two to explore the major and decide from there.” For Sader though, what’s ultimately important is that the University is “transparent with the admissions process.” 

By restricting application to the major until students have completed the prerequisites necessary for it, the University can successfully limit the effects of the resource gap in CS high school education.

While hiring in the long run will likely catch up and make a return to the current program format possible, by using this moment to initiate a paradigm shift in the major, the CS department can enact positive long-term change. Though CS-adjacent programs like the School of Information (SI) and Data Science exist, the lack of a governing body over such majors makes sharing resources challenging. With a highly theoretical CS major, Michigan rigorously builds students’ foundations in the subject, but often fails to provide significant industry experience. SI, on the other hand, offers a curriculum highly relevant to the industry but suffers from a presumed lack of prestige relative to a CS degree. 

By creating a College of Computer Science in the mold of SI, for example, the University could offer a richer experience by offering concentrations, more CS-related majors and greater opportunities to explore electives. By investing in CS-related majors like data science and UX design, the University can cater to a wider set of interests and free resources in many current core classes. Additionally, because of the inherent crossover between majors, CS students in each major would be able to explore concentrations that overlap across disciplines, allowing for a well-rounded CS education.

Furthermore, this shift in program structure could also alleviate the hiring challenges the department faces. With more industry courses, a School of Computer Science could have PhD students and guest lecturers make up a greater component of its teaching staff. While most theoretical courses would still be taught by professors, many practical ones are better suited to the pedagogical style of those with industry experience. Overall, this long-term shift could prove tremendously beneficial to both the prestige and utility of a U-M CS degree, maintaining fair, competitive admissions standards while increasing the resources available to each student.

After struggling for years to maintain small class sizes and support the onslaught of new CS students entering the major, the University clearly needed to make a change in the program. While the closed major is a step in the right direction, it’s important that the University addresses the equity concerns stemming from this approach and works toward a more sustainable department structure in the long run.

By capitalizing on the breathing room a closed major allows the department, the CS administration should also work toward the long-term creation of a “School of Computer Science,” or other consolidated program, that would allow U-M students to obtain a CS degree while pursuing different concentration areas across the field. By doing this, the program could expand its national prestige and distribute resources more effectively to prevent future hiring shortages. 

Though there remains significant work to be done to improve the CS program, taking the bold step to close the major has the opportunity to accelerate meaningful change. In order to effectively accomplish this, however, the department should revamp its admissions process and restructure the current program format to alleviate pre-existing shortcomings.

Have thoughts about our pieces? The Michigan Daily is committed to publishing a diversity of Op-Eds & Letters to the Editor. Submission instructions can be found here.