Delbanco's three cases for college
And what each one implies about CS education
In Chapter 1 of College: What It Was, Is, and Should Be, Andrew Delbanco describes three cases for college. In this post, I’ll consider what each one implies for CS education.
1: Economics
College is an economic investment in two ways. First, the median college grad has a higher annual wage than the median high school grad (by roughly $18K in 2022), so for individuals seeking higher salaries, going to college can make sense. Investing in education can also improve the strength of the overall economy. As President Obama once said, “[T]he countries that out-teach us today will out-compete us tomorrow.”
What does this imply for CS education? This one’s the easiest to answer. The obvious implication is that CS departments should help students improve their career-related skills such as interview preparation, résumé building, networking, acquiring in-demand knowledge, etc. There should be more emphasis on large programming projects and less on traditional written tests. Different courses could focus on different job titles, like “Web Development 101” or “Database Administration 201.” Colleges and companies could collaborate to design curricula, syllabi, co-op programs, and so on.
One catch is that, even if we accept “maximize the median graduate’s income” as the CS department’s only goal, I don’t think it’s obvious that the direct approach described above would be an optimal strategy. It might work in the short run, but perhaps in the long run, graduates would be better off if their college education had helped them acquire a better understanding of fundamental CS principles. Foundations have more staying power than trendy tools, and it’s hard to predict what students will need to know for their careers. Also, implementing this strategy would be costly since professors generally don’t have tons of experience in industry, and hiring instructors from industry forces colleges to compete (more directly) with companies for workers.
To be clear, I think it’s perfectly reasonable to attend college with the primary goal of making more money. But I agree with Delbanco that college fulfills other purposes too, and I suspect that even the most career-oriented student would benefit by reflecting on those purposes.
2: Democracy
Delbanco’s second case for college is that it supports our democracy. He cites Thomas Jefferson and John Adams, who despite their differences, agreed that an educated populace would be critical for the nation’s success. Delbanco claims that “this is more true than ever,” since “all of us are bombarded every day with pleadings and persuasions, of which many are distortions and deceptions” in the form of advertisements and political appeals. (Coincidentally, CS graduates have significantly contributed to the proliferation of these bombardments.) He goes on to say, “[T]he most important thing one can acquire in college is a well-functioning bullshit meter.”
What does this imply for CS education? This one’s trickier to answer. But I’d say that in many CS classes, we already hone our bullshit meters quite a bit. For example, in a typical algorithms class, the main activity is to propose an algorithm that solves some problem and persuade our peers (and ourselves) that it indeed solves the problem. Each step of the argument needs to be clear and logically sound — no bullshit.
A more direct way for a CS course to promote democratic ideals is to study civics problems that have a computational bent to them. For example, one area in mechanism design is on voting and considers questions like “How should we aggregate people’s preferences in an election?” Another field of study, which has a more geometric flavor, is gerrymandering: “How can we divide voting precincts into districts in a fair way?" and “What does it even mean for a division to be fair?” Questions like these help shed light on the challenges that citizens face.
Finally, as digital technology continues to proliferate, more CS departments offer courses on the social impacts and ethics of computing. These courses directly teach the relevant facts and potential consequences of these technologies, hopefully steering our society toward better outcomes.
3: Liberal arts
Delbanco’s third case for college is a bit nebulous; he admits that it’s “harder to articulate without sounding platitudinous and vague.” But you’ve probably heard of something like it before: a liberal arts education can open your mind to new possibilities, gratifying experiences, and greater fulfillment. Delbanco quotes Judith Shapiro (former President of Barnard College), who put it nicely: “You want the inside of your head to be an interesting place to spend the rest of your life.”
This sounds like something reserved for “the idle rich,” but Delbanco emphasizes that it shouldn’t be:
Surely, every American college ought to defend this waning possibility, whatever we call it. An American college is only true to itself when it opens its doors to all — rich, middling, and poor — who have the capacity to embrace the precious chance to think and reflect before life engulfs them. If we are serious about democracy, that means everyone.
What does this imply for CS education? This one’s the trickiest to answer because people’s interests vary wildly (even within CS majors), and properly inculcating liberal arts values requires an interdisciplinary approach. But one thing to note is that CS is, in my opinion, uniquely capable of impacting other fields; there’s an area called “Computational X” for many different values of X (e.g., biology, economics, even humor). So if a CS department wanted to take a liberal arts approach, it could collaborate with other departments to develop courses on a broader range of topics. There would also be smaller class sizes and more reading, essays, and discussions — these are all typical characteristics of liberal arts programs. Maybe there wouldn’t be grades (!).
Conclusion
Each of Delbanco’s cases for college has different implications for CS education. If we want to maximize graduates’ wages, it seems we should teach one way, but if we want to cultivate democratic and liberal arts values, we should probably teach a different way. Maybe there’s room to pursue everything at once; that’s certainly a popular strategy among colleges (e.g., via major and general education requirements). But regardless of what we do, I think it’s good to be aware, perhaps explicitly, of the tradeoffs that we make.