Yes, Supply and Demand Applies to Computer Science Degrees
it's still a great field, but "learn to code" has created headwinds
For a long time, I’ve been talking about the importance of supply and demand in the labor market. It was a key part of my first book; the effort to push everyone into college is a bad idea, among other reasons, because flooding a market with something (like college degrees) inevitably degrades the value of that thing. We know that this applies to degrees in general. For like the twentieth time, I’ll share this paper from the NBER. Apologies to people who’ve gotten this shpiel from me before, but I really see it as something of a skeleton key for the past 50 years of higher education and the labor market. Looking at 115 years of data, the authors conclude, “Overall, simple supply and demand specifications do a remarkable job explaining the long-run evolution of the college wage premium.” The more jobs for college degree holders out there, the higher the college wage premium; the more degree holders out there, the lower the college wage premium. Which, duh.
I think it’s clear that this also holds for specific degrees. If there are more workers with veterinary science degrees graduating, they will compete against each other in the labor market and drive down entry-level wages. The poverty of the “practical major” idea becomes clear when you look at outcomes for business majors. It’s hard to think of a major that better satisfies the vague criteria of practicality than business. And yet business majors only do alright in terms of employment rates and income, not great. Why? Because a major’s value is not the product of some Platonic ideal of practicality, but of supply and demand! The trouble for all of those business majors is that we graduate like 350,000 of them a year. All of those young workers compete against each other for the same jobs. Employers therefore have little incentive to raise entry-level salaries too much, as there’s always another potential worker to hire. “Practicality” makes no difference.
For a long time my go-to example of this phenomenon was pharmacy. Now, finally, I can go after the “learn to code” crowd, at least a little bit. This New York Times article details an increasingly competitive labor market for recent comp sci graduates, with many tech companies slashing jobs at the exact same time as more and more graduates pour into the field.
Over the last decade, the prospect of six-figure starting salaries, perks like free food and the chance to work on apps used by billions led young people to stampede toward computer science — the study of computer programming and processes like algorithms — on college campuses across the United States. The number of undergraduates majoring in the subject more than tripled from 2011 to 2021, to nearly 136,000 students, according to the Computing Research Association, which tracks computing degrees at about 200 universities….
…cutbacks have not only sent recent graduates scrambling to find new jobs but also created uncertainty for college students seeking high-paying summer internships at large consumer tech companies.
In the past, tech companies used their internship programs to recruit promising job candidates, extending offers to many students to return as full-time employees after graduation. But this year, those opportunities are shrinking.
Let me say this right up front: computer science majors still have enviable outcomes in the job market. A skilled and experienced coder will enjoy employment conditions that are among the best in the economy. Someone who really knows how to program is going to get a job somewhere. I’m not denying that. I am saying that dramatically increasing the number of computer science graduates we pump out will inevitably result in headwinds for applicants, making it harder to secure a job and reducing their initial wages. Yes, it’s great to be a professional programmer, I have no doubt. But contrary to what some hackers have angrily demanded in my comments section, computer science is not exempt from the supply and demand considerations that affect all educated labor. And as the interviews in that Times articles show, there’s been a culture in the field that has created a sense of entitlement among some, an expectation that good economic times are preordained.
Again, because people get very sensitive about this topic, this is about headwinds; it’s about salaries and employment rate on the margins; it’s about a very strong employment market that’s still very strong, just more competitive. And there’s a natural cap on the number of people who can enter the field: it’s hard to code. Not everybody can do it. It takes a certain degree of talent, and most people don’t have it. It’s a cognitively-loaded field in world where not everyone has the same cognitive gifts. (This was the other major reason sending everyone to college is a bad idea that I discussed in the book.) Certainly getting a computer science degree is a high-percentage play when it comes to getting a good job. But tripling the number of people you’re sending out onto the job market each year has consequences. It just does.
I’ll ask the people who’ve been shouting “learn to code!” for ten years again: what did you think would happen once people actually did?
Please check out my appearance on the Converging Dialogues podcast.
I can assure you that business is not a practical major.
The galling aspect of "learn to code" was when it was thrown at unemployed older workers. Like it's an easy retraining option, and tech firms are just racing to hire 53 year olds.
Meanwhile the competition is every programmer in India and China: one element of work from home since covid is firms realizing "home" can be Bangalore.