Ever since I began watching data about computer science graduation rates, I’ve been wondering when the bubble is going to burst.
I started out by digging into the percentage of women graduating with CS degrees. I found some Fascinating stuff in our software pipeline numbers, like the fact that the percentage of women in CS peaked in 1984 at a nearly unimaginable 37%, then dropped to a steady-state of about 27% for over a decade, then dropped again in 2003 to a new steady-state of about 18% for another decade. Definitely some social influences going on in the rate at which women filled the CS pipeline. It has been a few years since I posted data on this; here’s the most recently published data about where we are.
Percentage and number of women in computer science, 1971-2021 (larger image)
We’ve been slowly clawing our way back upward to about 22% since 2015, but I’m still mad that we haven’t even recovered to where we were when I got my own degree in 1999.
But aside from gender differences in the pipeline, the other thing that’s starkly obvious is the growth in the number of computer science graduates we’re producing. Amazing growth. In 1999 when I graduated, the U.S. produced about 30K CS degrees; in 2021 we crossed above 100K. Since 2009 the number of CS degrees has been rising fast. That’s the bubble I’m talking about: eventually it’s going to be more than the tech industry can absorb.
Has that CS degree bubble burst, yet? I don’t know. The tech industry is in an employment down-turn at the moment, with many layoffs and few open jobs. But the story we all tell ourselves right now is that this is driven by economic factors, especially a drop in advertising revenue. An upturn is coming Real Soon Now, once our CEO/shareholder overlords release the purse strings again, right? I can’t say whether we’ve hit a temporary change or a more permanent one. But certainly it’s the first big tech employment downturn since the dot-com bubble burst in 2001. In 2001 it wasn’t a large-scale disaster; most people were back on their feet in less than a year.
As an aside: I’m curious if the COVID years and their impact on the tech industry are changing the CS graduation rate. We won’t know for another year or two, since the freshmen who were choosing majors in 2020-2021 haven’t graduated yet. Even more than that, I’m curious about the impact of this tech down-turn. But we won’t have that answer for another 4-5 years.
Speaking of the dot-com bubble burst… Remember from the graph above, how gender diversity in CS degrees took a big drop in 2003 at the same time the CS degree count declined? If the current tech downturn leads to another decline in CS degrees, we’d better not lose ground in the diversity of our CS population! Don’t let those under-represented college freshmen get scared and opt out right now!
What’s going to happen when the CS degree bubble bursts? Wow, that’s a significant question for the tech industry and the people in it. I can think of a number of potential impacts:
- Tech industry compensation will stop growing, and maybe even drop. I doubt base pay will drop much – that’s hard to do when you already have tens of thousands of employees at major companies. But future bonuses and future stock grants are easy to shrink.
- There might be pressure / incentives for the older, more experienced (expensive) tech talent to retire early – to make way for younger, cheaper talent.
- On a more positive note: with lower compensation, tech talent would be more accessible to a wider range of industries than ever before. Suddenly more companies – non traditional “big tech” companies – will be able to hire more people with tech skills. I’d like to hope that this will lead to a new tech Renaissance of a sort, where quality software skills get applied to a wider set of problems than how to make the next social media. That would soften the blow on tech talent, since there would still be lots of jobs around, and hopefully a wider variety of areas to bring impact and meaning to a career.
- It’s going to be harder for immigrants to find jobs in the U.S. tech industry. We’ve been so under-staffed with American talent for so long that a really large portion of my co-workers are immigrants; I’d easily guess at least 25% (I see it as a good thing!). If we reach a point where we produce more CS grads than the tech industry can absorb, the U.S. government will probably give out a lot fewer visas for tech work. This will be a disaster to many immigrants, and a real shame for the diversity of the American tech talent pool. But I hope it similarly leads to a new tech Renaissance in other countries, too.
- On first thought, I was guessing that big tech companies will get even more profitable if they can compensate employees less. š On second thought, I am wondering if instead we’ll see a proliferation of competition from smaller tech companies. I can imagine both things happening, but I wouldn’t bet against the power of Big Tech to come out on top.
- Maybe we’d finally have enough high school CS teachers to go around, instead of having to augment with industry employees who teach on the side?
This might sound like a gloom-and-doom prediction, but it’s not. Software skills are as critical for our future as ever. We will continue solving problems with technology for the foreseeable future. I don’t think the jobs are going to run out. I’m actually saying they’re going to get more common, and more diversified. Pay might be lower, or at least stop getting higher. But the impact a CS grad can have is going to grow, because there will be more opportunities to solve problems in more industries.
That’s what I think. But, I’m nobody special, what do I know? š¤·āāļø We’ll have to wait and see.
