why a phd may just not be worth it, redux
With grad school finally winding down, I have a choice to make. Should I take a short break after my master’s, start drafting a CV, and applying for a PhD program, or should I just stick with what I have? On the one hand, I have professors who say that since I have the grades and they can see me doing research full time, finishing the process of becoming a proper computer scientist is a no brainer. But on the other, the economics seem quite set against more schooling. In fact, as mentioned last year, having a doctorate yields no premium at all for science and technology degree holders, and for the educators and businesspeople for whom it does have some economic benefit, the increase salary is only 3%, which is actually less than your standard cost of living adjustment in the grand scheme of things. And considering that most companies don’t want to pay for a PhD in computer science, and government labs that employ them are facing budget cuts, I have every incentive to stay put and enjoy the flexibility of being between a career-minded undergrad and a research scientist.
It all comes down to money and options. Median annual earnings for programmers are $85,430 and rising, and the employment outlook is widely considered to be excellent. Currently, there are over 1.3 million jobs on the market for them and another 300,000 or so are expected to be created in the next seven years. For your average computer scientist, the annual salary is just over $97,000 per year, but the job outlook is not nearly as cheery with less than 29,000 jobs available. Even with the predicted 24% annual growth in the profession, the next seven years will only see another 7,000 new positions, meaning that all the jobs meant for computer science PhDs will amount to just 0.2% of the employment opportunities for industry programmers. And this is not just a problem in computer science. Science bloggers who promote STEM majors often fail to mention or even investigate this concern in the scientific fields they promote, but it’s there. Sure, we could simply point to interesting scientific professions and show that they do pay a decent living wage for those employed in them, but we also have to point out that there could be only a few thousand such jobs even available, and growth in the field is not just slow but outright glacial. Neglecting to do that is a major disservice to our readers.
Now don’t get me wrong, it’s great that up and coming programmers with a passion for turning code into tools that solve a wide variety of problems are out there and want to do research, and it’s terrific that we have some wonderful explanations of what they’ll do in industry vs. academia, like this one from Mark Chu-Carroll. But if we take a moment to pick on Mark’s summary, nowhere do we see mentions of tenure and the competition in the academic application process. You may think that because you really love what you do, you’ll be one of the exceptions and get that research job with its freedom of experimentation, and that may well happen. But you’re also very likely to encounter someone even more driven or more resourceful, or just better mentored in direct competition with you and it would be most unwise to assume that you’re the smartest and most talented in a very, very large and diverse applicant pool. Of course the industry’s competition for jobs can also be very fierce but there you can specialize and branch out to become an expert in certain in demand tools, making it easier to get a foot in the door and pass a tech screen. Since industry needs to employ far more programmers than academia, you start with much higher odds of getting a job and beginning to manage your career, though you must keep in mind that working in industry means you won’t get very much freedom at work.
That’s the sad tradeoff we see when we no longer incentivize grad students to become scientists. Politicians and the military call for more STEM graduates and try to promote excitement about math and science. But at the same time, the very same politicians call the same subjects they supposedly want to promote a waste of time and effort, and use scientists as convenient scapegoats for partisan gain while eagerly cutting the already small budgets of research labs to pretend that they’re actually doing something about deficits. Rather than encourage students to become scientists with something other than rhetoric and create places for them to actually carry out their experiments and new jobs, the political establishment gladly sacrifices their future in partisan debates. Instead of being able to take their new skills and just see with what they can come up, they are given a choice between a poorly paid chase for eventual tenure during which they’ll have to pay massive salaries for administrators, or giving up a good deal of intellectual freedom until they advance to senior levels in industry and either find employment at a very large company which will allow them to experiment, or start a company of their own, delaying their potential breakthroughs. And that’s what happens when we ask for more scientists but refuse to give them more jobs. We’re bound to end up with fewer of them as a result.