First, I have a STEM Ph.D. and am in my 35th year at a U.S. corporation. Now my question: Just where do you expect all these newly minted Ph. D. grads to do that research?
At least in the US, the last two decades have not been kind to the once iconic corporate research labs. As global competition increased and the push for constant high quarterly returns by investors, more and more has been off-shored. The most recent casualty was DuPont CR&D but there has been constant erosion.
My colleagues and I have actually lost count of the number of downsizings we have been through.
Here is a test for you: Next conference you go to, look at the talks and posters. How many corporate people are there who are not instrument vendors or contract labs?
I'm not sure I get your point though. I said the pursuit of Ph.D makes sense if you'd want to do research. I never said it's a good return on investment or you'd have a job or you'd end up doing research in the industry.
Is the rise of more efficient competitors overseas the main thing that's made these labs disappear, or has something else more internal to these companies or to the domestic economy changed over the last 30 years?
No! Absolutely not. A PhD is a huge time commitment. It takes a whole chunk out of your life which is one of the most important times for building a career. You absolutely definitely should not do a PhD just because you think you'd enjoy it.
It really isn't such a great option because it has a huge knock-on effect on your future. You should only do a PhD if you either (a) are committed to having an academic career or (b) are in a field where there is a clear alternative career path. Doing a bit of internet stalking (sorry) I see that you have a CS PhD. That is one of the fields where there are many non-academic jobs available for PhDs. In my field, it's academia or bust, which makes the calculation of whether or not to do a PhD a very different one. If you're giving generic advice on whether or not to do a PhD, it's worth bearing in mind that CS is really the exception rather than the rule in this respect. In most fields, "do a PhD if you like doing research" is terrible, terrible life advice.
There is no career for most PhD students. This is the whole problem with how people look at it. If it were the case that most people with PhDs could obtain stable, low-paying academic jobs, then you'd have a point, but most people with PhDs can't do that.
I agree with you that counting on a stable job in academia is probably a mistake for most PhD students. But your previous categorical statement goes too far.
It's a perfectly valid choice for me to decide to do a PhD for no reason other than an intense desire to learn more about some topic and then plan on making a living as a kayaking instructor or elementary school teacher or a handyman for the following two or three decades. As long as I have no delusions about the feasibility of my plan.
Of course, if you want. But surely it's unwise to actively encourage someone along that path. If it's what they really really want, they'll do it anyway. Advice is directed at people who are undecided.
except there's nothing like the rigour of a good PhD to teach you good methods, stats and epistemology. Not all do that, but there are many people out there whose code, research and generaly understanding woudl be much better had they studied for a PhD.
Now, of course it's not necessary that these learnings have to come through a PhD, but few people pick up good research habits without one. And that includes masters students who at least in some fields definitely get Research-lite levels of understanding from the teaching given.
At least in the US, the last two decades have not been kind to the once iconic corporate research labs. As global competition increased and the push for constant high quarterly returns by investors, more and more has been off-shored. The most recent casualty was DuPont CR&D but there has been constant erosion. My colleagues and I have actually lost count of the number of downsizings we have been through.
Here is a test for you: Next conference you go to, look at the talks and posters. How many corporate people are there who are not instrument vendors or contract labs?