So 50 jobs a year appear in astronomy? Not too sure about that one. Seems a little high to me. I wonder how many of those were real job, and not just 2 year soft money positions somewhere. We have a real problem in science with this. If we assume that a retired professor’s position will be replaced, then each professor should only ever have 1 PhD student. An average professor will put through 30-40 PhD students in a career, so that means 29-39 students must get a job in a research institute. Sounds unlikely.
So what is the solution? More research money in industry might work, and leads to the bonus of making companies more efficient. But this poses a second question- why should the tax layer fund PhDs if the students are just going to go into industry. More money in education, and a stricter selection of grad student might work. But this leads to the accusation of throwing good money after bad.
In general education is always good. So I recommend that students pursue a PhD because they love science, not because they think it’ll be good for their career or will lead to a job. If money is your chief concern, better to stay clear of research. Or aim to become president instead.