One that has been widely discussed is the difficulty in selecting the best future researchers based on Physics GRE scores and transcripts from the huge variety of undergraduate programs.

A second problem is that decreasing the number of PhD students will directly lead to fewer faculty jobs in the future. To see why this is the case, imagine asking a Dean to replace retiring faculty in a department that just either eliminated its PhD program or cut its numbers by half. Since teaching graduate courses and supervising graduate students is part of what faculty at research universities do, there is less need for faculty, so why hire in the field without grad students. Moreover, a field with fewer students will be less vibrant because of the poorer selection as noted above.

The underlying issue is graduate school is about education. If this isn’t working in that few to no graduate students are prepared for anything other than replicating their thesis advisors, then one needs to look at changes in graduate education. I think the evidence favors the view that physics and astronomy PhDs are more broadly useful, but even if this were not true at some point, decreasing the number of graduate students does not really solve the problem.

“A PhD is Not Enough – A Guide to Survival in Science” P Feibelman (1993)

“Alternative Careers in Science – Leaving the Ivory Tower” (AP, 2nd ed, 2006)

(Not sure, but I may have mentioned the second in your previous article on this. Apologies for the repetition if so.)