“Almost entirely” is very different from “somewhat.” Whether or not it’s self-evident that most “science” PhDs don’t do science, there’s plenty of finite evidence of problems, like Saul Perlmutter’s claim that he couldn’t do the work that won him his Nobel today. Perhaps there’s not an uniform decline, but there’s enough evidence that the meaning of the relevant metric is not consistently reliable to make it pretty sketchy to use PhD as a proxy for doing meaningful scientific work.
“Almost entirely” is very different from “somewhat.” Whether or not it’s self-evident that most “science” PhDs don’t do science, there’s plenty of finite evidence of problems, like Saul Perlmutter’s claim that he couldn’t do the work that won him his Nobel today. Perhaps there’s not an uniform decline, but there’s enough evidence that the meaning of the relevant metric is not consistently reliable to make it pretty sketchy to use PhD as a proxy for doing meaningful scientific work.