[jjoonathan]

Markets don't know how to judge the value of science. Its main product is a non-excludable good -- you can't know if a train of experimentation or thought will ultimately be important until long after you have shared the results, but sharing thought/results almost entirely removes your market leverage. Therefore, the market says the value of science is... nothing. Which we've collectively decided is pretty obviously not true, so we fund science through the government, but that still doesn't give us any way to judge the value we're getting in return. Are we funding it too little? Too much? There's no way to tell the bang we're getting for our buck, we're just guessing(footnote). That guess determines the demand for scientists. Market corrections in the academic labor market target that guess, not any underlying market truth, so it's hard to say whether they're good or not, no matter how much of a True Believer in markets you are.

Contrast to, say, video game programming, where the market forces pushing people away from those jobs do represent a fundamental truth that isn't obviously baloney: there are more people who want to program video games than society needs/wants to program video games.

* Given the set of things that market forces tell scientists they should be doing instead of science, I'm pretty sure we're dramatically underfunding them, but that's certainly debatable.