Well, if by "demand" you mean tenure-track faculty jobs, there's a plot in the article of that. The answer, which as far as I know is applicable across all fields, is that the number of PhD graduates have risen dramatically while the amount of tenure-track jobs are flatlined or even dropping.

Unless job demand is driven by exponential growth (like physics during the Cold War), it's simple math to conclude that every professor can graduate one Ph.D. student during his entire career for the job situation to be sustainable. Most probably graduate an order of magnitude more than that. (This of course ignores non-academic jobs and the question whether getting a Ph.D. is a good investment for such jobs.)