I love #WhatShouldWeCallGradSchool. Hilarious.
I think the decrease in the percentage of PhD students funding themselves from their own resources has been partially due to a focus on programs in simply cutting spaces in their programs. They no longer let in students who don’t have funding in a lot of cases, not allowing them to fund their own way from personal resources. Or perhaps with the advent of the Internet, students are more savvy about funding and have chosen not to attend programs at which they were not funded, whereas before they would’ve.
One interesting note is that the 2011 chart on post-graduation salaries does demonstrate that across fields, on average people who go into industry make more money than people who start in academia. Postdocs in general hover right around $40K; academic salaries seem to range from $55K (humanities) to $75K (engineering), and industry salary averages range from $55K (humanities) to $100K (physical sciences). The only field that does not seem to have an industry bump is humanities - humanities PhDs make about $55K on average after earning a PhD regardless of whether they stay in academia or leave.
I will say, though, that most of my colleagues that have moved onto academic positions have been offered salaries quite a bit higher than the average for the social sciences. They do, however, tend to move onto R1 or elite institutions.
Another thing the chart shows is that the number of people going into postdoctoral positions has actually been increasing. They’ve always been more or less ubiquitous in the life and physical sciences - with about 60% of life scientists and 50% of physical scientists going to postdocs after the PhD in 1991. But in 2011, 20 years later, 70% of life scientists and just under 60% of physical scientists did postdocs after the PhD. The rate has creeped up in other fields, too. In 1991, less than 20% of social scientists and engineers did postdocs after graduate school; in 2011, 40% of social scientists and engineers did postdocs after the PhD.