>I can't help but suspect that sometimes it may be related to graduate school itself, which can be stressful and unforgiving, with minimal support, and where supervisors often hold both academic power over their students' futures and financial power over their livelihoods.
I dropped out of a PhD -- took the master's I earned for coursework, did my quals so it would be clear I chose to leave, then took an "academic-ish" job that paid very poorly. I'd hoped to do that a bit then get hired by a big tech company, but I found out that you have less free time in a job than grad school, and my tech skills began to erode, further sending me down a path I did not want.
What caused me immense, IMMENSE distress is that I felt, for lack of a better term "involuntarily destitute" -- my adviser in grad school had told me that she'd ONLY give me a positive reference for "research" jobs, and that trying to leave for industry was evidence I had lied my way into the program, and thus she could not give me a positive reference for any roles without a research component.
I feel that she purposefully tried to "trap" me with her -- she was having trouble recruiting new students as word of her behaviors and convictions spread (she'd racked up a DUI during the liminal period between my acceptance and starting school, among other gems).
I currently work in a job that has nothing to do with my field -- I had many, many years of strife because when I was fresh out of college, I looked around my hometown and found I couldn't even get a helpdesk job because my skillset was that of an open source nerd, and they wanted people who could answer questions about the UI of Windows like "How do I enable this printer" that, having not used it for years, I couldn't answer off the top of my head -- and it's not sustainable to "just Google it" on calls over and over, people will get frustrated with the wait times.
(That was the way people generally broke into infosec back then -- get a help desk job at a bank, hospital or university, study during downtime, maybe do some certs or try to do an interesting project to present at a conference, move up to sysadmin, and eventually security analyst/engineer)
I thought I'd found a third way -- I could do this PhD, and at worst leave with a master's, and sidestep the tedium of the help desk and the uncertainty of if I'd move up. (I knew people who got worked many hours, struggled to study up, and got trapped).
Anyways, academia can be incredibly abusive and downright medieval. That's not an excuse for violence, but it is an explanation.