Do students often go to grad school because they had a bad industry job?

Purely anecdotal: I've seen many people go to grad school because it is a path of least resistance, or because they can't find a job.

There may be another group of careerists that have it as part of the plan because it's a recommended qual for a job they want - though in CS I don't know how realistic this is. Really strong people that want a job are more likely to go get one with an undergrad.

A small minority are genuinely interested in research, and do grad school for that reason.

I worked a number of industry computing jobs starting at 14 years old. I worked at Apple, Palm, the Internet Systems Consortium, as well as a gov't lab (Lawrence Livermore.) Lawrence livermore was cool but high performance computing didn't really grab me. ISC was very cool and dealt with super interesting problems but I don't think the software was exactly what I wanted to work on. Apple and Palm were neat but what I was working on felt like it lacked importance. Consumer electronics did not feel like it was endowing my life with meaning. So, I went for a physics PhD and now a postdoc and am pretty happy. I get to be pretty self directed and get to be the first to see a natural phenomenon and figure out how to probe it. This gives me a fulfillment that I couldn't find in industry. Maybe I'll go back to industry as a research scientist -- who knows.

I went to grad school because the company paid for all but a couple thousand dollars of it. I thought it would help me move up and maybe even into management.

It was a waste of time and money. It would have been more productive and lucrative for me to spend a couple extra hours every night doing work projects rather than going to night school.

I don't think this is particularly common at least among people I know. When people in the industry aren't happy with their job, they find a new job. If they aren't happy with the whole industry, they usually change to a different industry.