Is there any research/study/survey that looked at the main reasons why academics leave academia?
I did read a few articles explaining why some particular academics left academia, but I would like to have some statistics to see what are the most common reasons invoked.
I mostly interested in the computer science field (machine learning) in the US, but curious about other fields and locations as well.
I retired immediately after completing my Ph.D., but would have switched back to industry if I had continued working.
I am much, much happier and more effective doing technical work than teaching or managing. The computer industry has well-established technical tracks that allow career advancement without becoming a manager. The academic world, at least in the USA, seems to require teaching and administrative work from everyone, regardless of individual preferences and talents.
No hard stats on how often each reason occurs, but from anecdotal evidence:
Some people simply choose other carreers - either they're disillusioned with their research topic, or with some specific people/managers, or found a much better paying job in other domain. This is pretty much the standard set of reasons for any other jobs.
The only academic-specific reason that I have seen - it's often hard (or subjective) to say if it's "not enough money" or "you and your research are not good enough to compete for the money", but it certainly happens - some research project ends, a new one doesn't get started (yet), then people get other jobs to feed their families, and don't come back afterwards if/when new funding arrives.
For machine learning specifically, I think a major reason is the high demand for such skills in industry (both in existing businesses, in startups/spinoffs or as consultants), which makes leaving easier compared to some other fields. This applies to other fields that are close to the market as well (for instance engineering).
Innovation and tech transfer is important for universities, so I am not entirely sure how to classify researchers that 'leave academia' to start a spin-off to valorise the IP generated during their research career. That said, many machine learning scientists take this 'exit', which can be unavailable to researchers in more fundamental fields.
