Is there any extensive research/study/survey that looked at what percentage of people who left academia were happy with their decision after X years?
I mostly interested in the computer science field (machine learning) in the US, but curious about other fields and locations as well.
People leaving academia can be PhD students or after (tenure-track, tenured, soft-money research positions, national lab researchers, etc.).
I believe some of the national longitudinal surveys would cover this for the US.
For instance, the National Longitudinal Survey of Youth 1979 cohort started tracking "middle school" age students in 1979 and is still tracking them. They have hundreds to thousands of attributes including education and many measures of health and happiness (though not all are updated every year).
Another example is the National Longitudinal Study of Adolescent to Adult Health (Add Health).
Well, I cannot think a reason that such a study should be either thought of doing or published... It lacks in many aspects (strictly speaking of study) and has really many dependencies (e.g. economical, political, cultural, scientific discipline, field in the discipline and others).
What I think that you actually seeking is some sort of justification for preferring one after the other. Since I can only speak of my self, I will try to present my point of view.
Professional and academic "worlds" are interconnected. Neither is better, "scientificier", more correct or anything than the other. Actually, (I believe that) it is supposed that academia is the more risky research department of a society, where by "risky" I mean research that mostly not results in a commercial ready product. But academia involves also teaching, where by teaching means that you have to pass to others what you know.
On the other hand, professional world is not only sales. It has development, implementations and other really interesting parts (at least to me). Of course, without sales there would not be any actual "sales"... but that's something that I cannot discuss because I think my self as a little biased on that subject. Professional world has also patents, applied knowledge, problem solving on implementation (and not in simulation) and actually making things that work all the time.. not only in the lab or at the presentation. I think that you can image what would happened if your car, your kitchen, your laptop and all the other facilities worked as most things in academia presentations, applications etc.. Armageddon!
Finally, a PhD is not only a way for an academic career. It is also a good way for creating a start-up company! Or obtaining patents! Especially of you get paid during your PhD.
So, what someone should choose must not be based on "how most people felt" but on which enviroment feels better and what goals this person have, i.e. wants to create things that actually work? things that are used as they are by many? Or teach and conduct research for unsolved problems or in order to optimize existing implementations?
And a final tip, if you are thinking on starting a PhD for pursuing an academic career then you choose it for the wrong reason (at least on my perspective). A PhD should be started for satisfying the inner need of research and not because taken as a career solving degree.
