I agree with MitchConnie that you've got to subtract out those who have such long commitments that it would be financially stupid to bail afterward. (Think ROTC+USUHS, second residency folks, long GMO tours then long residency etc.)
The 95% I throw around is from my service, and my specialty and comes from the specialty leader and the residency program directors. But so what if it's off 5%? How does that change anything?
And looking around my hospital (~55 docs), I can only think of 7 who are beyond their initial educational commitment:
1) Group commander O-6, IM subspecialist
2) Squadron commander O-6, Ortho
3) Squadron commander O-6, Flight doc
4) Came in from civilian life to do flight med, by his own admission wasn't cutting it in civilian life, O-5
5) Came in from civilian life to do ophthalmology, most of us suspect because he wasn't cutting it as he certainly isn't patriotic, O-4
6) Dermatologist (did second residency after finishing initial tour as an FP), still under second educational commitment, O-5
7) FP who is non-deployable and truly loves military medicine, O-4
8) Anesthesiologist who per the recent residency grads in their first commitment is "completely incompetent", O-5 or O-6.
Being in the ED, I think I know just about all of the docs.
That's it. 8/55=15%. It seems a bit higher in the Navy, but I'm not sure how much of that is the fact that these are folks who are at a major medcenter (more likely to be higher ranks), the fact that commitments are longer due to GMO tours so financially a better decision, or perhaps there is something different about Navy medicine. Whatever the number is, it's pretty low. I'd bet money that at least 75% of those eligible to get out in any given year in any given service in any given specialty do.
But let's examine Navy FPs numbers a little closer:
He says 10% get out in any given year. What does that mean? How many would we expect to leave in any given year if EVERYONE who could leave, did? What we need to know is what percentage CAN leave every year? Let's throw some educated guesses around and see what we come up with. Let's assume a typical doc stays "in the system" for internship, a 3 year GMO, a 3 year residency, and a further 3 year payback. That's ten years. So every year, 10% are eligible to get out. And apparently, about 10% get out per year. Seems awfully close to everyone to me. Even if you assume some people have shorter commitments, say a straight through 3 year residency and a 4 year payback for 7 total, and lower the average to 9 years, you only end up with 11% each year who CAN get out, and apparently 10% do. That doesn't look much better.
The numbers are rotten no matter how you paint them. Would you like to be in a job where even half of the people leave when given the choice? I wouldn't. I want to be in a job where not only do those who have it hold onto it with gritted teeth, but there are dozens of people banging on the door outside to take their spot. That's the way most of medicine is, but not military medicine. If the job was so good, why do so many people leave at 20 years? Why not stay 25 or 30 if it is so great?