Indeed, and in my field, it even had a name: The "Birth Control Problem." In fact, the problem extends to post-docs. Professors were even training multiple post-docs. I saw that quite clearly.

I was lucky to have role models: My parents were both industrial scientists, so the idea wasn't foreign to me. And I figured that I could always become a coder if things didn't work out. ;-)

Re-reading, I may have been unclear: I meant that some people take temporary post-docs with no intention of staying in academia.