What is the minimum time one should spend in a post-doc?

Question

Suppose one has been hired for a three-year postdoc. Three years give sufficient opportunity to get some work done, without moving to a new country every year. Possibly, an opportunity might arise to move to something permanent before the three years are finished (either in academia, or in industry, or in government research labs, or somewhere on the edge of those). However, the postdoc supervisor, who has invested time and money in a postdoc in order to get a project done, might be disappointed if the postdoc leaves a job unfinished.

On one extreme, I know someone who has been a postdoc at the current place for 5½ years, and they were a postdoc somewhere else before that. I suppose spending 8 years in postdocs does not help when applying for permanent positions. On the other hand, I know someone who has been a postdoc for only seven months or so, and is already applying for a primarily teaching position at a somewhat lesser known university, possibly a position where major post-PhD results are not essential.

What is the minimum one should spend in a postdoc, before even starting to consider applying for possibly permanent positions? I mean this question broadly; an answer may either be expressed in time (one year) or in results (one major and one minor first-author paper). Of course, without any significant results, finding a faculty position at a major place will be hard if not impossible. But the question here is related to the moral obligation to complete some of the task one has been hired to do.

Answer 1

Postdocs are cheap labor. Furthermore, unlike grad students, they show up already fully trained and ready to do productive work. This is all a wonderful deal for the lab that the postdoc is working for.

If a postdoc is applies for a job after 7 months, they will probably still have been in the postdoc position for about a year before they actually leave. This is perfectly reasonable. Tenure-track jobs are so scarce and hard to get that any reasonable supervisor's reaction in this situation should be, "Wow, that's wonderful that you lined up a tenure-track job so quickly!"

In any case, this will presumably come up when the postdoc asks the supervisor for a recommendation for the job. If there needs to be a discussion of how long is long enough, the discussion will happen then.

Three years give sufficient opportunity to get some work done, without moving to a new country every year.

My perception is that 2 years in a single postdoc used to be standard. It may now be common to do 2 or more postdocs, each of 3 years or more, but that's an expression of market conditions, not of any moral obligation that the postdoc owes to the supervisor.