I don't know if he was doing research or not. (And I'm not trying to judge him either way, since I know almost nothing about him.) But when he was doing his N

I don't know if he was doing research or not. (And I'm not trying to judge him either way, since I know almost nothing about him.) But when he was doing his Nobel-winning work, he was publishing on average about one paper per year. Then he published one paper in 13 years, and then zero papers in 17 years (and never again). He could have been doing research but unsuccessfully. He didn't lose his job. (One might argue that he should have, but I won't.) He just said that he would have a hard time finding a different job. Isn't that obvious? If you are looking for a new job at a research university with zero publications for 17 years, you can hardly expect departments to be dying to hire you. I think most people in that situation are not doing any research. Most researchers with ambitious research projects still manage to publish occasionally, not just for "the system" but for themselves. You want to know that you are making at least a little progress on something. Andrew Wiles, for example, published every year or two throughout his work, with the exception of one four-year gap. [1] No 17-year gap. [1] http://web.math.princeton.edu/WebCV/WilesBIB.pdf http://web.math.princeton.edu/WebCV/WilesBIB.pdf