Call it postdoc creep. In four years, the number of newly minted US geophysical PhDs who went on to become postdocs rose by nearly 50%, according to a survey conducted annually by the American Geophysical Union and the American Geological Institute. The survey, released this month, says that for the class of 1999, just under 40% opted to do a postdoc. Four years later, 58% of the class of 2003 decided to take a fellowship rather than a permanent position.

Meanwhile, as the number of geophysics postdocs increases, the gap between postdoc salaries and those for other positions remains wide. The median salary for a university postdoc — the lowest remunerated category — is about $39,000 whereas for a newly minted PhD working in industry it is $70,000.

The news for geophysicists isn't completely bad, though. The majority of 2003's graduates found work in their field, despite an economic downturn, with most reporting positive job satisfaction. And the percentage taking postdocs is still lower than it is for new PhDs in the life sciences, where a postdoc following a PhD has become the default.

But the growth in the number of geophysicists doing postdocs is a concern. More than scientists in other subdisciplines, geophysicists tend to work between receiving their undergraduate degree and starting their PhD, which means they tend to be more skilled and, theoretically, more likely to be hired without doing a postdoc.

So if fewer of them are finding permanent post-PhD positions, and there is still a disparity between postdoc and industrial salaries, there is less incentive for geophysicists with just an undergraduate education to go on for more training. The only way more permanent positions will open up — at least in government and academia — is with more funding from the US government for geophysics, an unlikely prospect, at best.