[data_acquired]

>Although NIH wages aren’t inherently tied to those training grants, if wages increased significantly for NIH fellows, universities might “raise holy hell”, Wiest says, because it could incentivize researchers to apply for NIH positions instead of those at universities.

This is a pretty weak point. There's hundreds of research areas and problems not worked on at the NIH that are explored in the hundreds of universities in the US. I don't see anyone switching research areas for better pay --- your skill sets are often tied to a certain class of problems. Sure, yes, there may be such effects in certain research areas. But what fraction of university budgets even go towards post-doctoral fellow stipends (when weighed against every other expenditure in a university budget)?

[kkylin]

Not disagreeing with your larger point, but a nitpick: at these levels (postdocs, research scientists) the relevant university positions will be funded by grants, in fact largely from the NIH. University budgeting doesn't have all that much to do with it.

Now, postdoc salaries are set by a combination of university policy + whatever the grant budget says, but grant proposals are also assessed based on how much they ask, and postdoc salaries will be set by whatever NIH actually awards the PI. In the end, for NIH to pay their early career scientists more without creating problems for itself (by competing with groups it funds), it will likely have to increase postdoc pay across the board. This will likely add some upward pressure to postdoc pay across fields, since many universities do look at entry-level NIH postdoc pay as a reference.

[data_acquired]

Fair point about how the funding actually works. The UC strike did allow for a good increase to the post-doctoral stipend, and the institutes likely had to move money around in order to make the increases possible.

What's also interesting to me in the concerns raised in the article about how budgets (either from grants or universities) will cover pay increases is that this concern never comes up when research consumable costs increase. I had a friend at a major embedded systems supplier for researchers who spoke of the 90% mark-up they charged labs. This gets into a messy issue of course about pricing power and monopoly in scientific supplies, but its telling that as much of a concern is not raised in the public domain about these sorts of cost increases?

[LatteLazy]

If private companies did this, it would be a huge lawsuit: we mustn't raise wages, we have a tacit agreement with the other big employers to keep them low!