Recently, we opened the floor so you could ask us anything. Here is the last question and answer.
How would you restructure all of North American academia so that the best research gets done? Please consider issues including, but not limited to, the length of time it takes to get a tenure-track job (and associated “leakage” of talent along the way), the difficulty of getting grants, judging people primarily by their publications, failure to train graduate students in people management, public communication, and teaching, and the split mission of universities to both educate lots of undergraduates and do research. (Margaret Kosmala)
Brian: In four words – get rid of indirects. The fact that grants in the US give money to the university as well as the research has been hugely distortionary – it diverts the discussion from the mission of the university (blended teaching, development of human potential and research) to grants measured by dollar size. As to what you would replace this system with, I have no doubt in my mind that the place that has this best figured out is Canada. Key features in Canada:
1) Universities are funded directly by the government for fulfilling their mission – not indirectly via research dollars (and not via ridiculously quantitative ranking processes as in the UK and some other places)
2) Faculty are paid 12 months salaries (which means society owns and gets the benefit of a full 12 months of these high trained workers – can you imagine a business investing years of training in somebody and then only using their training part of the year?
3) Students are funded much more by government fellowships (grad and postdoc) rather than on RAships awarded to a PI. This tips the power balance more toward students which is a good thing
4) Core research funding is awarded to people evaluated on past productivity, not on projects, and is given out in a much more egalitarian system (everybody gets a little rather than the highly concentrated fashion in the US) and with much lower transaction costs (5 pages once every five years!). This accords with all rational thinking and recent empirical results. There is a saturating response to research money (throwing a 2nd million at a researcher buys a lot less than throwing that million divided over 10 researchers who have no money). The best predictor of future performance is past performance. Review panels have little ability to assess what are the most promising research projects, etc.
If you want to put this in a larger context, part of the reason indirect dollars has become so big in the US, is because there has been a gradual decay of the social contract of society valuing (and paying for) research and a trained workforce. This has been converted into an individual good. The fraction of the university budget that comes from the states (and non-research federal dollars) has dropped precipitously over the last 20 years. Part of this is a change of attitudes towards government in the US. But I think universities have to hold a mirror up and look at ourselves too. We need a new social contract (sensu Jane Lubchenco’s 1998 piece in Science but also sensu the 1862 Morrill act that established land-grant universities). Universities have to pay more attention to providing what society needs and society needs to step up and fund this.
Jeremy: Sorry, too big a question for me to do justice to. Because I think the ultimate issues are ones Brian gets at in his last paragraph, plus macroeconomic issues. Those issues go way beyond academia.
That doesn’t mean that there aren’t useful reforms that could be pursued. For instance, NSF and NIH could move some way towards the Canadian NSERC model simply by reducing the average size of awards and constraining how many awards a single PI can hold at once (there’d be no legal obstacles to such moves as far as I know, though I’m no expert). I agree with Brian that the Canadian system makes a lot of sense (see here, here, here, and here, for instance). Though I’d also note that the same trends as in other countries are happening in Canada too. And that some countries in which academia and research funding are organized very differently than in Canada (e.g., the UK) also do very well in many measures of the effectiveness and efficiency of their research.
p.s. Two minor corrections/clarifications to what Brian said: in Canada, most grad students are funded as TAs during the academic year, and then supported by their supervisors’ research grants in the summers. Government fellowships to students are smaller and harder to get than they used to be. And while it’s true that very few postdocs are paid by research grants (very few people have grants big enough for that; that’s a necessary consequence of the NSERC funding system), few postdocs are available from any other source, either. As with studentships, there are substantially fewer federally-funded postdocs than there used to be.