Sunday, March 16, 2014

The End of Science

Eli in his long life has seen major changes in the funding models for universities and national laboratories in the US, and Europe.  And you know what bunnies, while different in detail, the end results are pretty much the same. 

The problem is the economic model which fuels health science and increasingly physical science in the US and elsewhere. By only providing a percentage of salaries, the schools and national labs force the faculty and staff to become entrepreneurial at the expense of scholarly.

In order to reach the level of support needed, faculty and staff build large groups of graduate students and post-docs (many of them international students) who then flood the market when they graduate, placing unsustainable stress on the research grant support mechanisms. Of course, some of this overpopulation finds homes at new or lower ranking places who try and populate the streams as lesser piranha  but more often get swallowed by the larger fish.  Starve the poor is the entire point of league ranking.  Rather kill the poor and starve the rich.

The large research groups need to be fed.  This puts the most successful PIs into a constant cycle of grant writing, program manager schmoozing and weekly conferences with little to no time remaining for research or teaching.  There are advantages to staying away from research institutions and burrowing in at a teaching college if you want to do your own research.

At the same time legislatures, seeing the flow of funding from national sources into research at universities, decided that they could cut back on direct support.

A perfect storm.

There is a choice, smaller department with smaller research groups, fewer new doctoral students and a return to a base funding model.

Michelle Goldberg in The Nation, castigates Columbia University's Mailman School of Public Health for firing Carole Vance and Kim Hopper for not bringing in enough money to support their salaries.  She sees this as a splendid club for beating on Nicholas Kristof, but the problem isn't Kristof, it isn't Columbia, it is the sea in which Columbia swims.
Kristof is right that universities have become inhospitable places for public intellectuals, but he misses the ultimate cause. The real problem isn’t culture. It’s money.
Like many schools of public health, Mailman operates on a “soft money” model, which means that professors are expected to fund much of their salaries through grants. (Many professors there, including Vance and Hopper, work without tenure.) Recently, the amount expected has increased—from somewhere between 40 and 70 percent of their salaries to as much as 80 percent—and professors say that it’s become a hard rule, with less room for the cross-subsidization of those who devote themselves to teaching or whose research isn’t attractive to outside funders. Meanwhile, the National Institutes of Health, the primary source of grant money, has seen its budget slashed. These days, only 17 percent of grant applications are successful—a record low.

The result is an increasing focus on the bottom line over a broad engagement with social issues. “One of the costs of this push for federal funding is going to be a depoliticization of the scholarship at Mailman,” says a professor there who asked to remain anonymous for fear of administration reprisals. “You can’t do great public health without engaging with politics.
And you certainly cannot support a science with scientists who have to spend all their time shaking the cup.



The inverse problem is institutions hiring public intellectuals in the long term interest of Federal grantsmanship, or when political fanbois feel like funding a chair or two in the next big thing in PC disciplines.

Anonymous said...

Disturbing story on privitization of science funding in Sunday NYT

Bill W

And Then There's Physics said...

All I can really say is that I largely agree. The current funding model is encouraging universities to see research (and students) as a way to generate income, rather than seeing the money as being what we need in order to properly educate students and to do our research. As it stands at the moment, universities want to expand (both students and research ativities) so as to generate more income. However, as you say at the end of your post, we could quite easily have smaller departments and smaller research groups doing (hopefully quality research and in which the funding level is set so as to support such a structure.

Victor Venema said...

Projects were intended as a bonus to people doing good work, it was never intended to be the main source of science finance. Which you can also see in the way the project funding works, there is a lot of noise in the system. If you want to be sure to keep your good people, you have to apply for projects for a much larger group, because you cannot be sure to get funded no matter how good the research proposals are.

They call it a competitive system, but actually the people making the decisions have no interest in the outcome. That is not a market. It is worse than socialist planning. At least they had a plan. It is a wonder that reviewers and commissions give such decisions some of their time and energy. The system does not reward them for good decisions.

In Switzerland they have a nice rule that makes the system a little better. One professor can only have a limited amount of funding from the national science foundation. In that way at least the professor will set priorities and only submit proposal for their most promising ideas.

Michael Tobis said...

There's also another very stupid aspect of the system: that performance is judged by publication record. Unfortunately, peer review is done by people stuck in the very same treadmill who don't have time or motivation to do a good job.

So review at best is lackadaisical and sometimes actually malicious - the editor will call on a reviewer who is from a group explicitly hostile to the submitting group.

In the end everything that looks like a paper gets published somewhere, and any real progress that happens is relegated to a social process that is indifferent to where a paper was published.

Telling the good stuff from the bad becomes an in-group of adepts process, which is better than nothing but defeats the democratizing purpose of publication. And measuring success is something of a "we don't want nobody nobody sent" process of an overlapping in-group of influencers.

Yet journal "impact" and paper count are the key metrics in the competition to which you refer.

Both grant writing and publication thus become matters of quantity rather than quality. The most valuable skills in science then become typing, typesetting, graph drawing, and tolerance of incompetent web interfaces. Ultimately the people doing the best work, who actually take pride and care in their publications, are systematically penalized.

The idea that what we have now is anything like a marketplace where the best ideas are rewarded is increasingly delusional. I agree with Victor on this. I add that in addition to the broken motivational system, the metric of success is also badly malfunctioning, partly as a result of the funding pressure driving out all other activity.

And as a consequence of this demeaning atmosphere, many or most of the conversations among scientists are about money. Given that a big slice of the remainder is still basically gossip, that doesn't actually leave much room for, you know, science.

Anonymous said...

Same at the other side of the Atlantic.


Sub-Boreal said...

Meanwhile, up here in the Great White North, the beloved Discovery Grant system of our NSF-equivalent NSERC is well into drain-circling mode.

Not that it was perfect by any means. A widely-circulated 2009 critique had the self-explanatory title of "Cost of the NSERC Science Grant Peer Review System Exceeds the Cost of Giving Every Qualified Researcher a Baseline Grant".

But more recently, a major bureaucratic restructuring in the NSERC review system system brought about a significant deterioration in success rates, striking folks in smaller institutions particularly hard. And new criteria now explicitly rewarded/punished grantholders for their output of graduate students, especially PhDs. So if you want to run a small lab - for a host of perfectly good reasons - you're automatically viewed as an underachieving loser.

Pity. We had something reasonably functional here.

Mitch said...

I have spent the last 35 years in the academic research environment, and find that most comments here are way off mark. NSF was set up in order that most research was done at universities, not at large government labs. This was done in order that there was a synergy between expanding our knowledge base and educating our children. We still need the knowledge--if we don't do it at universities, it will be done at NOAA or USGS, or other government agency.

Also, when I started back in the late 70's the meme had already changed from 'publish or perish' to 'get a grant or get lost' to make up for lost income from state subsidies. It is ironic that most people actually think that research is taking resources from the university, when it actually has been a major supporter of students, state-of-the-art facilities, and faculty with up-to-date knowledge.

Finally, the overproduction of students is primarily a result of university administrators who set mindless metrics on number of PhD's produced, not on quality of research.

Anonymous said...

I respectfully disagree somewhat with the comment from Mitch about USGS. They already have their hands quite full with the day to day work involved by seismic risk and other natural geological risks, they can do *some* research (and they produce outstanding science I shall add) but they can't replace easily for example geologists at Boulder.
Not without a massive increase of funding, that is. and we all know the probability of such an increase. Especially since USGS has grown an habit of embarrassing frackers by stating the obvious (never inject large amounts of water in a fault).