In A Zero-Sum Funding World, Proposals To Aid Some Should Also Describe Who Gets Hurt

Otherwise, you’re a jackass. Recently, eLife published a report about a workshop held at the University of Wisconsin, “Strategies from UW-Madison for rescuing biomedical research in the US“. The title tells most of what you need to know: it is yet another report in which the authors argue the future of science should revolve around how they do science–except their scientific path should receive more of the funding. Not only are there incorrect factual statements, but there are a whole host of implicit assumptions that are, well, awful. In short, most of the recommendations are, to be charitable, poor. I haven’t read something about the funding infrastructure of science that has made me so angry in a long time. So I’ll go through the article and critique, and then wind up with some thoughts about this.

Let’s fisk this motherfucker (where added, boldface mine).

From page 4:

First, the NIH should limit its support (from all NIH sources) for investigator salaries to a maximum of 50%, an idea endorsed by others (Alberts, 2010). This recommendation would counter the growing trend of requiring that faculty raise a significant percentage of their salary on their grants: it would also ensure that faculty expansion is accompanied by a long-term institutional commitment.

And the transition plan for mid-career soft money researchers is? Moreover, most universities haven’t budgeted for this, so the money will have to come from somewhere. Essentially, NIH should confine its research allocation to academic faculty paid by universities, as opposed to finding the best place to do research.


Second, in addition to offering grants to fund projects, more NIH institutes should offer grants to fund investigators, such as the new MIRA award being introduced at the NIH National Institute of General Medical Sciences. When investigators select this funding route, they will spend less time writing grant proposals and more time engaging in research and mentoring trainees. Such awards will also introduce flexibility into how research funds are used, thus maximizing research output per dollar spent.

This works fine, if you’re on the academic fast track. Slow starters, PIs with childcare responsibilities, and probably minorities will get screwed.

Still on p. 4…

Third, NIH institutes should limit the total direct funds from all NIH sources awarded to a given lab. For example, a million dollar limit could be applied to investigators in basic biomedical research.

Fine, but this will result in very limited savings–there just aren’t that many PIs getting this much funding. The entire fucking science internet has already been over this.

This toasted my noodle:

Fourth, the NIH should increase the proportion of its budget directed to Research Project Grants, Center Grants and Training, and it should decrease the proportion directed to R&D contracts, Requests for Applications (RFAs) and intramural research. These changes would redirect funds towards investigator-initiated research and allow funding of a greater diversity of projects. R&D contracts and RFAs place limits on the topics and approaches that can be pursued, so a shift away from them will lead to fewer intellectual constraints being placed on researchers. We emphasize that this is not a recommendation to eliminate R&D contracts or RFAs, but rather to reduce their number, which will sharpen their quality and provide the funds needed to award more investigator-initiated grants.

As someone who worked in public policy related to basic research for a couple of years, I would argue (and have) that some more “intellectual constraints” might be needed. There is nothing more frustrating than some version of the phrase “this is a preliminary study.” We often don’t another preliminary or exploratory study; instead, as I’ve eloquently put it, sometimes we need to go Manhattan Project on something’s ass. Readers might disagree, but I don’t think cutting RFAs and R&D contracts is an unalloyed good. Oddly enough, an academic-oriented group that lives on R01s calls for more investigator-driven R01s.

Ok, last gripe about page 4:

Last, the NIH should continue to set higher pay-lines for ‘early stage’ investigators and to prioritize ‘first-time renewals’ in order to promote the research of these junior investigators.

So they can lose their funding support when they turn 45 or 50? Ok, onto to p. 5, which deals with training:

Cutting the number of students would limit the number of trained scientists entering the workforce, which was not favoredAnd there was little support among faculty for the proposal to shift funding for graduate students from research project grants to training grants. There were three main reasons for this: training grants only cover a portion of training costs; they fund few non-US citizens; and the relatively narrow focus of training grants would limit intellectual diversity. The prevailing view was that dramatic cuts to graduate student numbers and a shift to training grants would do more harm than good.

Instead of cutting the number of graduate students, we recommend a narrowing of the workforce pipeline at a later stage. Our first recommendation is that fewer PhD students continue as academic postdocs. Most workshop participants preferred this option and supported the broadening of PhD programs to include experiences relevant to non-academic careers.

We don’t just have an overproduction of ‘academic’ biology PhDs, but of biology PhDs in general; the biomedical sector is plateauing. Surprisingly, a workshop consisting entirely of academics has no understanding of the larger biomedical job market. There isn’t exactly a shortage of PhDs. It’s also shocking that faculty members wouldn’t want training grants: imagine if a cheap source of labor was available only if the program had to compete with other programs on the basis of how well they train students? Can’t have that. By the way, a training grant system is the only way for prospective students to have leverage on the quality of training, which includes “the broadening of PhD programs to include experiences relevant to non-academic careers”, by tying the availability of student support with training. Also, can we stop with the narrow focus argument? I was on a training grant (admittedly many, many moons ago), and there were few, if any, limits on “intellectual diversity.” What does limit intellectual diversity is the view that students are replaceable labor.

While the call for support for staff scientist positions is good, the funding mechanism seems to rely on Magic Ponies. When one reads “the NIH should provide funds to cover the extra cost of replacing a trainee position with a staff scientist; the mechanism for doing this could be similar to the supplements used to promote the diversity of the research workforce”, one wonders where the money comes from–and if I were in one of those ‘promoted’ groups, I might ponder much about that. Just saying.

Then there’s this on page 6:

More radically, we suggest that the NIH raise the pay-line to 20%, regardless of the funds available, with a sliding scale of funding that depends on the score received by the application and on the other funds available to the investigator. For example, those with higher scores might receive close to their full request, while those with lower scores might receive only half the amount requested (and be expected to achieve half of the aims listed in the application).

This might work for PIs as more will be able to claim a grant, but, in terms of getting science done in some sort of cohesive way, this is a disaster. We’re underfunding research programs as it is, which has serious effects on the quality and integrity of that work, so let’s slash budgets further? As inconceivable as this might seem, people outside academia expect NIH research to accomplish stuff, not serve primarily as a jobs program for wayward PIs.

Thankfully, we can move to page 7:

US spending on research as a percentage of its GDP has dropped precipitously over the past two decades, from second to tenth among developed nations.

Actually, as a percentage of GDP, if we compare 1994 to 2014, NIH funding has increased by a third. It has dropped from the end of the doubling era, but that statement simply isn’t correct.

So let’s wrap this up. I realize readers might disagree with me on some or all points. But what made me angry about this article is that it reflects one very blinkered experience, one that is limited by experience, not to mention some implicit class, gender and racial biases. Moreover, most recommendations don’t indicate who would be hurt by these changes. In an essentially zero-sum environment, that’s crucial. I don’t have any magic solutions–there aren’t any*–but here’s my take on things:

The problem isn’t the overall funding, it’s that the NIH (and to a lesser extent, the NSF) have created an unsustainable system built around a cheap supply of graduate students and post-docs (especially since the idea that they work only forty hours per week is ludicrous). Compared to many other government contractors–and in the most reductionist sense, that’s what we extramurally-funded researchers are–we come really cheaply (and faculty at universities get most of their salaries paid by the university, so they’re cheap too). This is a great thing over the short-term for NIH. Who wouldn’t want a highly-motivated and -skilled labor force whose wages are lower-middle class?

So how is it unsustainable? Well, at some point, all of these cheap workers eat their spinach, grow up big and strong, and want to sit at the grownups table. Unfortunately, the NIH model doesn’t allow for enough PIs (grownups). So funding for R01s has become really tight. A lack of leadership by the funding agencies is critical here: there wasn’t (and barely is today) any comprehension that the constant generation of new PhDs is unsustainable, that at some point, the perpetual motion machine stops moving. NIH needed to realize that the increase of funds, if not managed responsibly, would increase the PhD ‘birth rate’, while there was no equivalent increase in either positions or in the PhD ‘death rate’ (retires). This is probably mostly due to lack of foresight as opposed to cynical malevolence, though I could be persuaded otherwise.

Until we fix the glut, both through limiting PhDs and by moving more funding to extra- and intramural research institutes, or else convince elected officials to increase funding dramatically (which merely kicks the can down the road if there are no systemic changes), everything else is basically shuffling chairs on the Titanic.

More funding would help–at least remove the damn sequestration caps. But attempting to shore up a fundamentally unsustainable system won’t work. At least I’m admitting who gets screwed–and I realize that the ‘losers’ won’t be happy. The Wisconsin proposal doesn’t even recognize there will be losers.

That’s not helping.

*Hoping that U.S.-ians realize that, at the federal level, we can’t run out of money and are only constrained by real resources (which include people) and inflation, and thus decide to massively increase science funding seems as likely to occur in my lifetime as does messianic return. Which is to say, not very likely.

This entry was posted in Funding, NIH. Bookmark the permalink.

3 Responses to In A Zero-Sum Funding World, Proposals To Aid Some Should Also Describe Who Gets Hurt

  1. Pingback: RFAs do not narrow the NIH portfolio, quite the opposite. | Drugmonkey

  2. Morgan Price says:

    My cynical view is that as long as PhD programs are the easiest way for foreigners interested in science to immigrate to the US, the oversupply of PhDs will continue. In other words, the current system *is* sustainable, even if it seems wasteful.

  3. jmz4gtu says:

    Mike, you assert that most faculty have a substantial portion of their salary covered by hard money (in the section where you talk about how cheap the labor force is). Where’s your source for this? I was under the impression this was not true, but I don’t have any numbers one way or the other.

    Secondly, I don’t get the all or nothing approach to tightening the pipeline. There are valid reasons to do both, so why not do both, but allow some wiggle-room.
    First, I would shift some of the training grant money to post-bach money for independent fellowships. So you graduate, tech for a year or two, and write a fellowship that, if received, would essentially allow you to go to any school with an accredited grad program. You could stipiulate minimum criteria regarding academic accomplishment (GPA, GRE scores). I’d make the program for the first two years (i.e. guarantee a masters), with a relatively easy renewal (found a lab, have a project, all clear from the PI and dept. chair) for another 3 years. Let the training grant holders know this is coming, and that the bottom 20% are getting culled. This approach would increase the quality and maturity of grad students, since they’d have technical skills, writing skills, and some idea of a research focus going into their program. This is contrast to the current criteria for training grant appointment, which, at my old uni at least was, “American with good grades”. Then, and this is key, you actually *keep track* of the trainees on each kind of support, and adjust funding levels accordingly over the ensuing decade.

    Concurrently, I would phase in a prohibition on allowing tuition to be charged to RPGs. This will reduce the PhD candidate pool substantially (about 15-20%, judging by the supposed costs of tuition, which may be inflated). Eventually, you’ll want to phase in some sort of university contribution to stipend/salary. I know this is cost shifting and universities swear they don’t have the money to cover it, but worst case, they just charge the tuition directly, which will still serve the purpose of limiting the grad school population.

    For postdocs, the solution is simple. You simply allow only a fixed number years of salary support off of an RPG (I’d suggest two, taken at any point), while expanding the fellowship pool to meet a target number of postdocs (you might also want to expedite fellowship review). You can be rougher on the PD population because they’ve already gotten their degree, and had ample time to consider their career options and where they best fit in the scientific world.

    Couple this with mandatory salary contributions from the universities (again, phased in gradually) and you’ll have a system that doesn’t balloon so much in times of feast (because the trainee workforce will not be proportional to the number of R01s), but is relatively robust (due to multiple lines of salary support) during times of famine.

    The trick is to do all of this slowly, gradually granting the NIH, and not the universities, control over the size of the biomedical workforce. As a further point, I’d suggest that any of the likely forthcoming increases to the NIH budget be used to supplement the size, and not number, of RPGs, until some reforms to the training pipeline are made.

Comments are closed.