As mentioned in my previous post I spent the last three days at the AAAS-CASE workshop in Washington, DC. The purpose of the workshop was to introduce junior scientists to the legislative process and the ins and outs of science funding at the federal level. After two days of intense introduction, including a fascinating lecture from congressional expert Judy Schneider, we spent a day on Capital Hill meeting with staff members of our congressional delegations. I made the rounds of the Washington delegation with Kelly Fleming, a graduate student in chemical engineering at the UW.
The Washington congressional delegation is strongly pro R&D. The University of Washington brings a lot of federal money to the state, as does the Pacific Northwest National Laboratory and NOAA’s Pacific Marine Environmental Laboratory. Additionally Boeing, Microsoft, and other large commercial interests in the state benefit heavily from federal R&D investment and a highly educated labor pool, and presumably lobby on behalf of sustained (or improved) funding.
Despite this pro-science stance I was initially surprised at how little congressional staffers understand about how funding works at the institutional level. In retrospect this isn’t surprising at all. After six years in graduate school I’m still not 100 % clear on how science funding works at the UW (see previous post on this issue here). It quickly became clear that what we could best provide is some perspective on how increased grant funding directly translates to more jobs at institutions that receive federal funds and an improved quality of life.
Perhaps contrary to popular opinion academic salaries are not high. For public universities salaries are a matter of public record, the salaries of UW professors can be found here. It’s a bit difficult to pull a mean out, because certain appointments receive no salary, and there are duplicate entries for faculty members who were promoted in a given year. I would guesstimate the mean at $114,000 per year – and this is dragged upwards a bit by some outlier salaries in well-paying departments (engineering, medicine, etc.). On the surface the mean seems like a healthy salary. Although it’s much lower than other fields with a comparable amount of graduate education (e.g. law, medicine), and much lower than what industry pays PhD researchers, this is offset by certain advantages (namely academic freedom and prestige). If we dig a little deeper however, we see that the reality is quite different.
Most faculty appointments are only funded at 50 % of the stated salary (taking the mean, a paltry $57,000 per year). That’s not a bad salary when you’re 30 – a bit worse if you’re trying to raise a family – but for a full professor at the end of a productive career its an undervaluation. Why the limit to 50 % funding? This guaranteed pay is for teaching classes and performing other service directly to the university. The remainder of a faculty member’s salary is supposed to be raised from grants. Salaries aren’t a great use of grant funding, but the system might be considered rational if grant funding was adequate. To determine if that’s the case lets pull a typical grant apart.
In the course of recently preparing a grant proposal of my own I determined the mean size of single investigator awards for the NSF Antarctic Organisms and Ecosystems program, within the Division of Polar Programs, to be ~$400,000 over three years. Here’s how a typical award to that program might break down, costs are very approximate but loosely based on UW rates:
Institutional Overhead at 56 % (i.e. indirect cost recovery, varies): $224,000
Equipment/analytical services (i.e. doing the actual science): $50,000
Consumables (the day to day lab stuff): $20,000
Graduate student operating costs (i.e. tuition for 3 years): $45,000
Graduate student stipend and benefits: $90,000
Partial salary for laboratory technician (someone’s got to teach the grad student): $30,000
Travel costs (assuming this isn’t fieldwork based): $10,000
Remaining for faculty salary: $-119,000
And therein lies the problem. Faculty work around the financial limits as best they can by not hiring laboratory technicians, cutting back on analysis, and in some cases just dealing with less than 100 % pay (for far more than 40 hours of work a week). Education, academic service, and research all suffer as a result. So how to fix this problem? The largest chunk of money that comes out of a grant is overhead, but how would the UW and other institutions function without the income derived from indirect cost recovery? Vastly more funding to NSF, NIH, NASA, and the other granting agencies would help (and is a laudable goal anyway), but this isn’t likely to happen anytime soon.
I think the solution might lie in pursuing the more modest goal of retooling the way funds are allocated to investigators, while slowly pushing for more overall science funding. Taking our previous example, by increasing the size of the award to $519,000 we at least break even (keeping in mind that the hypothetical project was pretty cheap, requiring no expensive field work, equipment, or analysis). Adding three months of support for our beleaguered faculty member each year brings the total to approximately $609,000. If we assume each faculty member has two active grants at a time (lucky them!) the math works okay, however two issues remain.
1) If every grant was 50 % larger, we need 50 % more funds to support the same number.
2) Generally optimists say that 10 % of submitted grants get funded. The number can be much lower for some divisions.
I don’t think that pushing to get 50 % more funds to the grant providing agencies is unrealistic (it’s a minuscule amount of the federal budget), but the political will to do this isn’t even on the horizon. In the meantime making sure that grants are distributed equitably will help, though this goes against the grain of the brutally Darwinian academic culture where what matters is that an idea and implementation be best. I think there is a balance to be struck here in terms of funding the best and funding what’s really good but also supports a promising researcher and their research team. Encouraging adequate institutional support (which for the UW really means adequate state support) is also important. Reduced overhead and alternate funding sources for graduate students and technical staff would really help stretch the federal grant dollars.
Number two is really the bigger issue. I’m not familiar with NIH grant applications, but NSF grants require, in addition to a great idea, weeks of preparation. I’m a reasonably fast writer and reader, but based on my own experiences I’d estimate that it takes about 120 hours of work to put together a good proposal and all the necessary supporting documents. That’s assuming that you’ve already got a reasonably well-baked idea in your head and some preliminary data to show that the idea is well founded.
In our scenario a faculty member needs to fund two of these things every three years, which means that they have to submit about 20 (of course there will be some resubmissions in this total). That’s 2,040 hours of work, or 680 hours each year – 17 work weeks at 40 hours per week, if we lived in a sane world where that was the norm. Given the requirements for teaching, advising, service, and oh yeah, doing the actual research, I don’t think this is realistic. One solution is to move grants to a five year cycle so that a faculty member needs to fund two grants every five years. Yes, that makes each grant a bit larger, but it also reduces the number of active awards and has the additional advantage of bringing the grant cycle in (approximate) alignment with the length of time it takes a graduate student to get a PhD.
I had a great conversation with my colleagues at the Blue Marble Space Institute of Science about this issue this morning during the monthly “Beer with BMSIS” event (to be fair it was the afternoon in most of the country, and we only discuss a beer of choice, we don’t typically drink it). These events are recorded, so you can listen here (and of course tune in the first Thursday of each month, you don’t need to be a member of BMSIS to join us – but it does help if you like beer).