A few days ago, Chris Newfield published an article on his Remaking blog on the proposed NIH cuts to university indirect costs. Now, Chris is the most knowledgeable person on university financing who is willing to speak openly about how university accounting works–to the extent that university accounting admits of being designed to, um, work.
Set aside for a moment Chris’s overt politics–even if you disagree with him–and consider some of his main points, as I paraphrase them (please–go read his article, too):
Most discussions of federal indirect costs don’t get at the “hidden budget reality of academic science.” Indirect costs support research infrastructure–stuff that can’t be easily charged out to each grant, like buildings and kitting out labs, and dealing with maintenance, and all the federal compliance issues, not to mention all the federal grant applications that have to be sent in to have a chance of getting a federal award. Dropping indirect costs means threatening the infrastructure, means threatening the resources scientists need to do research. Justifying an indirect cost rate is difficult–and expensive–but that’s a reality of modern science research.
The NIH reasons that if the Gates Foundation and others have set an indirect cost rate of 10% or 15%, that must be the market rate–which is nonsense. It is more the case that these private foundations are free-riding on infrastructure paid for by other sources of funding. In effect, these foundations, piled high with cash, want the public to subsidize the research they choose to fund. Nice thought–and mostly they get away with it. But it is not science research budget reality.
There’s a bigger point here, much bigger, waiting to come out. People think that universities break even or better on their research operations. But they don’t. Not hardly. Chris points out that UCSF, which receives nearly $800m a year in federal grants, and $2b a year from all sources, has as shortfall in its research operations of $250m. Rest a moment. UCSF spends $250m per year of its own money to subsidize its research programs. The sources of this additional funding isn’t reported. Chris is quick to point out that public research ought to be subsidized by the public–it is a public good that we have robust research programs. For Chris, it is not that UCSF shouldn’t be subsidizing its research programs (or finding funding sources that will) but that if the federal government reduces its subsidies to support research infrastructure, then programs such as UCSF’s will simply break, fail. Instead, the federal government should be maintaining the negotiated levels of indirect cost funding, if not increasing indirect cost funding, so that universities like UCSF don’t have to find ways to subsidize research that federal agencies choose to support. Chris argues that the only folks “who make money on science are in business.” The return on investment for public science is mostly a long way off.
Chris makes the point that only if universities open up about their losses in supporting federal research (and all research for that matter) will folks opposed to the NIH cut to indirect cost rates have a way to win.
Well, there’s good material in what Chris writes. But there is another approach to the situation that I want to consider, and will use Chris’s excellent critique as part of it, though without the same politics regarding who is to blame. And this other approach has to do with opportunities and outcomes, not just funding (or lack thereof), not just improving the status quo and the current systems of research, and not just focusing on the intent to benefit the public.
This other approach starts with a broader but just as incisive critique of research progress, or outcomes that affect the public (and science, and our imaginations, and the personal narratives we live by, however casually. Let’s lay some things out. More money for research does not mean better research or more productive results or even better scientists. Money attracts all sorts, and lots of money attracts folks just as attracted to money and status as it does folks wanting to do research in the public interest.
In my two decades or so working in major research universities, my experience has been that the people who have most wanted to help others (“the public”) didn’t get federal funding, were looked down on by those that did get federal funding, that federal agencies didn’t much fund science and medicine that directly helped plain folks (wasn’t breakthrough science, not “high risk, high reward”–why fund plain help to folks when the reward was apparent, the risk low–where’s the bragging status in that?). Even when an agency did fund something that helped, it was only to start a program up (as an “experiment” or something) and after a year or three, the agency would lose interest. So, things like a telephone service to provide caregivers and pregnant women with the latest information about what chemicals and drugs cause birth defects had to be self-funded.
If “science” gets done, but there’s no one to support that science getting “out” to people who would benefit (and no, publication and conference presentations are a lousy way to move science into use), then we end up with a huge research spend and next to nothing to keep programs that connect science with the public or with companies. I’m not talking about “science communication” and I’m not talking about “technology transfer”–I’m talking about having the time and resources to respond to questions from the “outside” or to assist in adapting, or testing, or figuring out what someone in the big wide world needs. Much like when a farmer drives his flatbed truck with a dead cow and a can of blood onto the University of Wisconsin campus on a snowy weekend night and gets directed to the lab of Karl Paul Link, where the students set about to figure out why the farmer’s cows were bleeding to death. Not with any funding, initially. Never any federal funding. Because it was a real problem, and they could help. Problem identified–mold in sweet clover hay, produces dicumarol, result the blood thinner warfarin, saves President Eisenhower’s life after heart attack, kills a heck of a lot of rats, that bleed to death.
That sort of story repeats. Stuff gets done without the systematic, proposal-based, competitive research grant, where there is so much to be done to get a grant at all–announcing a grant competition, with a deadline, on a defined topic, following application rules down to the number of pages allowed, point size of font, margins, to be reviewed by a “panel” of other researchers basically representing the status quo of science, or the science approved by the federal agency conducting the grant competition. The proposal-based approach to federal funding goes back to the debates surrounding the formation of the National Research Foundation (which morphed into a strange doppleganger called the National Science Foundation)–should federal agencies fund people or projects? That was the administrative debate. Projects won out. As Charles Kidd tells it in American Universities and Federal Research (1959), funding people would put people at risk of being turned down–and that would ruin their career–and it would be much easier to turn down a proposal, on, say, the grounds that it wasn’t written as well as some other proposal, and the person could try again, having done a better job of writing A line of reasoning only an administrator could love. Missing is how Karl Paul Link would get support for helping a Wisconsin farmer with a dead cow on his truck in the middle of winter.
The present system of project-based federal science now dominates. If a company wants to work with a university lab, the lab is all but forced to write a proposal in the federal format. No informal collaborations allowed. Our Wisconsin farmer can sponsor research if he has a cow problem. This present system is unbearably complicated, and requires huge diligence under time pressure to get done properly. Yes, we can throw more money at this system, as Chris argues, or we can work to tear it down, simplify it, remove all the unproductive requirements, and even introduce alternative approaches to compete with this project-based system that is so expensive to operate. There’s an opportunity here. Rather than fight with the NIH and sue the bejezus out of them to try to retain the current, expensive, narrow, wasteful system, why not propose alternative systems that do only cost 15% overhead? If the federal government spends, what?, $45b per year on university and “basic” research, isn’t there some opening now to move some of that money into other approaches to doing science? Does every university have to have the same expensive lab facilities to do much the same research in a given area? If they are to have that same expensive lab facilities, ought they be pursuing different or competing or contrary research to the status quo, especially the status quo as established by government agencies?
Think about it. We have an opportunity to make a case for something new (or perhaps a return to somethings old that did work after a fashion, at least much better than the present bureaucrat and politician designed system. Vannevar Bush wanted to use the best and the brightest to get scientific and medical basic work done asap. Senator Harley Kilgore wanted to spread the work around so poor universities in poor states could get better–or, federal research funding as a species of pork. Reminds me of Python’s Ministry of Silly Walks grant program. Anyhow, Kilgore won, Bush lost. We have a chance here to revisit. How badly do we want to find out what’s killing our cows? Or is federal research really about careers for academics who study what the government wants to have studied, using a complicated proposal system reviewed by folks favorable to what the government wants studies–where there’s 100 applications for 10 awards?
Now let’s talk about outputs. Here are the stunning numbers. Since 1981, universities and affiliated foundations and institutes have acquired over 140,000 US patents on research inventions. At–conservative estimate–$10k per patent, we are looking at $1.4b in direct patenting costs. But it’s more than that–universities file 2 to 3 times more patent applications than issue as patents. And there are maintenance fees every few years to keep each patent, and there are foreign counterparts if a university wants to mess with the lives of people in other countries (or license/sell that right to mess to a company). Easily looking at $2b in patenting. And then there are the operating costs for the patent and licensing offices. That’s another (using very coarse estimates) of $200m per year. Forty years and we another $8b in expenditures–so maybe $10b has been spent by universities on patenting and licensing attempts, along with all the staff to operate another complicated, difficult, expensive side of research operations (I know, I worked that side of things for well over a decade).
And of all these 140,000+ patents, nearly 70,000 of which cite federal funding, and $10b of patenting and related expenditures–what has come of it? Next to nothing. Most university patents are unlicensed–meaning that those patents serve to prevent use of the patented inventions. Companies design around or simply go another direction–the patents work against the use of the research results from all that complicated funding that now is apparently threatened because of a change in indirect cost rates. Even where universities license patents, they mostly license exclusively–so all the companies who don’t get that one license have even more incentive to avoid the university’s research and to actively compete to make it obsolete. As two university licensing office directors told me, if they were told they would have to license an invention non-exclusively, they would refuse to handle it. And even when a university manages–it’s rare–to license some invention exclusively, it still rarely comes to anything–not developed, never a product. And in those few instances where there is a product, it is often priced as a monopoly–10x the price if it were a profitable non-exclusive, competitive product–and we might ask how such licensing behavior serves a public interest in competitively priced products, with multiple versions, choice of company to patronize, and the like. And we might sense we know the answer.
In this context, university administrators puffing about the potential loss of “life-saving” research looks thin, if not a mere rhetorical trick. They don’t have many examples of such outcomes. And the ones they typically cite often have sad backstories. UCLA’s “nicotine patch” fiction comes to mind. Or maybe these administrators are really that thoughtless and cruel. It is not the wordsmithed institutional intention that counts, but the outcomes–real advances at the frontiers of basic science, movement of people with talent to take these advances into account, and changes in practices (for the better, we hope) that give us better health, and better spam filters, and the like. Here, again, we might consider the opportunities rather than fight like the dickens to preserve or even improve a “system” that provides lots of careers for researchers, administrators, and lawyers but doesn’t much give a rat’s bottom for actual productive outcomes. “Research as an industry” is one phrase university administrators repeat. Just bring in the money, spend it, and get more money. I once asked a program officer at a federal agency what happens if in a three-year grant it is clear by year one that the research is not going anywhere. “Nothing,” he replied–we are too busy getting the next round of three-year grants ready to go. “Research as an industry” on the federal side, too.
So we presently have a research “system” that is over complicated, expensive, caters to the research directions favored by federal agencies, and produces piles of inventions that then are surrounded by patents and so are mostly never used, chasing away company use, or even professional use (some university inventions don’t require a product–ask the university lab medicine folks about the testing kits sold under licensed patents).
Yes, the NIH proposal to drop university indirect cost rates to a flat 15% threatens the present system of allocating research funding. But rather than fighting to salvage an expensive, unproductive, even cruel system of producing new science in the interest of the federal agencies–I’d rather see arguments for doing things much differently, drawing on past approaches that have worked (even if those approaches favor the random, the farmer’s dead cow, rather than a bureaucrat’s dream of competitive proposals on a deadline to be approved by the status quo, and then bordered by patents that chase most everyone off, making the research itself useless for 20 years or more).