27 May 2011

Dan Sarewitz on Senator Coburn's New Report

[EDITOR'S NOTE: This is a guest post by Dan Sarewitz, professor at ASU and co-director of the Consortium for Science, Policy and Outcomes (and pictured below). To his surprise, Dan found his work cited approvingly in a new report on the National Science Foundation just released by Senator Tom Coburn (R-OK). This post has Dan's reaction.  David Bruggeman has more here.]

Senator Tom Coburn (R-OK), who has a reputation as a straight-shooting, no-nonsense conservative, is also, it appears a supporter of national industrial policy, something that conservatives typically hate. He has just issued a report that alleges to be a hard-hitting critique of the National Science Foundation, but it’s mostly just an attack on government funding of social science research, thus continuing a conservative tradition that dates back to the debates over the initial creation of NSF in the late 1940s.

I’ll get to the social science stuff in a minute, but for now let’s focus on the fact that Senator Coburn prominently—and apparently approvingly—quotes, um, ME! (Please see page 12 of his report.) In a recent Nature column (among lots of other places) I argue that US civilian R&D agencies are not appropriately structured to catalyze technological innovation or progress rapidly toward desired societal outcomes, and that this institutional weakness remains significantly camouflaged by the legacy of DOD and the military-industrial-university complex,which powered technological innovation and economic growth in the decades following World War II.

I’m pleased that Senator Coburn finds this critique to be compelling, and can only infer, then, than he would agree that what’s needed is a much more coherent and strategic approach for linking knowledge creation to knowledge use and problem solving—a strategy that, in the olden days might have been called “industrial policy” and now we might term “innovation system policy.” It’s only slightly ironic, I guess, that the (still admittedly limited) understanding we have of how innovation systems work—the basis of my critique that he so flatteringly cites—is, well, rooted in the social sciences that he wants to de-fund.

(On this latter point, in part the Senator’s report is just another example of Republicans using the banner of fiscal responsibility to attack programs that they happen not to like but whose elimination can have no conceivable impact on fiscal responsibility. The entire social and behavior science budget at NSF ($252 million) amounts to all of 3.6% of the total NSF budget, 0.3% of the civilian R&D budget, and 0.006% of the federal budget. Attacking social science is good conservative politics, but it has nothing to do with serious budget policy.)

Moreover, much of Senator Coburn’s report details the sorts of random, petty abuses that are simply unavoidable in any complex bureaucracy like NSF (my goodness, an NSF employee was caught watching lots of porn! And another one scheduled a work trip so he could visit his girlfriend!). Yet the report does touch on a problematic aspect of civilian science policy that has managed to escape serious political scrutiny for 60 years, even though it is fundamentally incoherent. In specific, Senator Coburn is concerned that NSF’s research is insufficiently “transformative.” He cites survey work (more social science!!) showing that most NSF peer reviewers believe that only a small percentage of the proposals they review are “transformative.” He then goes on to list fifty or so examples of funded projects (“Are people more or less racially-focused when seeking love on-line in the Obama era?”) whose potential “transformativeness” he questions. This approach follows the tradition of Senator William Proxmire’s Golden Fleece Awards of the 1970s and 1980s, and it was probably fun to do. But could it be a coincidence that all of the projects he singles out have titles that a lay person can understand, and that many of them are social science projects? Or have Senator Coburn and his staff determined that all of the work that NSF funds in subatomic particle physics, deep mantle geochemistry, and molecular genetics is genuinely “transformative”?

The political rhetoric of basic academic science for the past 50 years has been basically this: leave us alone to follow our curiosity wherever it may lead, and the payback to society will be enormous—we’ll cure cancer, create the next industrial revolutions, clean up the environment, and everyone will get wealthier in the process. And if you try to tell us what to do you’ll only screw things up.

But as budgets have increased over the decades, as science and politics have increasingly come into conflict, and as the promised benefits of science have often proven elusive indeed, the question of how this path from knowledge to benefit really works becomes hard to avoid. Science advocates up the ante by promising more “transformative” research, but of course science is only one element of a complex set of factors that lead to progress on difficult social problems. When frustration sets in, what’s a politician to do?

I’m a big fan of (and have been generously supported by) NSF (including support for work that has informed my own critiques of the policy model that sustains NSF), but have long feared that the simpleminded elegance of the political rhetoric of academic basic research will someday turn out to be a source of serious political vulnerability for publicly supported science. Is Senator Coburn just being a heel, or has he discovered American science’s Achilles’ Heel?


  1. I am told that blogger has comment problems. Add that to the list of problems with blogger. I am told that if you unclick the "stay signed in" box that comments will go through.

    I may consider switching to WordPress if Blogger continues to have issues.

    Thanks for your patience.

  2. My experience with the NSF model (and other agencies that have adopted the NSF model) is that research panels composed of scientists not only determine whether the science is good (which they are uniquely qualified to do) but whether that discipline and approach will be helpful to decision and policy makers (which they are not).

    At an unnamed agency, I once tried to allow stakeholders to participate on panels, but that was against that particular research culture, and I was told that that was unacceptable. I think the resulting disconnect is a serious problem and leads to many wasted tax dollars.

    As does the fact that all science agencies seem to think that all research is within their mission, leading to immense duplication of effort. But that's another problem. I would recommend to Coburn that he pursue the idea of "one topic, one agency." For example, now we have several agencies.. call them agencies A B and C, who study the same thing, and we have B scientists applying for grants from A and C, with A scientists applying for grants from B and C, and so forth. By the time all the panels have been held and grants handed out, the overhead must cost as much or more than the grants themselves.

    If I were a Senator, I would propose that we take the science agencies and assign them topics, with the concept of ultimately moving all the scientists studying that topic into one agency, and that agency responsible for external grants in the field as well, so extramural and intramural efforts would be complementary.

  3. "leave us alone and .."

    * we will do lame rehashes of the last good published thing we did;
    * help funnel money to colleagues (who are not a threat);
    * burrow in, play the game and don't take risks despite what "our curiosity" might prefer

    Congressional faux outrage is part of the process. The greater risks inherent in government funding are politicized inertia, ideological capture and groupthink. Those are bigger threats than Coburn will ever be.

  4. "Science is only one element of a complex set of factors that lead to progress on difficult social problems."

    This reminds me of my favorite aphorisms on innovation policy: "It's not the 'R.' And it's not the 'D.' It's the '&'." (Philip Auerswald, as quoted by Lew Branscomb)

  5. "It's only X amount of dollars!"

    "It won't matter to the total budget!"

    Yeah. That's what they all say. Don't cut my little piece out of the budget - it's too small to matter. So in the end, nothing is cut, and everyone is happy.

  6. I argue that US civilian R&D agencies are not appropriately structured to catalyze technological innovation or progress rapidly toward desired societal outcomes, and that this institutional weakness remains significantly camouflaged by the legacy of DOD and the military-industrial-university complex,which powered technological innovation and economic growth in the decades following World War II.

    This actually parallels the Libertarian/Austrian school critique of market distortions caused by government funded R&D. You get too many folks doing basic research and not enough folks/infrastructure doing development/testing/sales/marketing/etc. to turn that basic research into a useful product or service.

    This ties in to your other emerging theme on the popular/academic disdain for the merely useful.

    Looking forward to more innovation policy stuff.

  7. I'm going to assume that Dan actually has something intelligent to say and that this rant was shot from the hip before his brain was fully engaged. Otherwise, I have to wonder about his ability to craft a rational argument.

  8. There's an interesting commentary here that makes some similar points to jtstults regarding the importance of development/testing/etc. in addition to basic research on pharmaceuticals.

  9. Mike, thanks for linking to the McArdle piece. This one quote is worth the price of admission

    " At Sanofi, the goal now is to strive for "open innovation," which involves looking for new research and ideas both internally and externally -- for example, at universities and hospitals. In addition, the company is focusing on first understanding a disease and then figuring out what tools might be effective in treating it, rather than identifying a potential tool first and then looking for a disease area in which it could be helpful."

    I think that that quote applies more broadly than the pharma biz. The conflict of interest involved in tool makers or model makers determining whether the tools or models are useful undermines the very empirical base that science claims for itself.

  10. I dislike the term 'transformative science'. It's not that such science doesn't exist, it's that it is very difficult to forecast in advance which results will be transformative.

    Most transformative science comes as an unexpected result of 'normal science'. And if government wants to produce transformative science, it will need to fund normal science at a level sufficient to keep it ticking over.

    Maybe that's hard to sell to the electorate. In that case, we will have to forgo the (unanticipated) benefits. But I think science is already suffering from the drive to fund mainly 'hot areas'. Funding 500 grants to look for new more efficient methods of solar energy conversion only makes sense if you have reasonable expectation that the next big thing in solar energy conversion is likely to come from that area. But, in fact, just because we've already spent a great deal of money and effort on directed searches in that area, it's narrowed the chances there is a big breakthrough waiting there.

    Government is no better at picking research winners and losers than it is at industrial policy.

  11. Roger,

    You should find this supportive of your ideas.


    "He concludes by recommending the Hartwell paper and Pielke Jr’s book The Climate Fix."

    later Prof Curry comments:

    "Message to climate scientists (especially in the U.S., and especially the climate/science establishment): now that the UNFCCC treaties do not seem to be desired by even the most progressive U.S. administration in recent (and likely future) decade, please rethink your allegiance to the UNFCCC/IPCC ideology. Let’s get back to doing climate science as it should be done: challenging every aspect of the climate science to broaden and deepen our understanding of the climate system and the full range of possible future climate scenarios associated with both natural climate variability and anthropogenically forced climate change. And supporting policy makers in developing and assessing a broad range of robust, no/low regrets policy options."


    And of course my favorite "no/low regrets policy option" is:

    Investment in "green" nuclear like China

    "China Takes Lead in Race for Clean Nuclear Power"

    "China has officially announced it will launch a program to develop a thorium-fueled molten-salt nuclear reactor, taking a crucial step towards shifting to nuclear power as a primary energy source."