I’m a friend of science and of government funding for scientific research (though I do insist that these are not the same thing). I spend much of my professional life explaining and advocating the very economic arguments for research funding laid out nicely in President Obama’s recent speech at the National Academy of Sciences. This is a thematically coherent statement of philosophy, likely to go down as a seminal document in the history of the administration, though in substance it contains little not already embraced on a bipartisan basis and telegraphed in dollars in the stimulus bill, the FY 2009 “omnibus” appropriation, and the FY 2010 budget, which is still in “outline” stage as of this writing.
The headline, I suppose, is a commitment to achieve a “societal” level of R&D of 3% of GDP, up from the most recently measured 2.57% and exceeding the 2.9% registered at the height of the Sputnik-fired space race. Societal R&D means all activities (basic, applied, development), all funders (federal, industrial, state, and philanthropic), and all performers (government, university and industry). As you might expect, there’s a complex interplay among funders and performers, cross-cut by research type. To understand the distribution of funding across these dimensions, see materials available from the NSF Division of Science Resource Statistics, including the masterful annual summary in Science and Engineering Indicators.
Since the government component of R&D is the only one the President and Congress can directly control, and since the thematic emphasis is on civilian R&D, what this comes down to is a theoretical justification for the 10-year doubling of the research budgets for the physical science agencies (NSF, NIST and DOE) — analogous to what was done for NIH in the 1990s and already authorized for these other three agencies in the America COMPETES Act but not until now committed to in a presidential budget.
On the plus side, the president’s speech evinced a clear understanding of the positive externalities of basic R&D that cause industry to under-invest in it, and which leave basic science an obligation of government worldwide: Most of the G8 nations typically achieve a rate of societal R&D of at least 2% of GDP, again with private investments in applied and development work supplemented by substantial government-funded contributions to basic research. However, I grimaced a bit at several elements:
First, while it’s true we’ve been foolish in not making the R&D tax credit permanent until now (and really, the refusal of the Congress to do so has been mainly a matter of budget gamesmanship, having to do with making projected “tax expenditures” look lower than they eventually will be when the credit is renewed each year at the last minute), all that the uncertainty about its status does is affect the “hurdle” rate for corporate R&D projects. Except when paid by the government, companies fund mainly product development. The main effect of the uncertainty over the credit has been to make corporate R&D marginally more short-term and risk-averse than it would be otherwise — but that’s really not saying much any more. The economic arguments that the president is making are mainly about basic research (perhaps directed basic research), which is now essentially exclusively funded by government, and performed either in universities or government laboratories.
Second, color me still skeptical about the chances of ARPA-E, the new Department of Energy agency supposedly to be modeled on the world-known DARPA, the Defense Advanced Research Projects Agency. In the first place, the ARPA-E is budgeted an order of magnitude lower than DARPA. Next, ARPA works so well as a tool to shape industrial innovation because it is understood that the DOD weapons-procurement budget stands behind innovations funded in the ARPA program. However, innovative energy technology is purchased primarily by end users rather than the federal bureaucracy itself, which is primarily a steering entity rather than a customer. There will be a disconnect. Precisely because of differences in what each agency is accountable for, the cultures of program management are very, very different. I frankly do not expect that DOE has the management capacity right now to run a truly innovative advanced research program, though I will be happy to be proved wrong. UPDATE (5/2): Don’t believe me about DOE? Read Secretary Chu’s own words. UPDATE (5/19). It has been pointed out to me that under the America COMPETES act, the ARPA-E director may appoint between 70 and 120 program managers exempt from all civil service laws. That’s more than I’d realized, but I suspect still inadequate to the cultural challenge.
Finally, and I’ll come back to this in a future post, I am concerned about the implications and unintended consequences of doubling the physical-science agency research budgets. I actually don’t think that doubling has worked out all that well at the NIH, and I believe there are reasons to doubt we’ll be getting all or precisely what we expect if we try it in the physical sciences. Unlike the DOE, the NSF is one of my very favorite agencies, and is truly committed to eventual economic outcomes of the basic research it funds, but it labors in the face of some daunting societal roadblocks. More in a follow-up post as I have time.