Science policy analysts usually distinguish between “science for policy” and “policy for science.” The former denotes a unique role of science in informing policy on all aspects of the national agenda – national security, foreign affairs, the economy, public health, education, natural resources, hazards, environmental protection, and much more. The latter covers the national strategy for investments in science – not just the amount and allocation of funding across disciplines, but also the balance of funding for universities versus in-house federal laboratories, the emphasis on STEM education, etc.
World War II, and development of radar, the atomic bomb, rocketry and jet aircraft, and even penicillin put policy for science in the national spotlight. Since then, political, military, and corporate leaders as well as the broader American public have supported substantial, sustained federal investments in research. With some oversimplification, the funds have been widely distributed across the sciences. In general the basic premise or policy has been that scientists and researchers within each of the many disciplines were best positioned to judge research opportunities and research quality, and to allocate their apportioned funds most effectively.
Though U.S. funding for science has been broad, it has been uneven. Some of the roots for this are simply historical. Others reflect the reality that experimental science tends to be more expensive than theoretical work, or the field of mathematics; or that the requirements of physical sciences for laboratory equipment such as particle accelerators and mainframe computers exceed those for, say, social science, which are more labor-intensive. Some of the allocations reflect political opinion on the maturity or utility of certain branches of science relative to others.
Not surprisingly, policymakers have changed the apportionment of funds from time to time. For example, in the 1950’s the Cold War was on and the country invested heavily in particle/nuclear physics. Soon the space race would consume resources for similar reasons. In the 1990’s investment in the biological sciences and technologies accelerated. In general, these shifts in funding have been few in number, broad in nature, and enduring. As America has looked over its shoulder at the experience of other nations, it’s seen evidence that governments aren’t expert at judging winners and losers – that is, which fields of endeavor will yield the greater or quicker payoffs.
Over time, however, the costs of science have been rising. Increasingly, research falls into the category of “big science.” De facto, political leaders appear to have settled into a policy default that science funding should never amount to more than a certain percentage of GDP. This has resulted in a squeeze on science. In turn it has encouraged some in the policy world to break ranks and propose stagnation or declines in funding allocations to certain fields in order to prolong growth in support for other science. A recent round of such proposals has put funding for geosciences and social sciences in the crosshairs. That in turn has prompted a search for counter-arguments from stakeholders in those fields. Such discussions are no doubt inevitable and are probably a good thing. However, given the polarized nature of today’s politics, the conversation runs the risk of making federal allocations for science a political battlefield rather than a non-partisan discussion.
The equation. A large part of the discussion centers around the idea of “innovation.” And that’s where my proposal for a new (?) “equation” comes in. The word “equation” is in quotes, because this is not a true equation such as f=ma, or e=mc2, with clearly defined physical parameters and precise mathematical relationships. Rather, it’s more in the spirit of the “equation” familiar from risk analysis that Risk=HazardxVulnerability. Here, risk, hazard, and vulnerability are not single, easily measured variables so much as they are conceptual ideas. And the equation itself is really more of a guide to thinking than a rigorous mathematical statement. It reminds us that the risk posed to society by floods or a K-T-like asteroid hit depends not just on our vulnerability (easily managed) to local floods and (the-end-of-life-as-we-know-it) vulnerability to the asteroid hit but also the frequency of that hazard: (everyday, somewhere) for the flood versus (every 50 million years or so?) for the asteroid strike.
With that preamble, then, consider this equation, offered in a similar spirit:
(SOCIETAL BENEFIT)i = INNOVATION x APPLICATION
This equation reminds us that the societal benefit (economic growth; national security; public health and safety; environmental quality, and more) resulting from innovation (and hence denoted by the subscript i) depends critically on how extensively that innovation is applied –not just on the innovation per se. Thus, the invention of the transistor, by itself, has arguably paid for all the science that has ever been done or ever will be done. But that is not just because of the invention as such. It is because of the extensive application of that invention across every field of human endeavor since. It follows that if the goal of publicly supported science is societal benefit, we ought to give as much primacy to applying science as we give to advancing it.
This is where the geosciences and the social sciences come in. Both contribute substantially to innovation intrinsically. But they make at-least comparable, and arguably far larger, contributions to application, and hence to societal benefit. In particular, they help us be as disciplined in our approach to application as we are to innovation itself. Our past attention to application has arguably been more happy-go-lucky. We see application as kind of miracle that follows effortlessly and inevitably from innovation, in the spirit of Sidney Harris’ famous cartoon.
More on the implications of this in a future post. In the meantime, better-educated and more-well-informed readers can tell me where they’ve seen this equation or something like it in economics or some other field.
 Look for an aside on the definition of “big” science in a future LOTRW post.
 Note that this is a choice. Instead of being resigned to limiting research funding to 2.7-2.8% of GDP, the U.S. might instead have opted (or could still opt) for raising that by as much as, say, a percent, and tested (or test) what difference that might make in GDP growth.
 Not to get too pointy-headed here; there is obviously societal benefit intrinsic to pure innovation and the joy it brings to the innovator or the innovator’s audience; but in cases of interest I’d argue that these benefits are dwarfed by larger societal benefits to much broader publics.
That’s unsupported by data or peer-reviewed analysis, but prove me wrong. To appreciate the scale of this: there are something like 100 million transistors in every cellphone integrated circuit. Intel estimates that this year the number of transistors worldwide is 1.2 sextillion. That’s 1.2×1021 (or roughly 2×1011 – 200 billion – for each human on the planet).
Numerical weather prediction didn’t break new ground in physics at the start. It took existing equations of motion for fluids and applied those equations to modeling the atmosphere. Not long after, however, the insights gained from that application generated innovation, creating a new field of physics: chaos theory.
In theory, I like the equation; but we don’t live in a theoretical world. In practice, it’s way too easy to swing and miss at what the “Applications” might ultimately be. Like the President of Digital (remember them?) saying that there was really no market for personal computers. Or the physicist who said that everything that would be invented had been – around the turn of the 20th Century. In the words of the Yogi [Berra], “It’s hard to make predictions, especially about the future.”
Your equation is a great way to look at the reasons for success after the fact; but I’ve never quite found it all that useful for choosing where to invest – and that’s what we’re really talking about. In my career heading up R/D organizations, I’ve taken a different approach to investing, which I offer for you and your other readers to pick at. I’ve always tried to build a portfolio – about 95% applications-driven, and 5% innovation-driven. Applications are selected in terms of what the organization can do, the [perceived] potential for meaningful impact, and the [perceived] potential for continued funding (i.e., is there a business in there?). Innovations are selected again on the basis of potential impact.
If we used something like this on a national scale, in your field I’d be investing heavily in weather prediction – huge impacts and certainly the need won’t go away. I’d be funding R/D to improve accuracy and precision of forecasts, and to lengthen the time for which we can make useful forecasts. Along with other nuts-and-bolts tasks (e.g., climate data collection), I’d spend 95% of my money there. The other 5% I’d spend on innovations and discovery-based research. That would include climate modeling, but not exclusively – I’d bet there’s lots of innovations in sensing still to be made, for example. I’d limit my climate modeling budget to about 2% of my total – there’s too much “turning the crank” and not enough real innovation to justify the amounts we’re spending.
Obviously only my opinion, but I’d defend it as better stewardship of the public moneys than we have now. Now if only we could get the politicians to really make choices instead of picking meaningless numbers out of the air…