The Importance of Studying the Obvious
When Congressman Jeff Flake recently proposed eliminating NSF funding for political science research, it marked the third time in the past 6 years that a member of Congress had introduced legislation to cut federal support for some or all of social science.
To make his case, Congressman Flake picked out some examples of political science research that had been funded by the NSF and argued that the research in question— $600,000, in one case, to figure out if policymakers actually do what citizens want them to; $200,000 in another to understand why politicians make vague statements — was of little value to society, and therefore not worth the expense to taxpayers.
At first glance, the cost of these projects does seem hard to justify. But let’s imagine we performed the same exercise with grants from other NSF divisions, like Physics and Biology: We could easily find many examples of funded projects with equally hard to quantify benefits. $200,000 for “Quantum Phases and Non-Equilibrium Dynamics of Strongly Correlated Photons” anyone? Or how about $215,000 for “Characterization of Translation Silencing Complexes During oskar mRNA Transport and Localization: Trans-acting ZFactors, RNAi and RNA Decay in the Drosophila oocyte”?
I have no doubt that these are worthy projects, having passed a rigorous and highly competitive process of expert review. I’m not prepared to second-guess this process, and neither presumably is Congressman Flake. But NSF social science proposals are evaluated in exactly the same way as those in the physical and biological sciences. So why is social science research uniquely and repeatedly singled out for this sort of criticism?
Because it’s about us.
In brief, we don’t have any experience being ants or atoms, so if I tell you something about them that you didn’t know, it sounds exotic and non-obvious. It sounds like science. But everyone has experience being human, and so the vast majority of findings in social science coincide with something that we have either experienced or can imagine experiencing. The result is that social science all too often seems like common sense.
As social scientists have long pointed out, however, common sense can easily support opposite conclusions — which is why politicians on both ends of the political spectrum invoke it in support of their arguments, even as they disagree bitterly.
For the same reason, it is easy to come up with plausible sounding hypotheses about even highly complex social and political problems — the causes of the recent financial crisis, or the likely impact of the new health care law — but extremely difficult to prove any of them right or wrong.
All of this puts social science in an awkward position with respect to public perception: Answering even the simplest social science questions is painstaking work; yet the answers tend to seem obvious. Worse, when results from social science do not conform to our intuitions, our reaction is not to be surprised and impressed, but rather to dismiss them.
Recent research, for example, strongly suggests that most success stories, from Facebook to Shakespeare, are accidental products of randomness and cumulative advantage; yet try telling that to a social media pundit, or a Shakespeare buff.
In the same vein, no matter how many times economists explain that cutting government spending in the midst of a recession is generally bad for economic growth, politicians continue to pursue austerity policies based on the flawed but intuitive analogy that government debt is the same as individual debt, and so should be handled the same way.
How can we better appreciate the limits of our intuition, and hence the need to support the scientific investigation of human affairs? One interesting possibility is raised by the arrival of “big data,” increasingly derived from digital communications, social media, mobile apps, and e-commerce sites. The potential for all these data to yield insight into human behavior is tantalizing; yet, as recent results about social contagion and the role of “influencers” highlight, the insights are often at odds with our intuition. Clearly a more rigorous, scientific approach is needed.
For this reason, companies like Facebook, Google, and Microsoft, where I now work, are beefing up their research labs both with computer scientists, who have the technical skills to handle huge datasets, and social scientists, whose job it is to ask the right questions. In fact, the emerging intersection of computer and social science–what some people are calling computational social science — is one of the hottest areas of research today.
Given the increasing interest that supposedly economically ruthless corporations are showing in the social sciences, it is particularly ironic that elements of congress continue to see so little value in them.
Surely politicians, of all people, do not believe that social, political, and economic issues are unimportant. The impact of illegal immigrants on the US economy, the origins and consequences of wealth inequality in US society, and the dynamics of economic and political development in the Middle East and Africa — all questions that the NSF funds social scientists to think about — are clearly of tremendous relevance to Americans’ collective well being and security, and hence of concern to politicians.
What they need to understand is that they cannot solve these crucial problems on their own. Finding answers will require serious investment, and that will only happen when the politicians grasp that social science is real — and essential — science.
by Duncan Watts
Originally posted at The Harvard Business Review