Political control of research

By | May 17, 2013

polar-bears-standingThere has been an interesting exchange on one of the Chronicle of Higher Education blogs relating to a story on the efforts of Rep. Lamar Smith (R-Texas) to manage the funding of research grants by the National Science Foundation.  This effort is part of Rep. Smith and the Republican majority in the House to politicize and eventually eliminate NSF research funding . One source describes it thusly:

“The bill says that any research done using federal funds (which is the majority of research done in the United States) must have its results and finding approved by the Committee on Commerce, Science, and Transportation of the Senate and the Committee on Science, Space, and Technology of the House of Representatives. If the findings are not agreed to, the research is taken from the researchers and disposed of by Congress as it sees fit.”

Rep. Smith’s long-term agenda is quite clear: kill off any independent research in the country that does not explicitly support the Republican political agenda, made abundantly clear by his attitude toward climate research.

As usual, I had to have my say ; here’s my response to the original posting:

I served as an NSF Program Manager in the early-to-mid 1980’s; I’ve also been both the recipient of NSF grants and had proposals declined, so I’m very familiar with the game from all sides. While I might not be one of Manoflamancha’s “best and brightest”, I and my colleagues conducted thoroughly professional and thorough reviews of the research proposals that we received, obtaining reviews from a wide range of highly competent scientists who took time from their own research to help us evaluate the next generation of projects in their field, and then synthesizing these comments into helpful guidance for the proposers and working closely with them to shape their projects more effectively. As a result of this very demanding and time-consuming process, we helped to generate a large number of ground-breaking research projects, including the first empirical studies of the impacts of personal computing on white-collar work.

There is no question that political elements enter into NSF’s funding of research. At the time, our “adversary” was Sen. William Proxmire of Wisconsin – I put that in quotes because despite his stern criticism of some NSF grants, he firmly believed in the value of science and its ability to improve lives; we just occasionally disagreed on how to do it. There was also internal academic politics, involving pressure to fund large well-established researchers rather than innovative new proposals, so we always had to juggle these elements as well. All program managers learn quickly how the research funding game is played; but the result of the game is that almost all of the research funded by NSF is of high quality, and a lot of high-quality proposals are not supported.

Rep. Smith is trying to impose an entirely different political agenda on research funding. He is hardly agenda-free; he is an explicit denier of climate change, among other things. He is seeking to impose a political review on all federally-supported research to be sure that the findings confirm the prejudices of his political supporters. If he succeeds, there will be no research supported that is not cleared first by his committee, which includes numerous self-proclaimed anti-science members. Independent research and research that might reach independent conclusions will be effectively dead in this country which is entirely in accordance with the Republican agenda. This whole dialogue needs to framed on terms of what its real purpose is – not screening for quality, but for conformance to a rigid and generally anti-scientific political agenda.

Following a brief and extremely general diatribe against all social science research and an accusation that I was furthering my own political agenda, I responded further:

I am trained as an organizational psychologist, largely through the Institute for Social Research at the University of Michigan, one of the most respected programs in the country. I have taught research methods and statistics from beginning through structural equation modeling at several well-ranked universities, and I will go toe to toe with you on any statistical or measurement issue you care to debate (although I doubt it would interest anyone else). And yes, I do use “canned programs” such as SPSS, SAS, AMOS, and SmartPLS, largely because I don’t see all that much virtue in spending the next 15 years inverting matrices by hand, as apparently you feel is desirable (although I also doubt that you practice.)

Behavioral/social science is not physics (despite the degree of “physics envy” that sometimes leads my colleagues to drape their research in complex statistics that the data won’t support. It is a science, in that it has established rules of inference and procedures in place to reasonably protect against bad reasoning. Statistical analysis is helpful, as long as it’s adapted to the data. Statistics at bottom is simply a way of helping us avoid committing the inferential errors we’re all prone to. It doesn’t “prove” anything – but neither does it do any damage. Ultimately, the utility and validity of any research study, in “hard” or behavioral sciences, depends on the quality of the research questions asked, the nature of the data collected relating to the question, the acceptability of the research procedures, the suitability of the data to the analytical procedures employed, and the reasonableness of the conclusions based on the analysis. This is not a series of assessments to be made off the cuff by someone with no training or experience in research – particularly not by a congressman from Texas who cavalierly dismisses all scientific evidence in the case of climate change and whose publicly proclaimed agenda includes the systematic subordination of women, the suppression of gay people, and, it might be reasonably inferred, white supremacy.

Of course I am and have been political; no one can operate in any organization today without being sensitive to both organizational and national politics. Politics is, as Robert Dahl once commented, all about “who gets what, when, and how.” When everyone else is devoted to acquiring your resources for their purposes, you’d simply be a fool not to join the process, if only to protect yourself in the clinches. Obviously, you have your own political agenda, which may or may not coincide with Rep. Smith’s, which includes deriding or suppressing any behavioral or attitudinal research that might call your agenda into question. The difference between us is that I encourage research -good research – regardless of its direction, and if it’s well done, I’ll give credibility to the findings. I’m not afraid of new knowledge, as I’m afraid all too many folks today are.

I apologize for taking up so much of the blog with reprints of dialogues engaged in elsewhere, but I feel strongly enough about this potential perversion of research at NSF and elsewhere that I’d like to see the dialogue broadened. I’m sure there will be further exchanges on this subject; watch this space.