Bending Science: How Special Interests Corrupt Public Health Research (review)

Identifying robust principles for using science to inform government regulation is among the most contested issues in health and environmental policy today. It is also one of the most intractable. Battles have been fought—repeatedly—over the objectivity of science, the degree to which science is separable from policy, and purported epidemics of “junk science” infecting public health and environmental laws. These debates have played an important role in revealing the limits of science and have made it more difficult for government officials to avoid public scrutiny by hiding behind a veneer of supposed scientific objectivity. They have also, however, encouraged indiscriminate attacks on the scientific bases of government policies and calls for standards of proof that are rarely attainable in the environmental and health sciences.

Professors Tom McGarity and Wendy Wagner, both prominent legal scholars at the University of Texas, are perhaps best described as scientific realists. In their new book Bending Science: How Special Interests Corrupt Public Health Research, they seek to strike a balance between acknowledging the epistemological limitations of science—it is not purely objective and often cannot be cleanly separated from policy—and recognizing the distinctive capacity of science to provide insights into the natural world and to help enhance human understanding over time. This is an admirable but precarious position.

Bending Science moves beyond traditional debates over the use of science in health and environmental policy. Recognizing the malleability of science and the inevitable intrusions of judgment into it, McGarity and Wagner expose systemic factors that make policy-relevant science uniquely vulnerable to manipulation and then document the strategies that are used to “bend” science towards specific ideological or economic ends. Bent science, under their definition, “works backwards from a result” and is incompatible with processes of “testing hypotheses with an open mind and using methods that scientists have accepted as valid and generally capable of replication.” While one could quibble with this characterization, the authors avoid a definitional morass by noting that no description of science includes research based on predetermined results. In other words, by defining bent science negatively, they seek to sidestep these deeper philosophical questions.

Early on, McGarity and Wagner observe that protecting science from manipulation is an age-old problem, but one that has taken on heightened urgency as reports of improper influence, most recently in federal agencies, have grown. They illustrate the perennial nature of these threats through a play by Henrik Ibsen, in which economic interests in a small town trump the warnings of a conscientious but politically naïve scientist. This basic scenario repeats itself throughout Bending Science in depressingly predictable terms. A great virtue of the book is the pains that the authors have taken to compile well-documented cases of bent science, by which they mean “either intentional or reckless ends-oriented” manipulation of scientific methods or results.

McGarity and Wagner’s use of unambiguous cases is intended to avoid debates over the legitimacy of the specific conduct in question. This decision is not without its costs, though, as their account captures only a relatively small fraction of the potential abuses, and it relies on a select sampling of often the most egregious cases of manipulation. I am sympathetic to this approach, however, as significant constraints are unavoidable in this context. Completeness is impossible, because by definition manipulation and distortion of regulatory science are difficult to track. Clearly aware of these limitations, the authors acknowledge that outright fraud and gross misrepresentation are rare. They argue instead that the cases they describe are simply more extreme examples of otherwise common tactics.

Many of the examples and statistics reported in Bending Science are shocking. According to a 2001 survey of epidemiologists, 50% were “harassed by someone with an ideological or economic interest in their research.” Similarly, a 2005 survey of more than 3,000 scientists reveals that 15% of the respondents admitted to altering research plans or results in response to pressure from a...