Protecting Scientific Integrity
Sheldon Krimsky reviews Thomas O. McGarity and Wendy E. Wagner’s Bending Science: How Special Interests Corrupt Public Health Research.
Thomas O. McGarity and Wendy E. Wagner. Bending Science: How Special Interests Corrupt Public Health Research. Cambridge, MA: Harvard University Press, viii + 384pp.
Scientists, whether employed by universities, government, industry, or research institutes, have all received an education based on a common set of principles and methodology. And they have all been exposed to the same corpus of knowledge unique to their discipline. The universality of science demands nothing less. A molecular biologist whether in China, Croatia, or Kenya reads the same core journals, uses identical nomenclature, and follows similar laboratory investigation techniques.
Nevertheless scientists conducting their work in different sectors of society are faced with different social and political contexts, and they may not exercise the same autonomy to investigate the natural world and report on their findings. In the United States, for example, government scientists who are expected to submit papers for talks and publications may be required to submit their papers to be vetted by policy members of their agency. In turn, the policy members may censor the language and conclusions prior to publication or distribution. This practice is illustrated by the recent controversies over “political editing” of scientific documents on global warming.
The rights and responsibilities of government and industry scientists are not as clear as they are for academic scientists. And while considerable attention is paid to academic freedom and investigator autonomy within universities, there too scientists can exercise self-censorship when they wish to please an external funder who has a financial or political interest in the outcome of their research.
Bending Science explores the multifarious ways that science has been distorted when its goals and practices are superseded by profits, issues of liability, politics, and industrial competitiveness. The book is structured around six core themes: “Shaping Science” (the use of contract research to acquire support for preexisting views); “Hiding Science” (suppressing knowledge that politically or economically motivated funders dislike); “Attacking Science” (manufacturing uncertainty around sound scientific results); “Harassing Scientists” (using litigation to force scientists to defend their published results in court); “Packaging Science” (selecting scientists to reach a predetermined outcome); and “Spinning Science” (reinterpreting or falsely interpreting scientific results to meet non-scientific agendas). A wealth of both vignettes and historical cases illustrate these six themes.
This book is a welcome addition to a new body of work that explores the ideological underpinnings of efforts to construct or reconstruct a scientific record by methods that compromise scientific integrity. Among this new generation of work I would include my own book Science in the Private Interest, Seth Shulman’s Undermining Science, David Michael’s Doubt Is Their Product, and Chris Mooney’s The Republican War on Science.
The authors, both law professors at the University of Texas, contribute a unique legal and regulatory perspective on current trends that compromise the integrity of and the public trust in science. A notable discussion is the conflict between trade secrecy and the public’s right to know. The conflict becomes palpable when companies negotiate out-of-court tort settlements that block public access to discovery documents that may provide valuable knowledge about certain products’ occupational or public health dangers. The authors question whether courts should accept non-disclosure agreements as easily as they have, rather than balancing corporate confidentiality against “the public interest in disclosure of policy-relevant scientific information” (p. 123).
Notwithstanding the voluminous cases of manipulated science presented in the book, the authors are clear that the practice is the exception and not the rule. “Independent scientists can and regularly do conduct policy-relevant research in a disinterested way, without input from affected parties or financial inducements that cause them to tilt or skew the research toward a particular end” (p. 65). Where I might find some disagreement with the authors is in their assessment of the extent and depth of the problem of “bending science.” They state, “Only when the stakes are very high and parties become desperate does the temptation to shape science sometimes become irresistible” (p. 65).
My reading of the effect that corporate sponsorship of research has on tilting science is far less optimistic. Consider a 2005 Nature survey funded by the National Institutes of Health in which several thousand early- and mid-career scientists in the United States were asked to report on their own behaviors. When asked whether they ever change the design, methodology, or results of a study in response to pressure from a funding source, 20.5% mid-career scientists and 9.5% early-career scientists answered affirmatively. These and other studies show the extent to which corporate sponsorship and political censorship have contaminated the scientific record.
The final chapters offer potential solutions to the litany of science abuses that the book illustrates. Despite the authors’ parenthetical comments that most of science is sound, the stories do present a dismal state of affairs. One of the oversight reforms they suggest would protect against the politicization of federal advisory boards. There are over 1,000 advisory boards that serve federal agencies. Two rules apply to these advisory boards. Rule one is that scientists who have a substantial conflict of interest are excluded from serving. Rule two is that rule one can be waived. The stacking of members on advisory boards has been well documented. And while awarding waivers to conflicted scientific advisers is now well recognized, some agencies have not begun to address the problem.
One of the authors’ primary remedies is passage of legal requirements for disclosure of financial interests for scientists who provide scientific information or analysis to courts or regulators. Such disclosure should cover post-market research conducted on the safety and efficacy of federally regulated products. Of course, scientists providing evidence in litigation can be cross-examined. As one of my colleagues has said repeatedly, the most rigorous peer review he has experienced comes from cross-examination of his claims in the courtroom. But the authors suggest that judges, as gatekeepers of expert testimony, should pay more attention to the conflicts of interest of scientists before their words reach the jury.
According to the authors, disclosure to courts and regulators should be required by law and failure to disclose should be met with sanctions and criminal penalties. Federal regulatory agencies would then be able to ensure that “conflicts are factored into the relevant decisions without losing the value of the underlying scientific information altogether” (p. 239).
Another important governmental reform proposed by the authors is a requirement for data sharing in studies submitted to regulatory agencies. Currently trade secrecy protections prevent independent scientists from reviewing the data that support product safety and regulatory claims made by stakeholders.
Bending Science offers a clear and penetrating diagnosis of the problems that have arisen when academic science and business interests coalesce. The reforms proposed in the book, which are intended to promote data transparency and conflicts of interest within the fields of public health and the environment, are sober and sensible and deserve serious consideration. Disclosure and openness, however, will only get us so far toward a solution. During their graduate training, scientists must be made to understand that their profession demands autonomy and independence. All too often corporations treat scientists like lawyers—as their paid advocates. The responsibility of scientists must be, uncompromisingly, to their profession and to producing knowledge.