Commentary

The Threat to the Scientific Method

The legacy of Franklin Roosevelt is harming American science.

At the end of World War II, President Roosevelt asked Vannevar Bush, who oversaw the explosively successful Manhattan Project, if there was a way that the horde of scientists recruited to produce The Bomb could somehow be kept in government employment.

Within eight months, Bush sketched out a blueprint in which the Universities, not the government, would be the employers, but that the pay, either for faculty or for hired researchers, would actually originate from federal science agencies, cabinet departments, or the clandestines.

The consequences were obvious. Universities charge 50 percent overhead on federal grants, using these profitable science Department monies to pay for unprofitable Art and Music Departments. The seeds of political correctness—which requires big, expensive, expansive government—were planted as the schools became addicted to federal welfare.

The search for knowledge has become the search for funding, and funding agencies tend to frown upon negative results.”

Under unforgiving competition to secure funding for their institutions (and promotion for themselves) some scientists are behaving badly.

Last week, a technical publication, Journal of Vibration and Control, retracted sixty papers, after an internal investigation revealed a fraudulent “peer review and citation process” that greased the skids for a small number of authors to have an enormous number of citations in what is a prestigious engineering specialty. At least one of the authors even managed to review his own papers under an alias.

That’s symptomatic of a larger sickness raging in what should be our most sacrosanct of institutions. If we can no longer trust science, what do we have as the basis for knowledge?

It is a fact that the policy world—particularly the environmental policy world—claims to base policies on “science,” such as the reports of United Nations’ Intergovernmental Panel on Climate Change, or the U.S. Global Change Research Program’s periodical “National Assessments” of the impact of climate change on our country.

These influential documents are essentially large reviews of a voluminous scientific literature. The tragedy is that literature is being insidiously poisoned by the incentive structure for science itself.

The evidence is increasingly compelling. University of Montreal’s Danielle Fanelli has written several comprehensive reviews of the content of published science and he found, in the last twenty years, that the number of “positive” results is increasing dramatically. That’s when the data confirm a proposed hypothesis rather than suggesting rejection or modification.

In a real world where scientists are answering real questions, that would be impossible. People have not suddenly become smarter, except, perhaps at how to advance in academia. There, candidates for promotion in the sciences are basically asked two questions: What did you publish, and how much taxpayer money did you bring in to support your research?

If an Assistant Professor, up for tenure, answers either insufficiently, he’s likely to be looking for another job. It’s amazing how many of these wind up staffing Congressional Committees, or better yet, on programmatic committees for the big science agencies.

The money part is of paramount importance. At a tier-one University, to publish the requisite number of papers for promotion in, say, the Environmental Sciences, probably requires a minimum of $2.5 million. That’s a lot of overhead for the Germanic Languages Department.

Does anyone seriously think that a young researcher is going to get that kind of funding by going to federal agencies with a proposal that global warming’s amount and effects have been dramatically overblown (as they have)? The mere proposal threatens to derail everyone else’s gravy train. It won’t get funded, and the researcher soon won’t be paid.

Dr. John Loannidis, now at Stanford, may have been the first person to detect the illness when he wrote, in 2005, a then-iconoclastic paper, “Why most Published Research Findings are False”. His thesis is the demands to publish and get funding are so strong that many studies are poorly designed in order to force a positive result and rapid publication.

Since then, the number of retracted papers has gone through the roof. The winner of the 2013 Nobel Prize in Physiology or Medicine, Randy Scheckman—on the eve of his award—wrote a Guardian op-ed, “How journals like Nature, Cell and Science are damaging science”, and he vowed to never send them another manuscript.

Scheckman noted that publishing in Nature and Science is a ticket to tenure and enhanced research funding, but that these two magazines—the most “impactful” science magazines on earth, gravitate towards “flashy” science to draw attention themselves (and thereby inflating their “impact factors”). This is done at the expense the day-to-day grind science that is perhaps more important, but won’t land you on CNN. Knowing this, people will gravitate towards flashy fields, like global warming, at the expense of others, and burning our scientific talent for a mess of pottage.

So the search for knowledge has become the search for funding, and funding agencies tend to frown upon negative results. Who is seriously going to get a federal grant that can ultimately diminish the power of the federal government? No, instead we read in the recent National Assessment, silly positive associations, like global warming is associated with more mental illness. This can only mean that people in Richmond are loonier than here in Washington DC, and that they must be crazy in Miami.

That makes about as much sense as compromising science to serve the federal funding behemoth.

Patrick J. Michaels is the director of the Center for the Study of Science at the Cato Institute.