In its first year, the Biden administration launched a fast-track Scientific Integrity Task Force, intended to “lift up the voices of Federal scientists of many perspectives and backgrounds” and put scientific integrity “paramount in Federal governance for years to come.” The task force took a “whole-of-government” approach to ensuring the scientific integrity of federally funded research and included representatives from the 21 federal agencies that maintain scientific-research programs. For those with a high pain threshold, the final report may be seen here.
Prominent among the move’s critics have been the Council on Governmental Relations (a consortium of research universities) and the Association of Research Integrity Officers (university staff who conduct in-house investigations into alleged research misconduct). Together, these groups submitted nearly 200 comments representing their respective institutions, most of them opposing the proposed rule changes.
As with all things governmental, one looks at this spectacle and asks “why?”As with all things governmental, one looks at this spectacle and asks “why?” It’s not like anyone is in favor of scientific misconduct. As I have written elsewhere, the rules for trustworthy science are pretty straightforward: don’t lie, cheat, steal, or be beastly to your underlings, basically. The problem is that there seems to be a lot of scientific disintegrity nevertheless. Some of it is high-visibility, as in the recent forced resignation of Stanford’s president, Marc Tessier-Lavigne. Other manifestations appear to be systemic, as in the alarmingly high number of retractions of scientific papers: more than 10,000 in 2023, according to Nature. Nor is the problem an isolated one. An alarmingly large proportion of the roughly 460,000 scientific papers published in 2022 have never been read by anyone, nor is it clear that this voluminous output does anything to advance scientific knowledge. The scientific literature is rife with irreproducible or shoddy publications, many of them finding refuge in so-called predatory journals, which will take anything from anyone looking to beef up a CV, provided one is willing to pay hefty “page charges.”
Nor has there been an absence of means to detect and punish research misconduct. Since 1981, the Public Health Service has maintained an Office of Research Integrity (ORI), charged with—well, the name says it all. The National Science Foundation (NSF) polices research integrity through its own inspector general’s office. And, as mentioned above, universities maintain their own offices and policies charged with ensuring that the scientists employed by them follow the rules. Little of it has been effective: The National Institutes of Health (NIH) asserts that the problem is small, with only about 1.5 percent of NIH-funded researchers estimated to have engaged in research misconduct. That amounts to more than 2,300 cases of misconduct per year, however. Other surveys put the misconduct rate as high as 23 percent, raising the estimated number of cases alarmingly. For all that, the ORI typically finds only 10 instances of misconduct a year, and sometimes fewer. No wonder there have been recent calls to criminalize research misconduct. No wonder, too, that some scientists are suspicious of the government’s attempt to solve with an oversight body a problem that has been besting established oversight bodies for years.
There is a bigger picture in play, however. As demoralizing as research misconduct is, we should hardly be surprised by its occurrence. The unscrupulous, like the poor, will always be with us. More demoralizing is how research misconduct is actually incentivized in our modern science ecosystem. Few people want to acknowledge that.
In 1950, the U.S. federal government embarked on a radical experiment: to federalize university science. Whereas prior to World War II government involvement in university science had been minuscule, the basic knowledge that had built up in the universities had proven invaluable. Thus it was determined that government should mobilize university science to help secure the peace. This was the rationale behind establishing the National Science Foundation, which would accomplish this task through issuing grants to university scientists, so that they could pursue their curiosity as they explored science’s frontiers of knowledge.
Research misconduct is actually incentivized in our modern science ecosystem.As is often said when things go sideways, it seemed like a good idea at the time. The “NSF model” has since spread throughout the federal bureaucracy (hence those 21 representatives to the Biden administration’s Scientific Integrity Task Force). Some of this simply extended the existing science-oriented missions of agencies like the NIH, the U.S. Department of Agriculture, the Department of Defense, and the U.S. Geological Survey. Whereas these agencies had always carried out their own in-house (“intramural”) research programs, they all now have developed “extramural” research-grant programs to fund university research. The biggest player here is the NIH, which pumps about half of the nearly $100 billion of research funds now streaming into universities, followed closely by the Department of Defense. The NSF, which kicked off the whole experiment, presently accounts for about 10 percent of university research funding. Government funding now accounts for the majority (about 60 percent) of university research funding.
Such enormous streams of money will always attract grifters, which feeds nicely into the bad-actors narrative for scientific misconduct. Such people are using public funds in a fraudulent manner, after all, and they should be held accountable. The bad-actors narrative by itself, however, does not account for how misconduct is actually incentivized by our present system of research funding.
Only about two-thirds of that $100 billion actually supports the work of scientists. The other one-third streams into university coffers as reimbursement for overhead costs that come from having scientists do their work. These are known as indirect costs, and they account for the administrative and infrastructure costs that go along with scientists doing their work. Reimbursement for overhead costs is a perfectly legitimate thing for universities to claim, but, in the American science ecosystem, they look to have become disproportionate to actual overheads. In research grants, indirect costs are assessed as a surcharge on the monies granted to scientists, i.e., the direct costs. Presently, indirect costs average out at about 53 percent of direct costs (or as high as 90 percent at some universities). Compared to overhead costs in other countries with similar national research programs, overhead costs at American universities are two to four times higher.
Because administrative costs can legitimately be claimed as an indirect cost, university administrations are strongly motivated to maximize indirect-cost revenues, which allow research grants to fuel administrative growth. The instrument administrators use is pressure on scientists to bring in ever more grant money. If “publish or perish” is a contributor to scientific misconduct, as it has been claimed to be, you need look no further than this: Scientists are now judged by how “productive” (measured by numbers of papers published) they are and no longer by what they discover. In this scenario, the science itself no longer matters: If you don’t keep the cash flowing in, you don’t have a career. Hence the pervasive and perverse incentive to churn out publications, never mind whether they report the discovery of the century or are flotsam cast upon the vast sea of the mordantly named scientific “literature.”
Since the federal toe was jammed into the door in 1950, the entire culture of science has changed.Indirect costs are only the visible tip of a much larger problem. Since the federal toe was jammed into the door in 1950, the entire culture of science has changed, and not for the betterment of scientific discovery. Prior to the war, science operated in what we might call a “small science ecosystem.” Government support of scientific research was small and mostly looked upon with suspicion. Scientists would approach research ad hoc as interesting questions or opportunities came up, and these questions could range from the habits of mayflies to the structure of the atom. Universities would tap into institutional funds to support their efforts. Private foundations would support work that interested them. Private companies could provide support where needed, as in developing the first cyclotrons. Overall, the small science ecosystem was dominated by an ethic of discovery: Scientists’ behaviors and incentives gravitated toward that.
Since 1950, however, the science ecosystem has morphed into a “big science cartel,” united through an interwoven network of self-aggrandizing actors who hold a common interest, not around science but around capturing research funds. University administrations are one such actor, but there are many others. A spendthrift Congress is another, comprising legislators whose election prospects are tied to delivering the goods to constituents, which, in many districts, includes universities who look with favor on representatives who can keep the research money flowing in. Those 21 federal agencies represented on the task force constitute another crucial player: bureaucratic entities whose value and very existence is tied to capturing dollars from the federal budget. In some instances, such as the corrupt relationship of NIH bureaucrats to the pharmaceutical industry, science-based agencies may have their fingers in some very profitable private interests. Lately, even the academic publishing industry has been drawn into the big science cartel, fueled by generous page charges that are increasingly built into research-grant budgets (thus inflating direct costs).
We can now begin to make sense of the dust-up between the bureaucracy-based Scientific Integrity Task Force and the university-based Council on Governmental Relations. Neither is concerned so much with protecting the integrity of science; they merely differ on who shall be the enforcers—universities or federal bureaucracies—over the working scientists who bring in the crucial cash. Neither have any interest in dealing with the most significant scientific-integrity issues facing us today, ranging from overt politicization to corruption verging on criminality at the highest levels.
At this point, it’s apt to recall Henry Kissinger’s famous quote about the 1980s Iran-Iraq war—that it’s a “pity both sides can’t lose.” Yet, while I’m no friend of the shenanigans of university administrations and the games they play, I’m far more concerned about the Biden administration’s move to complete the federalization of university science begun in 1950, which may finally squash the very people who are the most effective custodians of scientific integrity: scientists themselves.
[Note: An earlier version of this article slightly understated the number of misconduct cases found by the Office of Research Integrity. We regret the error.]
Scott Turner is director of the Intrusion of Diversity project at the National Association of Scholars.