Harvard Should Protect Whistleblowers

Harvard Business School professor Francesca Gino, a prominent academic studying honesty, has recently made headlines for allegations that she authored papers containing fraudulent data. The accusations stem from four posts written and published by Leif D. Nelson, Joseph P. Simmons, and Uri Simonsohn on their blog, Data Colada.

The Data Colada posts present a careful and devastating indictment of Gino’s work. For instance, one post describes a dataset containing a demographic question asking students to report their class year. Instead of a typical response (e.g. “junior” or “class of 2016”), 20 respondents gave the nonsensical answer of “Harvard.” The data collected from these respondents overwhelmingly supported the study’s hypothesis, strongly suggesting that the data were tampered with. As one Harvard professor put it, “We are never going to get stronger evidence of fraud, short of the researcher cackling maniacally as they manipulate the Excel spreadsheet on camera.”

Data Colada’s report led Harvard to undertake an internal investigation into Gino’s work and ultimately place her on unpaid administrative leave. Gino responded by suing Harvard — as well as Nelson, Simmons, and Simonsohn — for at least $25 million, due to the impact of the allegations on her career and reputation.

Gino’s lawsuit is likely to fail — similar lawsuits have not succeeded, and many consider hers unconvincing — but the involved parties now face the prospect of a protracted legal battle and hundreds of thousands of dollars of legal fees. While Harvard has ample resources to defend itself, Nelson, Simmons, and Simonsohn do not; they are private citizens who are not backed by a $50.9 billion endowment.

We, along with many members of the academic community, are deeply concerned that this lawsuit will deter future efforts to uncover fraudulent research. A GoFundMe aiming to finance Data Colada’s legal defense has raised more than $357,000 from thousands of people as of this writing, including multiple Harvard professors, but we believe that Harvard should cover these costs. Simine Vazire, one of the GoFundMe’s organizers, agrees, recently telling The Crimson that a goal of the GoFundMe is to “put some pressure on Harvard and on the three data quality researchers’ universities to step up and support.”

The allegations against Gino occur amidst broader challenges related to ensuring the reliability of research. Influential social science experiments have been recreated but failed to yield the original results. In the biomedical sciences, nearly 2 percent of papers include figures with images that appear to be deliberately manipulated, while a similar percentage of scientists admit to having engaged in fraud. The combined effect of un-replicable study results and fraud significantly undermines the credibility of academic research.

Blogs like Data Colada represent an important and all too rare check on suspect research, but there is little professional incentive to do this kind of work. The threat of lawsuits likely makes others even less willing to report suspected research misconduct. Universities that employ academics accused of fraud should therefore pay whistleblowers’ legal fees for three reasons: They benefit from outsiders exposing their employees’ misconduct, they are responsible for their researchers’ behavior, and they should ensure others can afford to speak up by subsidizing this kind of public good.

First, universities have a stake in uncovering their professors’ research misconduct. Harvard’s reputation is tied to the quality of its research. In addition, fraud can expose universities to financial risk: Duke University recently paid the US government $112.5 million after a whistleblower alleged that falsified research had been used to attain federal grant funding. While it is difficult to quantify the financial value of Data Colada’s report, Harvard should at least ensure Data Colada does not pay to perform this valuable service.

Second, Harvard is responsible for the conduct of its researchers. As such, it should ensure that people who catch its employees doing their jobs poorly – and perform an important public service in the process – are not financially penalized as a result. Notably, this is also an argument that Harvard should cover Data Colada’s legal expenses rather than Nelson’s, Simmons’s, and Simonsohn’s own institutions.

Finally, the work done by blogs like Data Colada has always been important, but not lucrative; now, it may become extremely costly. Covering whistleblowers’ legal fees would set an important precedent. While the Data Colada GoFundMe campaign has exceeded its fundraising goal, crowdsourcing legal fees is not a sustainable long-term solution, nor is it one that is likely to reassure other academics considering making similar allegations. (It does, however, show how much the public values Data Colada’s work.)

Harvard has the opportunity to set an important precedent, fulfill its obligations to Data Colada, and perform a valuable public service by covering Data Colada’s legal fees. It should seize this opportunity before academia as a whole pays the price.

Leah L. Pierson is a sixth-year M.D.-Ph.D. student at Harvard Medical School and the Harvard T.H. Chan School of Public Health. Haley K. Sullivan is a fourth-year Ph.D. student at the Harvard Kenneth C. Griffin Graduate School of Arts and Sciences.