Headquarters of the National Science Foundation in Alexandria, Virginia

National Science Foundation

New NSF rules on sexual harassment leave many questions unanswered

The National Science Foundation (NSF) in Alexandria, Virginia, hopes that its new policy on sexual harassment will spur universities to deal more aggressively with the pervasive problem. But the additional reporting requirements, which will be officially published Monday in the Federal Register, are far from a definitive statement about how NSF plans to deal with this complex and sensitive subject.

The carefully worded notice, for example, doesn’t address whether a scientist found guilty of sexual harassment should automatically be removed from a grant. And it would not require universities to tell NSF when they launch investigations into allegations of harassment.

The eight-page notice in the Federal Register is designed to flesh out and seek public comment on an “important notice” that NSF issued on 8 February. It proposes adding two new components to the “terms and conditions” that universities and other institutions agree to follow when they accept an NSF award. (Grants are awarded to institutions, not individuals, although scientists invariably refer to “my grant.”)

One new section requires the institution to tell the agency if it has determined that the principal investigator (PI) or co-PI of an NSF grant has committed sexual harassment. The other says that an institution must inform NSF if it puts a PI or co-PI on administrative leave in response to an allegation of harassment, even if an investigation is not complete.

The proposal highlights the legal complexity—and institutional sensitivity—of the issue. Federal funding agencies have traditionally addressed sexual harassment in academia under the gender equity provisions—known as Title IX—of a 1972 education law. That law gives the federal government the ability to withhold funds and impose other sanctions on institutions found in violation of its terms.

Historically, NSF has not investigated specific allegations of sexual harassment made by individuals. Instead, it has referred harassment cases to the Department of Education. (In contrast, NSF investigates allegations of scientific misconduct through its Office of Inspector General. But harassment is not part of the federal definition of misconduct, although the American Geophysical Union has recently expanded its definition to include harassment.)

Even so, NSF has some leverage over universities. It requires all institutions receiving NSF funding to certify that they are complying with Title IX. NSF also conducts so-called compliance reviews of a subset of institutions to make sure their gender equity practices are up to snuff, and it can target an institution for further scrutiny if it sees evidence of widespread gender discrimination.

Confusion and uncertainty

Last month, NSF Director France Córdova admitted that the agency sometimes learns about sexual harassment charges involving NSF-funded researchers through the media, saying “that’s a pretty poor way to find out.” And a recent case involving an NSF-funded researcher at Boston University (BU) highlighted the uncertainties surrounding its current approach to handling harassment issues.

Last November, BU found geologist David Marchant had sexually harassed a former graduate student during a field expedition in Antarctica from 1999 to 2000, and placed him on paid administrative leave while he appealed the university’s decision. But confusion immediately arose at both the university and NSF around the question of whether Marchant could remain as the PI for an NSF grant, according to emails made public last week by the science committee of the U.S. House of Representatives, which has been examining the case.

[W]e know there’s still a lot of heavy work to pull off in the arena of sexual harassment.

Rhonda Davis, National Science Foundation

In early December, according to the emails, BU officials asked NSF “to confirm” its view that Marchant couldn’t continue as PI while he was on leave. The university said it wanted to retain the grant, with a different PI, to support the work of a graduate student. On 18 December 2017, an NSF program officer wrote back that “there is no NSF policy that supports [your] statement” regarding Marchant’s removal, adding that NSF’s grants and contracts division believed “this is an internal BU issue.”

Five weeks later, BU informed NSF that it had “decided to keep [Marchant] as the PI pending further developments.” BU’s email noted that Marchant “is providing mentorship to a graduate student on the project” and that the university believed “no formal notification to NSF is required at this time.”

A few days later, however, the NSF program officer reversed course. Citing “clarification from NSF upper management,” the officer wrote that BU is “required to appoint a substitute PI … without delay.” BU immediately chose another faculty member in the department, and NSF approved the change. (Last week, BU denied Marchant’s appeal to keep his job.)

“We want to know”

NSF officials say the new reporting requirements are intended to eliminate such miscommunication, as well as help NSF better monitor its grantees. “We want to know about it” if a university has placed someone on administrative leave because of a matter involving alleged sexual harassment, says Rhonda Davis, head of NSF’s Office of Diversity and Inclusion. “We don’t care if it’s a finding of sexual harassment or if they are still carrying out the investigation.” In either case, she notes, the development “changes the terms and conditions of the grant,” so NSF should be notified.

The new rules only address when NSF should be notified, however, not the broader issue of how instances of sexual harassment should be handled by either the university or NSF. For example, Davis says, a university is not required to inform NSF at the time it opens a sexual harassment investigation of a PI on an NSF grant. Nor must it remove an NSF-funded PI if that person is put on administrative leave.

One reason for the ambiguity, Davis says, is the wide variation in the meaning of “administrative leave.” At some institutions, administrative leave means that a faculty member is banned from campus. That step, in all likelihood, would make it impossible for them to carry out their NSF-funded research project, requiring the installation of a new PI. But at another university, Davis says, a faculty member on administrative leave might retain access to research facilities, or even be allowed to interact with students under appropriate supervision. In those cases, she explains, their altered status might not affect their ability to carry out the research.

Another significant unknown is exactly how NSF will respond to a grantee who is found to have committed sexual harassment. Davis demurred when asked whether the new policy means a harasser should automatically be removed as a PI. “Our response would be based on what the person has done,” she replied. “Sexual harassment is not OK. But there have been cases where an institution took action commensurate with what the person did—although we’d prefer that it hadn’t happened at all—and the PI experienced some serious consequences that were sufficient to change their behavior.”

At the same time, Davis says the new rules reinforce NSF’s duty to be a responsible steward of tax dollars. “It makes clear that universities don’t make the call [on whether a PI can remain on a grant], we make the call,” she says.

Not the last word

Davis emphasizes that the new notice is far from NSF’s last word on sexual harassment. To date, she says, an internal NSF task force has focused “on things that we felt we could act on very quickly, and avoided things that require a heavier lift. Now we’re ready to move to the second and third phases,” which could include making PIs found guilty of sexual harassment subject to some of NSF’s most severe penalties, including so-called debarment, which bans researchers from seeking federal grants for some period of time. “That’s a serious conversation,” Davis says, “and we know there’s still a lot of heavy work to pull off in the arena of sexual harassment.”

The new policy is part of a series of steps NSF is taking to stay abreast of developments relating to sexual harassment in academia. For example, Davis says a new website provides “one-stop shopping” for those seeking information. The site is still “a work in progress,” she notes, and there are plans to install a secure mechanism for individuals to file a complaint.

Her office also now has someone with extensive experience investigating Title IX complaints, she says. And it has the resources to query universities about individual cases of harassment that NSF learns about from the media or from an informal communication. “We want to do more than simply check the boxes” when such cases come up, she says, describing plans to use such interventions as a “mini–compliance review.”

Other voices will soon be weighing in on how federal agencies should address sexual harassment in academia. A study by a committee of the National Academies of Sciences, Engineering, and Medicine—funded in part by NSF—is expected out this summer. And the chairman and top Democrat of the House science committee have asked the Governmental Accountability Office (GAO) to examine the ability of five major federal research agencies, including NSF, to ensure that their grantees comply with Title IX. The legislators also want to know how the agencies deal with alleged sexual harassment by grantees in situations not covered by Title IX. The GAO study is expected to get underway this summer.

In the meantime, NSF will be accepting comments on its proposed reporting requirements through early May.