Policy ForumDiversity and Inclusion

Leadership to change a culture of sexual harassment

See allHide authors and affiliations

Science  27 Mar 2020:
Vol. 367, Issue 6485, pp. 1430-1431
DOI: 10.1126/science.abb5791

When we look at surveys of the scientific community, we see an overall trend: The presence of women, minorities, and other groups that have been historically subject to harassment tapers off in later career stages. This occurs despite those groups' strong interest and proficiency in science and engineering. As a primary public funding resource for science and engineering research and training, the U.S. National Science Foundation (NSF) has a responsibility to address such deficiencies. Shaping the research landscape is a strategic process that normally requires planning on the order of decades, but this is an issue that requires more immediate leadership and action. How, then, does an agency like NSF—which has considerable influence but limited direct authority—work with the community and other institutions to implement change on issues that cannot wait? The case of NSF's work to combat harassment in the science community, a persistent problem for decades that remains shockingly widespread, is illustrative.

A Challenge for Leadership

As a university administrator and president, I encountered the issue of sexual harassment in science, but efforts to address specific cases were overshadowed by pervasive roadblocks to changing a culture that enabled harassment. After my appointment as director of NSF in 2014, I took the opportunity to confront this issue when then–NASA Administrator Charlie Bolden and I simultaneously posted letters to university presidents emphasizing that our agencies had no tolerance for harassment. We reminded universities that they could lose grant funding if they were not compliant with federal Title IX regulations against sex-based discrimination.

In late 2016, a National Academies of Science, Engineering, and Medicine (NASEM) committee convened to begin studying the prevalence, nature, and impact of sexual harassment in science, engineering, and medicine. NSF funded this effort because we needed an accurate accounting of the current scope and scale of the problem. As the NASEM committee worked through 2017, the #MeToo movement brought new light to the problem of sexual harassment. The sustained outcry created a moment where action on NSF's part could draw a greater level of public support than was previously possible. The detailed insights from the NASEM report would be an important factor in how we proceeded once it was released, but we did not need to wait for the committee to complete their work to begin our own. Ample existing data and anecdotal evidence show that harassment deters talent, destroys careers, and impedes the science and engineering enterprise. I decided that NSF needed to act quickly to find a way to give the agency a more operative role in addressing harassment, and that we would not be delayed by calls for more data.

Embedded Image

How would our work this time be different, to achieve meaningful change on an issue that had defied correction for decades? To begin, we removed any uncertainty about our commitment to action and progress. I made it clear to NSF's leaders that this was a top priority for our entire organization and that staff were empowered to engage stakeholders, develop ideas, refine proposals, and make progress toward implementation of an antiharassment policy.

Then, we needed to create a robust mechanism to combat harassment. A campaign to simply reiterate our existing policy would be insufficient. We needed a solution that could be integrated into NSF's legal and policy framework without running afoul of other legal structures (including Title IX). Attempting to apply some new regulatory policy could create legal and technical challenges and would take considerable time. Instead, we identified our statutory grantmaking authorities as the optimal solution. Adding a new term and condition (TC) directed specifically at harassment to our Proposal and Award Policies and Procedures Guide would place it alongside numerous other requirements that nearly 2000 institutions already agree to every year when they accept NSF funding.

Developing a Course for Action

The resulting harassment TC draws a connection between researcher behavior and NSF funding. It creates accountability and is auditable. It requires an NSF-supported institution to notify the agency within 10 days of taking any administrative action or making a final determination against a principal investigator (PI) or co-PI related to harassment, including sexual harassment or assault. After consulting with the awardee institution, NSF will determine whether it is taking appropriate action, such as requesting NSF approval to remove a PI and appoint a substitute. This determination would be based on factors that include the need to safeguard students and other researchers. If the institution's work is satisfactory, no further action from NSF is called for. If not, NSF may take action, including reducing or suspending the award. Notifications come to NSF through a secure, electronic system, and the information is sequestered within our Office of Diversity and Inclusion, separate from our other data systems and provided only to staff with a specified need to know.

Describing the TC is straightforward, but the process of crafting it required careful thought about how to create a unified requirement that accounts for gray areas that might arise due to the interplay of existing policies and procedures at institutions we support. Through the TC, NSF sets an expectation for accountability but provides flexibility. Like NSF's other award requirements, institutions have some discretion in how they approach the TC. Disclosure to NSF is mandatory, but our hope is that we can approach this issue with the institutions as our partners and that the actions they take in harassment cases will help lead the way.

if there's a defining characteristic of how NSF operates, it's that our policies are shaped with research community input. We are cognizant of the need to avoid unnecessary burdens for proposers and awardees. After NSF leadership developed an initial proposal for the TC, we informed grantee organizations (1), then issued a call for public comment (2). The public comments were a tremendous help in shaping the final TC. Two examples include changing the reporting time frame from 7 days to 10 and requiring subawardee institutions to report determinations or administrative actions directly to NSF, rather than making the primary award recipient institution responsible.

Some comments raised concerns that institutions might avoid taking administrative action on harassment to get around NSF's reporting requirement. In response, NSF has developed an electronic, secure, anonymous harassment reporting system that can receive complaints directly from individuals. Although that system is new, individuals previously could report sexual harassment to NSF. That commenters were unaware of this indicates that we must communicate better.

NSF began outreach with stakeholder groups to provide assurances on some key issues of concern, such as confidentiality. NSF worked to clearly state what the TC is intended to do (ensure safety and security in NSF-supported projects and that proposed research can be carried out) as well as what it isn't (serve as a second investigation or remove funding that serves many because of the actions of a few).

NSF focused on feedback that helped improve the potential for successful implementation of the TC, not on comments predicated on harassment being overly complex as a basis for recommending inaction or delayed action. We devised what is, in our judgement, the best possible way for NSF to take immediate action when necessary. However, this process is not finished. We are still gathering data to help us refine the TC. We recognize that there are ongoing concerns from some parts of the community, some of which will likely result in modifications to our processes or be addressed through improved communication.

Some officials from the university community have questioned whether reporting is appropriate before their institutional processes to determine guilt or innocence are complete, including all appeals, which often take months if not years. NSF's concern at that stage is not the guilt or innocence of an individual—it's whether publicly funded research can continue under the conditions of the existing award agreement in a manner that ensures the safety of all personnel working on the award. We have heard concerns from those questioning how this TC interacts with Title IX and state laws, and the answer is that it doesn't—institutions are still expected to conform to all applicable laws and codes. And from some focused on specific cases, we have heard calls for NSF to take harsher or more dramatic action. For them, we are committed to explaining how and why we act in these cases and demonstrating that we do so consistently.

Ripple Effects Beyond NSF

Since the implementation of the TC in October 2018, NSF has received 24 harassment action notifications from institutions. The full implications cannot be expressed in a single number; one of our ongoing efforts is to refine metrics that better measure and express progress under the TC. But the impact has gone far beyond those particular cases. NSF has been proactive in our outreach to groups considering action of their own. We've already seen our actions ripple through the community as departments, universities, and scientific associations have set in motion plans to confront harassment in research environments. We have sought to amplify that work, identifying and sharing examples of codes of conduct developed by field sites, facilities, and other infrastructure we support. NSF seeks to build relationships that we can use to address other persistent cultural issues.

When one federal agency works on a culture-change issue, it creates opportunities for others to respond within their own domains. NSF has been in frequent communication with members of Congress and our partner agencies about exploring other possible ways to confront harassment. Shortly after we announced the TC, the National Institutes of Health expressed its commitment to new solutions (3). In September 2019, the White House Office of Science and Technology Policy established the Joint Committee on the Research Environment (JCORE) (4), with safe and inclusive research as one of its top priorities. JCORE's summary of objectives noted that harassment is “[W]here institutional leadership is extremely important” and called for an approach that addresses “cultural dimensions.” Centralized federal leadership on this issue has the potential to spur change at an even greater scale.

NSF leaders have also been meeting with foreign counterparts, exchanging information and discussing how to effect change in legal and ethical frameworks that are often quite different from those of the United States. Camila García, Chile's Science, Innovation, and Technology attaché, indicated that NSF input helped jump-start the process that led to the Chilean government passing a new law to combat harassment in August 2019 (5). NSF has engaged with the European Commission, and harassment will be the topic discussed by the Global Research Council's (GRC) Gender Working Group at the 2020 GRC annual meeting, an area of focus NSF suggested.

Lasting change will require sustained involvement, both in terms of maintaining and refining the TC, as well as in helping maximize its benefits. NSF is continuing to assess how the TC is being implemented, identifying potential improvements and ways to streamline processes that have been found to be inefficient.

Holding Ourselves Accountable

Culture change, within and especially across institutions, is never easy. When considering roadblocks, complications, and concerns from different sectors, the easiest option is usually inaction—and some will argue it's the most prudent as well. But we know that harassment is destructive for women and minorities, and for the research community writ large. For those who love science, research is where we can see humanity at its best, and it can be crushing to acknowledge that we often see the worst as well. We have a duty to live up to our ideals and make sure that science and engineering present real opportunities for everyone. We can reach that point by holding ourselves accountable as leaders, committing to taking as much action as quickly as we can within our ability to act. I hope that this serves as an example that others follow to inspire not just change, but change through leadership.

References and Notes

Acknowledgments: C. Lohse and R. Margetta contributed to the researching, drafting, and editing of this article. R. Davis, P. Hoyle, B. Cosgrove, J. Feldman, and A. Lupia contributed to review.
View Abstract

Stay Connected to Science

Navigate This Article