NSF Still Won’t Track Sexual Orientation Among Scientific Workforce, Prompting Frustration

The U.S. National Science Foundation (NSF) says it does not plan to include a question about sexual orientation in a major national workforce survey, prompting hundreds of researchers to send a letter of protest.

Last month, the agency submitted its plans for the 2023 National Survey of College Graduates (NSCG), a biennial survey of more than 160,000 U.S. bachelor’s degree holders with a focus on the science and engineering workforce, to the Office of Management and Budget (OMB) for approval. Many LGBTQ scientists were pleased that the survey will, for the first time, include a question about gender identity for all respondents. But the absence of a sexual orientation question is “incredibly disappointing,” says Ramón Barthelemy, an assistant professor at the University of Utah who has studied the experiences of LGBTQ scientists in physics. Speaking as a gay man, he says, “We have fought so hard for so long to try to get representation in the scientific community, and what NSF is communicating to us is, they don’t want us to have that representation.”

The agency had raised hopes in 2021 when it pilot tested questions about gender identity and sexual orientation on the NSCG. Advocates took that as a sign that in coming years the agency would fully implement those questions in its suite of workforce surveys, including the Survey of Earned Doctorates and the Survey of Doctorate Recipients, which focus on Ph.D.-educated scientists and engineers. Those surveys provide crucial data about which demographic groups are underrepresented in the U.S. STEM community.

Up to now, respondents have had the option of identifying their sex only as “male” or “female.” If NSF’s plans are approved, this year’s NSCG will split that question into two, first asking respondents to specify their sex assigned at birth and then their current gender identity. For the latter, respondents will have the option of selecting all that apply among male, female, and transgender; typing their own descriptor as free text; or both. This approach is consistent with OMB recommendations released this month detailing how to collect gender identity data, though some think the survey should offer more options, such as nonbinary and gender-fluid. Gender is “quite an expansive spectrum,” says Abby Ray, a microbiology Ph.D. student at the University of California, Davis, and vice president of communications for oSTEM, an organization that represents LGBTQ+ people in STEM. For oSTEM’s own data collection efforts, “We try to give as many options as possible so people have a chance to see themselves represented,” they say.

The addition of a gender identity question is a “small step in the right direction,” says Kaitlin Rasmussen, a postdoc in astrophysics at the University of Washington, Seattle, who has written on collecting data about marginalized genders. But it’s frustrating that sexual minorities won’t be counted as well, they add. “Gender and sexuality are really deeply intertwined and you might say you can’t consider one without the other.”

NSF told OMB it decided to omit the question about sexual orientation because it didn’t perform well on the pilot test, taking longer to complete than the gender identity question and resulting in more changed answers and respondents exiting the survey. But the agency provided no data to back up those assertions or show how the results compare with those for questions about, for instance, race, disability, or income. “If they really want to stick to what they’re claiming, then they should release the data,” says Jon Freeman, an associate professor of psychology at Columbia University who has spent the past 5 years advocating that NSF collect data on LGBTQ scientists. An NSF spokesperson wrote in a statement to Science that the agency “will make those and other methodological findings available to the public later this year.”

It’s “perplexing,” says Nancy Bates, a retired federal statistician. She points out that questions about sexual orientation have been asked in other federal surveys for years and have generally performed well. “There’s less debate and sort of worry about measurement error for that particular question, as opposed to gender identity,” says Bates, who co-chaired a National Academies of Sciences, Engineering, and Medicine working group that issued recommendations in 2022 detailing best practices for measuring sexual orientation and gender identity.

The NSF spokesperson said the agency plans to continue to explore whether to add a sexual orientation question to future workforce surveys. “The findings [from the 2021 pilot test] are already helping inform questions for future surveys. However, the results indicated that additional research and testing are needed to ensure it results in accurate data that protects privacy,” the spokesperson added.

Freeman hopes OMB will force NSF’s hand. In an effort to sway OMB’s decision, due next month, he drafted a 16-page letter detailing his concerns with NSF’s plans, which has been cosigned by 1700 other researchers. The letter, submitted to OMB today, argues that NSF hasn’t been transparent about its decision-making process and that the agency could still move forward with a sexual orientation question by borrowing methodological approaches that have already been extensively tested and vetted by other federal agencies, such as the Department of Education, the Census Bureau, and the Centers for Disease Control and Prevention.

“[OMB] could say, there’s no reason that you’re not also doing sexual orientation when that’s what all these other agencies have done,” Freeman says.