Back in the closet: Failure to count LGBTQIA+ individuals in STEM perpetuates underrepresentation

Editor’s note: The following series of stories highlights the social and academic landscape for LGBTQIA+ people in STEM, barriers to their success in STEM and potential solutions to these barriers to enhance LGBTQIA+ retention in STEM. This is part two. Read part one here. Read part three here.

I remember reading once that Alan Turing, the infamously gay pioneer of computer science, feared his homosexuality would be used to discredit his work. He outlined his logic in a letter to a friend, writing, “Turing believes machines think. Turing lies with men. Therefore, machines do not think. Yours in distress, Alan.”

This came only weeks before he was tried and convicted for gross indecency – being homosexual. After his conviction, Turing lost his job with the government of the United Kingdom and was forced to undergo chemical castration for his “crime”. He died from cyanide poisoning one year after his “treatment” ended, most likely by suicide, at only 41 years old.

Despite his fears, Turing’s ideas lived on; however, they often came without attribution. It wasn’t until the 1980s – more than 25 years after his death – that his cryptography work during World War II, development of the computer and ideas about artificial intelligence were rightfully acknowledged. Decades later, after petitions from scientists and mathematicians, Turing received a formal apology from the British Government in 2009 and a royal pardon from Queen Elizabeth II in 2013. In 2017, all men convicted for homosexual acts in England and Wales were pardoned by the Policing and Crime Act, known informally as the Alan Turing Law.

Given the growth of artificial intelligence today, his ideas are possibly more relevant than ever before. Furthermore, even today, his authentic representation of his sexuality, despite the consequences, is admirable.

The logic behind Turing’s fears, though ultimately somewhat flawed given his modern recognition, is present at the root of my own fears more than 65 years later. Will my sexuality negatively impact the way my science is viewed? Will it influence my career trajectory or interactions at work? Wouldn’t it be easier if I just hid my sexuality?

As “Turing’s children”, LGBTQIA+ individuals in STEM carry his legacy, facing many of the same challenges he did. These barriers can be summarized as under-sampling and under-representation.

Under-sampling
The first challenge for LGBTQIA+ scientists is counting, or rather, a lack thereof. As noted, there are very few studies about LGBTQIA+ individuals in STEM, perhaps because of the notion that LGBTQIA+ identities are viewed as demographic details rather than social identities and/or because the National Science Foundation (NSF) does not include measures of SO or GI in their research surveys.

Notably, at the urging of scientists and professional societies, the NSF began piloting SO/GI questions in 2021 for its National Survey of College Graduates, a survey focused on those with STEM bachelor’s degrees. Many suspected, and hoped, these questions would be formally integrated into the Survey of Doctorate Recipients and Survey of Earned Doctorates, which would provide vital information about Ph.D.-level LGBTQIA+ individuals in STEM.

Ultimately, these hopes were dashed when the NSF announced in January 2023 that it would only include questions on GI, excluding those on SO. This decision sparked frustrations for many LGBTQIA+ scientists, prompting two letters, one from 1700 scientists and one from Senators Tammy Baldwin (D-WI), Diane Feinstein (D-CA), and 16 others, calling for the NSF to adopt voluntary SO/GI questions on the surveys. The NSF has since announced a comprehensive pilot of 20 SO/GI questions for the 2023-2024 academic year, but the new survey received strong criticism on its methodology.

Since the NSF is often seen as the “gold standard” for STEM education research, professional organizations often follow their lead. For example, many organizations do not assess SO or GI, beyond a generic “male or female” dichotomy.

Commendably, some professional organizations, like the American College of Neuropsychopharmacology (ACNP), a group for psychiatrists, psychologists, psychopharmacologists and neuroscientists, have started to integrate SO/GI questions. ACNP introduced their SO question (“Do you identify as LGBTQ+?”) in 2020, but still only offered “male or female” options for GI. However, many other professional groups have not made any changes in their data collection.

Unfortunately, ACNP and many others took the common approach of combining all LGBTQIA+ individuals together under a single, yes-or-no question, rather than taking the opportunity to disaggregate and assess specific identities. This is important because while LGBTQIA+ individuals share many of the same experiences, there are also unique challenges for certain groups and intersectional effects.

Without sufficient data on SO/GI, the NSF and professional societies are unable to appropriately measure and address the numerous problems, which have been identified by the few studies available, that LGBTQIA+ individuals in STEM face.

"Turing’s story is both tragic and triumphant. It exemplifies the profound impacts LGBTQIA+ individuals have had in STEM, but also the ongoing challenges we face. It is my intent that bringing awareness to these problems will spark discussions, but ultimately these discussions must lead to actions and change."

-- Jordan Carter

Under-representation
An additional challenge for LGBTQIA+ professionals in STEM is a lack of representation. According to findings from the American Community Surveys and National Health Interview Surveys, men in homosexual relationships are more likely to have bachelor’s degrees than men in opposite sex relationships, but gay men are less likely to have a STEM degree or work in a STEM field. Interestingly, homosexual women have similar rates of STEM degrees and careers as heterosexual women. Unfortunately, there is not enough data on transgender individuals in STEM to conduct these comparisons – a perfect example of failures in data collection.

Lower rates of homosexual men, but not homosexual women, in STEM careers only tells half of the story though. While many individuals may be comfortable reporting their SO/GI in an anonymous survey, in one study, only 10% of faculty were “out” in their STEM department. In another study, 50% of LGBTQIA+ individuals were “out” in their STEM workplaces, 35% of which reported feeling like they needed to lie about or hide their identity.

The desire to conceal their identity is not surprising, given the negative career consequences, social isolation and explicit harassment faced by LGBTQIA+ scientists relative to their cisgender, heterosexual peers. These experiences have led 70% of those “out” at work to feel uncomfortable in their department.

This concealment directly contributes to the problem of lower LGBTQIA+ student retention in STEM because for many SO/GI minority students, working with faculty they identify with is a protective factor for retention and career advancement. However, faculty fearful of coming out in a hostile academic/STEM environment are hidden from students in need of these relatable mentors, while also silently reinforcing the notion that academia is unwelcoming for LGBTQIA+ individuals, fueling the same fears in the next generation of LGBTQIA+ scientists. This creates a self-sustaining cycle of de facto exclusion of LGBTQIA+ individuals from STEM research.

Further compounding the problem, resources reserved to support diverse trainees, like the Predoctoral Fellowships to Promote Diversity or Diversity Enhancements for other grants from the National Institutes of Health (NIH), do not include LGBTQIA+ individuals. Since many professional societies, like the Society for Neuroscience, use the NIH’s guidelines for their own diversity-related awards, LGBTQIA+ trainees are also excluded from this support. This impairs training and early career development opportunities for LGBTQIA+ scientists, perhaps contributing to attrition of LGBTQIA+ talent from academia.

Turing’s grandchildren
Turing’s story is both tragic and triumphant. It exemplifies the profound impacts LGBTQIA+ individuals have had in STEM, but also the ongoing challenges we face. As do the stories of Alan Hart, Ben Barres, Lynn Conway, Florence Nightingale, Bruce Voeller, Allan Cox, Sally Ride and many other LGBTQIA+ individuals, “out” or not, who made significant contributions to STEM.

It is my intent that bringing awareness to these problems will spark discussions, but ultimately these discussions must lead to actions and change. For this reason, in Part 3 I will discuss several actions at the individual and institutional levels that can be taken to begin to address some of these problems. With persistence, I know that STEM careers can be better for Turing’s “grandchildren” – the next generation of LGBTQIA+ scientists.