András Tilcsik – University of Toronto, Rotman School of Management
Michel Anteby – Boston University, Questrom School of Business
Carly R. Knight – Harvard University, Department of Sociology
Question 1. Drawing on theories of concealable stigma and stigma management, this article presents a comprehensive explanation of occupational segregation based on sexual orientation. It finds that gay men and lesbians concentrate in occupations that provide a high degree of task independence or require a high level of social perceptiveness, or both. What are the practical implications of this research? For instance, this article suggests that gay and lesbian workers might be severely underrepresented in occupations that utilize teamwork because they are low on task independence. Aside from reducing discrimination through legal actions (which takes time to take effect), how might we adjust work environments to ensure more equal representation?
What excites us most about our study is that it provides a comprehensive explanation for seemingly unrelated past observations of gay men and lesbians concentrations in select occupations. In short, we think we partly cracked the puzzle of these concentrations. Our theoretical explanation suggests that gay men might be overrepresented, for instance, in hairdressing and artistic occupations and that lesbians might concentrate in fire alarm installation and home appliance repairs not because these individuals are respectively “artistic” or “butch” but because these occupations offer high task independence and/or require a high level of social perceptiveness. Our theoretical explanation has several practical implications.
First, our study helps lesbians and gay men better understand the process of selecting or being selected into certain occupations. We often think of special talents, vocations, or even callings when trying to explain how people end up doing what they do. Our study suggests that what drives people into select occupations might not be as enigmatic as it seems. More specifically, we identify patterns to occupational segregation that speak to deep-rooted cultural dynamics (e.g., the stigmatization of sexual minorities), thus contextualizing individuals’ career “choices” or “pathways” with regard to larger trends. As illustrations, Ellen DeGeneres and Anderson Cooper might be particularly skilled at acting and news reporting, but the fact that these two occupations offer high task independence and require high levels of social perceptiveness should not be discounted. Lesbians and gay men might be over-represented in these professions because they build on their strengths (i.e., social perceptiveness) and buffer them from potential workplace harassment and other forms of discrimination (i.e., task independence). Once one notices these patterns, lesbian and gay professionals “choices” become clearer.
Second, our study also helps employers better understand how they might be typecasting lesbians and gay workers into certain lines of work. For instance, when hiring, many organizations might be looking for “team players” and therefore pause after reading the occupational background of a potential recruit. In everyday life, the term “team player” always carries positive connotations: e.g., “she is such a (good) team player and he gets along (so) well with his colleagues.” What this imagery misses is that not everyone might be “equally” eager to belong to a team. Some employees might—for very good reasons—be reluctant to interact with team members. Depending on how one was socialized, a team might prove very nurturing or extremely harassing. For a straight male worker, teams might evoke sports teams, bonding, and fun memories. For a lesbian worker, teams might evoke recess time on playgrounds, bullying, and skipped days at school. In that sense, recruiting only “team players” amounts to implicitly favoring the male straight applicant vs. the female lesbian one. A practical take-away from our study for employers is therefore to pause and think before reaching conclusions on what workers are or are not “good” at.
Finally, our study has practical implications for understanding, more broadly, the occupational segregation of any worker with a concealable stigma. We mention in the article that our theoretical lens might apply as well, for instance, to members of religious or political minorities. One could imagine extending the inquiry to the occupational segregation of Mormons or to light-skinned African Americans who can “pass” as white in the United States. Our theoretical lens would predict among these populations similar patterns of occupational segregation as the ones we noticed for lesbians and gay men. Ever since the publication of Mark Twain’s novel The Tragedy of Pudd’nhead Wilson (1894) describing the life of a black man passing as white, we have known of the unique circumstances faced by individuals with a concealable stigma. Our study shows more specifically how such individuals might find or be selected into lines of work that build on their strengths and mitigate potential backlash. In a strange way, our study provides faculty with an answer to the question of students with a concealable stigma asking what careers to embark on. The obvious answer is to tell them to embark on any journey they wish, yet a practical added answer might be to share this article with them.
Question 2. One of the underlying assumptions of the paper is that gay men and lesbians face on-going threats of discrimination and thus choose occupations that exhibit high levels of task independence to minimize this risk. Recent legislative changes suggest that minority sexual orientations are becoming more and more accepted in today’s society and hence gay and lesbian workers might feel comfortable in a wider variety of professions. Why do you think this was not reflected in the data from the National Longitudinal study of Adolescent Health? As de-stigmatization occurs in the US (and perhaps has already occurred in other nations), how do you think occupational patterns would be affected?
This is a great question. It is true that an implication of our study is that the de-stigmatization of minority sexual orientations might weaken the relationships we identify in our paper. And, by many accounts, recent trends in public opinion provide good reasons for optimism that de-stigmatization is rapidly occurring. When you look at the data, the shift towards the acceptance of gay men and lesbians has been astounding. This is particularly the case with regard to support of gay marriage. In the workplace, gay men and women have made remarkable strides and are continually entering occupations that were generally closed to them just a few decades ago. For instance, there are now seven openly gay members of Congress. And most research suggests the U.S. is not unique in this regard – worldwide trends point toward de-stigmatization. To the extent that this is the case, then, the key driver of the patterns we identify – stigma management – might fade over time.
Despite all this, we think one should be cautious in heralding the imminent demise of occupational segregation for gay men and lesbians. Part of the reason for this is that lots research we have on occupational matching points to mechanisms that make current demographic distributions sticky. For instance, even if it stigma in the workplace disappeared overnight, we would still expect a lag in the occupational landscape. People make educational and career decisions early in life that will have reverberations for occupational segregation for decades to come. Moreover, we have a lot of research that points to the degree to which networks and cultural-matching mechanisms serve to funnel similar people into similar jobs. These mechanisms serve to maintain some of the segregation patterns we document, even when stigmatization becomes weaker.
There are also reasons to be less sanguine about occupational segregation ending any time soon. Your question points to legislative changes that might lessen stigmatization. Some of these legislative initiatives, such as nondiscrimination laws, are still deeply contentious in many parts of the country. The majority of U.S. states have no explicit protections against employers discriminating against gays and lesbians, and the controversy earlier this year over so-called “conscience protection” bills should show that passing and enforcing anti-discrimination law is still an uphill legislative battle in many places. This suggests that, to the extent that de-stigmatization affects occupational segregation, this might crucially depend on geographic context. Even if there is a global trend towards de-stigmization, you could end up in a situation where occupational segregation continues in those places where worker protections are weak or where stigmatization remains high.
In fact, we have recently conducted another study that indicates that individuals who have a minority sexual orientation and live in non-urban areas report a greater tendency to monitor their self-presentation in front of others than sexual minorities who live in urban areas. This tendency is closely related to the social perceptiveness mechanism we posit in our ASQ article and suggest that these mechanisms vary with geography.
Finally, it is worth considering how occupational segregation would change, even if the processes we identify with regard to stigma were to disappear completely. As we mention in the paper, apart from our own theory, the occupational segregation of gay men and lesbians is characterized by a strong sex-typing tendency. That is, lesbians tend to enter male-dominated occupations and gay men tend to enter female-dominated occupations. While it’s not clear what drives this sex-typing tendency, it could be the case that occupational segregation based on this tendency would persist so long as gendered occupations persist, even if stigmatization decreases.
For all these reasons, we weren’t very surprised when a recent National Longitudinal study of Adolescent Health data, which focuses on younger individuals, revealed a similar set of patterns as our national-level Census data. Even with de-stigmatization, we anticipate that the patterns we identify here will be slow to unravel. Nevertheless, it would be extremely interesting to test what you propose. One could, for instance, continually update our findings as new American Community Survey data come in.
Question 3. We find it interesting that workers that experience stigma based on sexual orientation do not necessarily require an occupation to have both a high degree of task independence and a high level of social perceptiveness. Specifically, we are intrigued by the higher presence of stigmatized workers in occupations that feature above average social perceptiveness, but do not feature high task independence (e.g. flight attendants and training and development specialists and managers). Tasks in these occupations likely include regular work interactions with coworkers that also possess above average social perceptiveness—the implication being that the threat of discovery of concealable stigma could actually be enhanced in these settings. Do you have any ideas about what might enable gay and lesbian workers to do well in these settings despite this potentially higher threat of exposure?
Very interesting question. It’s worth noting that the occupations you mention are not grossly under the average in terms of task independence. So perhaps, it might be that what is required is a high degree of social perceptiveness coupled with at least some level of independence. As you suspect, it might be very difficult to be in an occupation that requires a high level of social perceptiveness but provides very little task independence. It’s an empirical question, and we haven’t yet explored this in depth. It is also possible that some of the occupations that require social perceptiveness have historically attracted gay and lesbian workers and have become more accepting, which could mitigate the risk of the potentially higher threat of exposure you mention.
Indeed, as we emphasize in the paper, while a high degree of social perceptiveness and task independence can help explain why an occupation becomes segregated in the first place, this pattern may actually be maintained not only by continued sorting on the basis of task independence but also by second-order mechanisms (e.g., processes related to the reputation of the occupation for tolerance or network-based mechanisms of employee referrals and information sharing). In other words, if task independence and social perceptiveness leads to a substantial gay or lesbian presence in the first place, that itself can draw additional gay and lesbian workers to the occupation in subsequent periods.
Question 4. What question did we miss? Please ask yourselves a good question, and answer it.
How did the paper evolve over time?
Our editor and reviewers gave us very consistent and thoughtful feedback that shaped the paper in three main ways. First, we worked hard to make it clearer what social perceptiveness and task independence really mean and how they are different from other, seemingly related concepts. Second, the final version was much more explicit about the potential mechanisms at work, including the more nuanced second-order mechanisms we mentioned earlier. Third, we had initially framed the paper in very broad “grand” terms, and the reviewers helped us find the right level of abstraction while also acknowledging the limits of our data.
Regarding this last point, we think there is a lot more to do in this area of research. One obvious avenue forward would be to collect data to look directly at finer mechanisms of application segregation (rather than just the broad outcomes in terms of occupational segregation). Roxana Barbulescu and Matt Bidwell have a very nice paper in Organization Science (2013, “Do women choose different jobs from men? Mechanisms of application segregation in the market for managerial workers”), which has done just that for gender segregation, so there are some helpful research models one could adopt to delve even deeper into the relationship between sexual orientation and occupational segregation.