Sexual orientation stereotypes and job candidate screening: why gay is (mostly) OK

Sterkens, Philippe et al. Article — Published Version Sexual orientation stereotypes and job candidate screening: why gay is (mostly) OK Journal of Population Economics Provided in Cooperation with: Springer Nature Suggested Citation: Sterkens, Philippe et al. (2025) : Sexual orientation stereotypes and job candidate screening: why gay is (mostly) OK, Journal of Population Economics, ISSN 1432-1475, Springer, Berlin, Heidelberg, Vol. 38, Iss. 1, https://doi.org/10.1007/s00148-025-01071-w This Version is available at: https://hdl.handle.net/10419/318426 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) Journal of Population Economics (2025) 38:16 https://doi.org/10.1007/s00148-025-01071-w ORIGINAL PAPER Sexual orientation stereotypes andjob candidate screening: why gay is(mostly) OK PhilippeSterkens1 · AxanaDalle1 · JoeyWuyts1 · InesPauwels1· HellenDurinck1· StijnBaert1,2,3,4,5 Received: 4 May 2023 / Accepted: 2 January 2025 © The Author(s) 2025 Abstract To explain the conflicting findings on hiring discrimination against applicants in a same-sex marriage, we explore the perceptual drivers behind employers’ evaluations. Therefore, we conduct a vignette experiment among recruiters, for which we test systematically selected stereotypes from earlier studies. We find causal evidence for distinct effects of same-sex marriage on candidate perceptions and interview probabilities. In particular, interview probabilities are positively (negatively) associated with the stereotype of women (men) married to a same-sex partner as being more (less) pleasant to work with compared to candidates in a different-sex marriage. In addition, interview chances are negatively associated with the stereotype of candidates in a same-sex marriage as being more outspoken. Furthermore, our data align well with the idea of a concentrated discrimination account, whereby a minority of employers who hold negative attitudes towards individuals in same-sex marriages are responsible for most instances of hiring discrimination. Keywords Sexual orientation· Signalling theory· Statistical discrimination· Tastebased discrimination· Hiring experiment JEL Classification C38 Classification Methods, Cluster analysis, Principal components and factor models· J12 Marriage, Marital dissolution, Family structure and domestic abuse· J71 Discrimination 1 Introduction Since the first empirical investigation of sexual orientation–based discrimination in economics (Badgett 1995), the socioeconomic outcomes of same-sex couples have, in general, improved globally (Badgett etal.2021; Drydakis 2021; OECD 2020). Responsible editor: Alfonso Flores-Lagunes Extended author information available on the last page of the article P.Sterkens et al. 16 Page 2 of 40 The scientific investigation of the labour market success of non-heterosexual individuals has branched into both a supply-side (Burn and Martell 2020) and a demandside research tradition (Burn 2018, 2020). Given that differences in labour market outcomes between sexual majority and minority individuals appear nowadays to be mainly driven by the demand side of the labour market (Fric 2017), it is necessary to further investigate labour market discrimination. As evidenced in earlier research, predominantly centred around gay men and lesbian women, sexual minorities are susceptible to hiring discrimination already in the earliest stages of the recruitment process. Indeed, both field (Hebl etal.2002; Drydakis 2009, 2015; Tilcsik 2011; Hammarstedt etal.2015; Jepsen and Jepsen 2015; Laurent and Mihoubi 2017; Patacchini etal. 2015) and laboratory studies (Horvath and Ryan 2003; Singletary and Hebl 2009; Everly etal 2016) have found evidence of such discriminatory treatment in the application process. For example, the first large-scale correspondence experiment on discrimination against openly gay men across the United States indicated that gay men were 40% less likely to be offered a job interview compared to their heterosexual counterpart (Tilcsik 2011). A laboratory study among 236 predominately White participants in the United States demonstrated a more nuanced image as gay men and lesbian women were evaluated more negatively than heterosexual men, but more positively than heterosexual women (Horvath and Ryan 2003). However, a substantial number of studies, again in both field (Bailey et al. 2013; Baert 2014; Patacchini et al. 2015; Acquisti and Fong 2020) and lab settings (Nadler and Kufahl 2014; Niedlich and Steffens 2015; Baert 2018a; Niedlich etal.2022), have found no such evidence. In general, two metaanalysis on this mixed evidence from field experiments in OECD countries conclude that gay and lesbian job candidates receive on average 35 to 40 percent fewer positive reactions than heterosexual candidates (Flage 2020; Lippens etal.2023). However, it must be noted that this unequal treatment varies between multiple contextual factors such as the occupational requirements, the gender of the applicant, and the country of employment. More concretely, the level of discrimination seems to be higher for gay men compared to lesbian women, in low-skilled (versus high-skilled) occupations, and in European countries compared to the United States. In the study of sexual minorities, bisexual individuals are currently underrepresented because it is unclear whether they are consistently perceived or ‘read’ as sexual minorities in the labour market, especially when they have different-sex partners (Badgett etal.2024). Nevertheless, a vignette experiment by Arena and colleagues (2017) shows that employers favour gay and lesbian individuals over bisexual ones. Similarly, Parnell and colleagues (2012) find that bisexual workers systematically report more career barriers than their gay and lesbian counterparts. To tackle potential hiring discrimination, it is crucial to develop insights into the stereotypes about candidates in same-sex relationships that drive potential differences in their hiring probabilities compared to heterosexual candidates (Fric 2017). The empirical investigation of these driving stereotypes is challenging and necessitates an in-depth analysis for three reasons. First, our own review of the multidisciplinary peer-reviewed literature (Appendix Table1) yielded no less than 70 characteristics associated with homosexuality—a sexual orientation featuring same-sex partners that has been studied thoroughly. The sheer number of such characteristics Sexual orientation stereotypes andjob candidate screening:… Page 3 of 40 16 is astonishing and raises questions regarding (1) the attributed relevance of each characteristic in contemporary hiring processes and (2) the complex detailed image (some) employers might have of candidates in same-sex marriages. Second, thus far we have mostly discussed individuals in same-sex marriages as one homogenous group. However, treating diverse groups of sexual minorities as one homogenous group could lead to incorrect conclusions about their labour market outcomes (Mize 2016). For instance, researchers have found heterogeneity in both the hiring chances and attributes associated with gay men and lesbian women. Different patterns of perceptions could be in play when explaining differences in the hiring probabilities of individuals in same-sex marriages and, consequently, a ‘one size fits all format’ of policy-making might be undesirable when supporting sexual minorities. Third, in addition to the large number of perceptions and the potential differences between men and women in same-sex relationships, the perception puzzle is further complicated by the mixed valence of perceptions. In fact, both positive and negative traits have been associated with homosexuality (see Appendix Table1). The latter may suggest that candidates’ hiring probabilities might be the result of an interplay between both positive and negative candidate perceptions associated with sexual orientations. Given the current state of the literature, instead of simply providing yet another data point on the instances of hiring discrimination against sexual minorities, research calls for a deeper understanding of the phenomenon—and its mixed findings (Neumark 2018). Through a vignette experiment among real-life recruiters, equally distributed between the United Kingdom and the United States, we contribute to the development of such an understanding by answering four research questions. First, we reply to the question ‘What are the average treatment effects of samesex marriage for (1) men and (2) women on their hiring probability?’ Second and most importantly, we question ‘What are the average treatment effects of a samesex marriages for (1) men and (2) women on candidate perceptions?’ and ‘What are the indirect associations between same-sex relations and hiring probability via candidate perceptions?’ to gain insights in the driving stereotypes. Fourth and final, we answer the question ‘For which (i) candidate, (ii) job, and (iii) recruiter are the effects of a same-sex marriages on hiring probability heterogeneous?’. By doing this, we contribute to the literature in three ways. First and foremost, we conduct an empirical and causal test on the stereotyping of men and women in same-sex marriages within one framework. Moreover, as a prerequisite to accomplishing this, we additionally contribute to the literature by reviewing and structuring the literature on potential stereotypes. As a second broader contribution, we go beyond a traditional investigation of moderators, i.e. variables that affect the strength or the direction of the relationship between an independent variable and a dependent variable (Hayes 2017). In addition to testing the candidate, job, and recruiter-side variables that strengthen or attenuate discrimination, we explore the idea of a concentrated discrimination account, whereby a minority of employers, who privately hold negative attitudes towards sexual minorities, are responsible for most instances of hiring discrimination. A third and final contribution lies in the data collection that took place in two similar yet different Anglo-Saxon countries (i.e. the United Kingdom and the United States) which allows us to explore possible culturally sensitive P.Sterkens et al. 16 Page 4 of 40 differences in the hireability of non-heterosexual candidates. For example, Adamcyzk and Pitt (2009) discovered cross-national variation in the public opinion about homosexuality which could explain stronger penalties in some countries. 2 Theory As described in the previous section, mixed results were found in the literature regarding the hiring chances of non-heterosexual candidates. However, as two metaanalyses indicate an overall negative impact of none-heterosexual orientations on the candidate’s hiring probability (Flage 2020; Lippens etal.2023), we hypothesise that for both men and women, same-sex marriages will negatively affect the candidate’s hiring chances. H1: Same-sex marriages have a negative effect on hiring probabilities for both men and women. Next, from a theoretical perspective, the two seminal (economic) theories of taste-based (Becker 1957) and levels-based statistical discrimination (Phelps 1972; Arrow 1973) can explain hiring discrimination against individuals in a non-heterosexual marriage.1 First, according to the theory of taste-based discrimination, employers may be prejudiced against sexual minorities and might, therefore, expect disutility in collaboration with such candidates themselves or from colleagues and clients (Bodvarsson and Partridge 2001). Consequently, prejudiced employers are, to some degree, willing to make sacrifices to avoid collaboration with employees. Specifically, they would rather hire a less competent heterosexual candidate than a skilled candidate in a same-sex marriage. Based on the theory of taste-based discrimination, we hypothesise that sexual orientations and hiring probabilities are negatively and indirectly associated via employers’ prejudices regarding collaborations between sexual minority candidates and themselves, colleagues, and/or clients. By doing so, we test candidate perceptions as potential mediators (Hayes 2017): variables that explain the mechanism through which independent variables (i.e. samesex marriage) influences a dependent variable (i.e. hiring probabilities). H2A: Same-sex marriages have a negative effect on the employer’s prejudices regarding collaborations. 1 Note that multiple biases from other academic disciplines align to some extend with taste-based or statistical discrimination. On the one hand, biases such as the in-group and out-group bias rely on the personal preferences for certain groups (Bertrand and Duflo 2017) which is also the starting point for taste-based discrimination. On the other hand, some biases have more commonalities with statistical based discrimination such as the halo-effect for example as this bias occurs when one wants to avoid uncertainty based on (in)accurate beliefs (Bello 2004). Nevertheless, our study focuses on the two theories of hiring discrimination that are rooted in the economic framework: taste based and statistical based discrimination (Neumark 2018). Sexual orientation stereotypes andjob candidate screening:… Page 5 of 40 16 H2B: Same-sex marriages are negatively and indirectly associated with hiring probabilities through prejudices regarding collaborations. Second, the theory of levels-based statistical discrimination provides a more rational explanation for discrimination against candidates in same-sex marriages. The starting point of this theory is that employment decisions are made under uncertainty. After all, employers do not possess perfect information on individual candidates. To aid decision-making under uncertainty, employers apply their general productivity beliefs of candidates in same-sex relationships as a group to the individual candidate. Negative productivity beliefs concerning minority groups then create advantages for heterosexual candidates and this, therefore, results in different hiring probabilities for individual candidates from both groups. Moreover, these beliefs need not necessarily match reality to lead to unequal treatment according to the theory of inaccurate statistical discrimination (Bohren etal.2019). Following the theory of levels-based statistical discrimination, we hypothesise that same-sex marriages and hiring probabilities are negatively and indirectly associated via employers’ productivity beliefs about candidates.2 H3A: Same-sex marriages have a negative effect on the employer’s productivity beliefs about candidates. H3B: Same-sex marriages are negatively and indirectly associated with hiring probabilities through productivity beliefs. Both the theories of taste-based and levels-based statistical discrimination, to some extent, rely on explaining hiring discrimination through the stereotypes which employers relate to the candidate’s relationship status. Specifically, stereotypes regard a socially shared set of (un)conscious and (in)accurate beliefs about characteristics of members of a social group such as sexual minorities (Judd and Park 1993; Banaji 2002). Employers could use such stereotypes as cognitive shortcuts in selection decisions since they simplify and justify social reality (Fiske 1998). Taken together, discrimination, prejudice, and stereotypes should be interpreted as related but distinct concepts (Fiske 1998). More concretely stereotypes represent beliefs about certain characteristics of group members (i.e. the cognitive component), prejudice reflects the emotional reaction to those stereotypes (i.e. the affective component), and discrimination refers to actions based on those stereotypes (i.e. the behavioural component). Indeed, many experiments indicate that stereotyping may affect hiring decisions, and subsequently, result in hiring discrimination against members of different social groups (Van Belle etal.2018; Sterkens etal.2021; Van Borm etal.2021). 2 As remarked by one of the reviewers, stereotypes, and therefore statistical discrimination, could vary at the intersection of gender and sexual orientation. This could also explain the disadvantage of gay men compared to lesbian women (Flage 2020). P.Sterkens et al. 16 Page 6 of 40 3 Data To make these contributions, we conducted a survey as a pre-study which was employed to set up a vignette experiment at a later stage. Vignette studies are controlled experiments that integrate experimental manipulations in a survey set-up. They are commonly employed to analyse human decision-making in the context of hiring decisions (Van Belle etal.2018; Sterkens etal.2021; Van Borm etal.2021). In such experiments, participants evaluate fictitious candidate profiles (vignettes) that vary across several characteristics (vignette dimensions, for instance ‘job-relevant experience’) on a predetermined number of levels (vignette levels, for instance, ‘yes, no’). Vignette experiments could be favoured over administrative data when studying hiring discrimination because they enable a causal interpretation of candidate manipulations—whereas administrative worker data could vary by confounding characteristics. Moreover, compared to correspondence experiments—the golden standard for measuring hiring discrimination (Neumark 2018)—vignettes facilitate the surveying of participants’ thought processes behind decisions. Hence, vignettes are more suitable for testing explanations for hiring discrimination. In contrast to prior controlled experiments on hiring discrimination against sexual minority candidates, which primarily featured student populations (Pichler etal.2010; Binder and Ward 2016; Baert 2018a), our study complements a small body of controlled experiments among genuine HR professionals. Compared to the two vignette experiments featuring HR professionals (Van Hoye and Lievens 2003; Barron 2009), our experiment is innovative because we not only measure hireability but also gauge the candidate perceptions related to candidates in a same-sex marriage. Although controlled experiments allow researchers to isolate variables and elaborately survey their participants, they are particularly susceptible to social desirability bias. Throughout Section3, we discuss measures taken to limit such bias such as multiple manipulations (SubSection3.2) and the administration of a social desirability scale (SubSection3.4). Finally, we discuss this limitation in Section5. 3.1 Pre‑study To investigate the stereotyping function of same-sex marriages and the explanatory potential of stereotypes for hiring probabilities, we reviewed the literature for potential stereotypes emitted by gay job candidates.3 More concretely, we searched the Web of Science international database for relevant studies published in economic, social, psychological, or interdisciplinary journals between 1965 and 2021. The keywords used included various synonyms and alternative combinations of the 3 The pre-study examines perceptions regarding gay men and lesbian women since the vignette experiment was initially targeted at gay and heterosexual candidates. However, a reviewer of the initial version of the manuscript correctly noted that our vignette experiment deals rather with same-sex couples (and, thus, by extension also other sexual minorities) by which the congruence between the pre-study and the vignette experiment is not perfect. Sexual orientation stereotypes andjob candidate screening:… Page 7 of 40 16 following search term: gay lesbian stigma (e.g. signals homosexuality). Based on the references in and to the studies found, additional studies were selected. In total, we identified 70 characteristics (see Appendix Table1). However, presenting each of these candidate perceptions as items to recruiters would have put unreasonable cognitive demands on participants. As Bethlehem and Biffignandi (2012) explain, research requiring excessive cognitive effort jeopardises both the response rate and data quality due to respondents’ satisficing tendencies (i.e. the ‘less attentive answering of items’). Consequently, we conducted a pre-study in which we applied item reduction techniques to filter out the stereotypes that fit three criteria: applicability, relevance to the work context, and limited overlap. First, when reviewing the literature, we traced back studies as early as 1965 but limited ourselves to the investigation of those stereotypes applicable to homosexuality as perceived by contemporary recruiters. Second, although the identified stereotypes span a broad range of characteristics, not all of these characteristics are necessarily relevant to the work context. For example, participating recruiters indicated that non-conformism and the need for security are fairly irrelevant. Third, we retained those stereotypes that showed a limited overlap with one another because research on social cognition has evidenced that there are dimensions underlying stereotypes of homosexuality (Fiske etal.2007). Therefore, we excluded passiveness as a stereotype as we already took the opposite stereotype, namely assertiveness, into account. 3.1.1 Data collection We conducted our pre-study in the form of an online survey and followed an approach comparable to the Sexual Prejudice Scale (Chonody 2013). Employing the services of the online panel service Prolific, 50 British and 50 American individuals experienced in making recruitment decisions (hereafter referred to as recruiters) completed our pre-study’s four sets of questions.4 In the first three separated batteries, recruiters indicated the degree to which they agreed with 70 statements concerning the characteristics of (1) the average gay man, (2) the average lesbian woman, and (3) the relevance of each characteristic to a hiring decision. For example, approachability, eccentricity, group orientation, honesty, intelligence, and social competence were assessed—the full list is presented in Appendix Table1—employing a 6-point response scale, ranging from strongly disagree (score 1) to strongly agree (score 6).5 The fourth battery was used to register the participant’s sociodemographics: gender (man, woman, non-binary or third gender, prefer not to say); age in years (numbers); nationality (British, American, other); level of education (no 4 By using Prolific, we could reach motivated and suitable participants based on the characteristics they entered when registering on this platform. In this case, we only invited participants with the American or British nationality who indicated that they have experience with hiring decisions (i.e. they have been responsible for hiring job candidates). 5 Our 6-point response scale did not contain a neutral option, thus forcing respondents to express (dis) agreement with statements. This is common practice to avoid social desirability bias when measuring socially-sensitive attitudes (Chonody 2013). P.Sterkens et al. 16 Page 8 of 40 diploma, high school, bachelor, masters, PhD); and sexual orientation (lesbian, gay, bisexual, heterosexual, other, prefer not to say). 3.1.2 Stereotype elimination procedure The item reduction process consisted of four subsequent phases—the results of which are presented in the fourth column of Appendix Table1. In the first phase, we strictly filtered out stereotypes based on descriptives. More concretely, items that were, on average, perceived by recruiters as irrelevant to a hiring decision or inapplicable to gay men or lesbian women (i.e. an average of below 3) were dropped.6 Hence, we eliminated 23 items. In this step, we dropped a further four items because they were close approximations of overall hireability and, therefore, of the hiring decision to be made in the experiment (for example ‘effective performance of jobrelated tasks’). Next, the second phase of the elimination procedure involved a re-examination of the item pool following factor analyses on the evaluations of stereotypes for gay men and lesbian women separately. Here, we dropped two items because they did not fit the factor structures emerging for either gay men or lesbian women (i.e. ‘nonconformist’ had a factor loading lower than the conservative threshold of 0.35 which is used to determine statistically meaningful factors; Comrey & Lee 1992) or were empirical opposites of other stereotypes (i.e. ‘passive’ was dropped in exchange for ‘assertive’). In the third phase of the eliminations, we discussed emerging themes within the stereotype factors from the previous step (‘factor interpretation’) and based the selection on item interpretations and their underlying correlations. Specifically, 17 original items were summarised in five newly generated items (‘empathy’, ‘creativity’, ‘loving and soft’, ‘self-awareness’, and ‘emotionality’) that fit the factor structures. Furthermore, after a re-examination of the correlation matrixes, we dropped another four items because they showed a substantial overlap with other stereotypes and had limited relevance to the hiring decision (average below 3.250). In the fourth and final phase, the face validity of each individual item was scrutinised. Consequently, we excluded five more items because, from experiences in the field, these candidate characteristics were less likely to be gauged from the earlier phases of resume screening (e.g. how individualistic a candidate is). The remaining 20 items were subjected to the evaluation of labour market experts. Based on their input regarding health stigma, we agreed on a reduction of the 70 initial items to the following 21 potential stereotypes of being a gay or lesbian job candidate: social skills, assertiveness, outspokenness, dominance, independence, competitiveness, leadership abilities, team orientation, empathy, lovingand softness of personality, emotional sensitivity, neatness, intelligence, open-mindedness, 6 Nevertheless, we acknowledge that some of these items might still be relevant in practice. Participating recruiters may agree less strongly with certain statements in our pre-study because they are aware that this could indicate recruitment discrimination which is prohibited by law. Sexual orientation stereotypes andjob candidate screening:… Page 15 of 40 16 Third, and central to our design, were the 21 statements measuring candidate perceptions. In this phase of the experiments, we implemented our systematicallyselected list of items from the pre-study (SubSection3.1) to collect causal evidence for the stereotyping function of candidates’ sexual orientation during the hiring process. Based on the theory of levels-based statistical discrimination, each of these potential stereotypes might be (part of) the explanation for hiring discrimination against candidates in a same-sex couple. 3.5.4 Post‑experimental questionnaire In a final step, recruiters completed a post-experimental questionnaire which is reported at the of the Appendix. We used the data collected via these items to explore recruiter-side heterogeneity of hiring discrimination and the execution of robustness checks. Socio-demographic variables surveyed were gender (man, woman, non-binary/ third gender, prefer not to say); age in year, educational degree (primary, lower secondary, higher secondary, bachelors, masters, PhD), and sexual orientation (heterosexual, lesbian, gay, bisexual, other, prefer not to say). Subsequently, we surveyed recruiters’ professional experiences, namely hiring tenure (less than 1 year, 1 to 5years, more than 5years) and frequency (none, between 1 and 5 times per year, more than 5 times per year).14 Next, we incorporated two measures of the recruiters’ experiences with sexual minority candidates. As such, we administered the West and Hewstone (2012) scale for contact with gay people. This scale contained four items (α = 0.888) scored on a 7-point Likert scale from no contact at all to very frequent contact (under nonCOVID circumstances). Each of these items referred to different contexts (at school/ work, daily superficial social contact, intimate social situations, all sorts of social situations) in which participants encountered gay individuals. Subsequently, we averaged participants’ scores into a single scale score, ranging from one to seven. However, we not only measured the frequency of contact with gay people but also the recruiters’ private attitudes, employing the Modern Homonegativity Scale (MHS), developed by Morrison and colleagues (1999). This validated scale measures the covert negative attitudes of participants towards gay individuals by statements as ‘Gay individuals seem to focus on the ways in which they differ from heterosexual individuals and ignore the ways in which they are the same’ for example. As such, we asked participants to which degree they agreed with each of the scale’s 12 items (α = 0.948) on a 5-point Likert scale. We merged the individual item responses into an average scale score, whereby a higher score indicated a stronger endorsement of modern homonegative attitudes (McCutcheon and Morrison 2015). Between these items, we implemented an additional attention check, asking the participants to indicate the option ‘strongly agree’ (SubSection3.4). 14 In hindsight, it would also have made sense to survey the political affiliation and specific job of the participant as, according to an anonymous reviewer, respondents on Prolific tend to be more liberalminded and self-employed which could skew the results. P.Sterkens et al. 16 Page 16 of 40 Another addition to our post-experimental questionnaire was Reynolds’ (1982) shortened Marlowe–Crowne Social Desirability Scale, which we used to measure the participant’s socially desirable response tendencies. For each of the 13 items (α = 0.790) expressing behaviour that is either culturally approved or sanctioned (e.g. ‘I sometimes try to get even rather than forgive’), we asked participants to indicate whether the statements were applicable to them (true, false). This scale is implemented and validated across different contexts (Beretvas etal.2002; Van Borm and Baert 2018). Its total score is calculated as the (standardised) sum of all statements indicated as true. The final scale we administered to the recruiters measured their level of risk aversion as might act as a driver of labour market discrimination against sexual minority candidates. For example, the study of Baert (2018a) supports this as gay men were less likely to be hired when the employers were more risk-averse. We implemented Baert’s adaptation of the Domain-Specific Risk-Taking Scale (Blais and Weber 2006) which contained six different items (α = 0.771), each describing a professional risk (e.g. ‘starting a new career in your mid-thirties’). Items were rated for the likelihood they would engage in this behaviour (1, extremely unlikely; 7, extremely likely). The weighted average of all item scores resulted in a global risktaking score.15 3.6 Data description Because of the random allocation of our vignette decks and fictitious vacancies, we expected low correlations between the fictitious candidates’ sexual orientation, job, and recruiter variables. The statistically insignificant t-tests and chi-squared tests presented in Table2 below confirm the success of our experimental setup. Moreover, the D-efficiency algorithm’s success is also demonstrated by the low correlation between candidate dimensions (maximum 0.093). The recruiter characteristics panel from Table 2 further describes the sample’s composition. On average, recruiters were 44 years old. They identified themselves as men (50.0%), women (49.0%), or other gender identities (1.0%). The majority of the recruiters considered themselves heterosexual (88.9%). Of the different levels of education, bachelor’s degrees (47.0%) were the most common. Participants possessed considerable tenure in making hiring decisions. In this study, 48.3% of the sample reported having more than 5years of experience. Notably, 49.5% of the recruiters did not evaluate any candidates in the last year—this substantial share might be explained by hiring freezes initiated because of the COVID-pandemic (Campello etal.2020). Furthermore, our descriptive statistics suggest that the average participant had—at least—occasional contact with gay individuals (average 3.680, maximum 7) and harboured non-negative attitudes towards them (MHS average 2.204, maximum 5). Our sample was comparable to the 394,644 US and UK HR managers who participated respectively in the American Community Survey (Census Bureau 2022) and 15 Note that we are measuring—instead of manipulating—recruiter characteristics. Consequently, inferences based on heterogeneity analyses with recruiter characteristics are non-causal. Sexual orientation stereotypes andjob candidate screening:… Page 17 of 40 16 Table 2 Description of the participating recruiters by experimental condition To test the independence between the participant characteristic and the experimental condition, a chi-square (indicator variable) or Kruskal–Wallis (continuous variable) test is conducted Proportion (indicator variables) or mean (continuous variables) Independence test (p-value) Full sample Experimental condition Man in different-sex marriage Woman in differentsex marriage Man in same-sex marriage Woman in same-sex marriage Woman 49.0% 50.4% 47.2% 46.6% 52.6% 0.369 Not heterosexual 11.1% 11.3% 11.7% 10.8% 10.0% 0.982 Age 44.156 44.261 44.108 44.063 44.148 1.000 No tertiary education 25.5% 24.9% 23.9% 28.0% 27.4% 0.530 Bachelor’s degree 47.0% 47.7% 47.5% 47.0% 44.8% 0.919 Master’s degree 22.8% 23.4% 23.2% 20.5% 23.0% 0.920 British 49.5% 49.0% 50.5% 51.1% 46.7% 0.697 Hired people in the past year 50.5% 49.8% 50.0% 48.5% 48.9% 0.974 Hiring experience of more than 5years 48.3% 48.1% 48.8% 48.5% 47.4% 0.985 Contact with gay people (standardised) 0.000 − 0.050 0.001 0.079 0.015 0.449 Homonegativity (standardised) 0.000 0.011 0.004 0.004 − 0.034 0.852 Social desirability (standardised) 0.000 0.022 0.019 − 0.013 − 0.070 0.786 Risk aversion (standardised) 0.000 0.006 − 0.023 0.046 − 0.009 0.647 P.Sterkens et al. 16 Page 18 of 40 UK Census (Office for National Statistics 2021) in 2021. More concretely, these HR managers were found to be predominantly women (US = 66.85%; UK = 61.05%) and between 40 and 49years old (US = 25.85%; UK = 30.74%). Moreover, the majority of the US HR managers have a Bachelor degree (39.97%). In addition, our sample aligns with the general US and UK populations in terms of sexual orientation. Specifically, Gallup data estimated that 7.6% of Americans identify as LGBTQ + (Jones 2024) and Census data from 2022 find 6.6% of the UK population do not identify as heterosexual (Office for National Statistics 2023). 4 Results In what follows, we examine the effect of same-sex marriages on hiring probabilities for both men and women using a multivariate regression framework (SubSection4.1). Next, through multiple mediation analyses, we investigate the stereotyping function of sexual orientation and its role in explaining differences in recruitment decisions (SubSection4.2). Subsequently, we explore heterogeneity in the effects of sexual orientation on interview probabilities and perceptions employing moderation analyses (SubSection4.3). Finally, in robustness checks, we test whether different subsamples hold the same perceptions regarding sexual minority candidates (SubSection4.4). For all analyses we conducted, the standard errors are corrected for clustering of the observations at the recruiter level. In the subsections below, we only discuss the results for the interview probability as this is the most proximal outcome in our experimental set-up (Sterkens etal.2021). However, the analyses on hiring probability yield similar results and are available in Tables4 and 5 in the Appendix. 4.1 Standard regression analyses To examine causal differences in interview probability between the investigated sexual orientations, we fit a multivariate linear regression model (see Eq.1) of interview probabilities ( Y) on the candidate’s sexual orientation (SO) and the other candidate ( CC) , vacancy (VC) , and recruiter characteristics (RC) discussed in Section3. The model’s estimates are presented in the first column of Table3. The other columns are further discussed in SubSection4.3. In general, we find a statistically significant causal effect of the candidates’ sexual orientations on the interview probability. That is, candidates in a female same-sex marriage have a 4.2 percentage point (p = 0.012) higher interview probability than the regression’s reference category of male candidates in a different-sex marriage.16 This equals an increase of 0.147 standard deviation in (1) Y=𝛼Y+𝛽YSO +𝛾YCC +𝛿YVC +𝜃YRC +𝜀Y 16 All beta coefficients could be interpreted as differences expressed in percentage points by multiplying these coefficients by 10. This interpretation is adequate because the evaluation and perception scales ranged from 0 to 10. Sexual orientation stereotypes andjob candidate screening:… Page 19 of 40 16 Table 3 Main and moderation effects with interview probability as the outcome Interview probability (1) (2) (3) (4) (5) A. Candidate characteristics  Sexual orientation (ref. = man in different-sex marriage)   Woman in different-sex marriage 0.224* (0.125) 0.206 (0.125) 0.223 (0.125) 0.228* (0.125) 0.208* (0.126)   Man in same-sex marriage 0.251 (0.167) 1.662 (1.656) 0.437 (0.377) 1.203* (0.653) 3.008 (1.823)   Woman in same-sex marriage 0.424** (0.168) 0.624 (2.400) 0.388 (0.350) 0.646 (0.638) 1.031 (2.475)  Age (c.) − 0.001 (0.008) 0.004 (0.010) − 0.001 (0.008) − 0.000 (0.008) 0.003 (0.011)  Experience (ref. = none)   Two years 2.988*** (0.143) 3.074*** (0.186) 3.007*** (0.143) 2.973*** (0.145) 3.086*** (0.189)   Five years 4.059*** (0.164) 4.521*** (0.195) 4.066*** (0.164) 4.047*** (0.165) 4.525*** (0.197)  Foreign language knowledge (ref. = none)   French 0.089 (0.125) 0.029 (0.145) 0.087 (0.126) 0.059 (0.124) 0.035 (0.147)   Spanish 0.164 (0.123) − 0.017 (0.155) 0.166 (0.123) 0.165 (0.124) − 0.020 (0.157)  Professional achievements (ref. = none)   Diversity ambassador 0.837*** (0.121) 1.285*** (0.159) 0.832*** (0.121) 0.864*** (0.122) 1.286*** (0.159)   Employee of the month 0.760*** (0.116) 0.944*** (0.151) 0.762*** (0.115) 0.777*** (0.117) 0.949*** (0.151)  Hobbies (ref. = none mentioned)   Wrestling 0.597*** (0.193) 0.700*** (0.227) 0.605*** (0.191) 0.582*** (0.189) 0.714*** (0.228)   Gymnastics 0.462** (0.201) 0.573** (0.255) 0.469** (0.199) 0.455** (0.202) 0.571** (0.256)   Tennis 0.365* (0.193) 0.656*** (0.238) 0.372* (0.192) 0.390** (0.193) 0.660*** (0.240)   Volunteers to distribute food for local community 0.716*** (0.176) 1.025*** (0.213) 0.731*** (0.175) 0.698*** (0.176) 1.024*** (0.215)   Volunteers at LGBTQ rights organisation 0.409** (0.189) 0.443* (0.229) 0.427** (0.187) 0.415** (0.192) 0.443* (0.231) B. Vacancy characteristics  Gender-type (ref. = neutral job)   Male-dominated job 0.230 (0.192) 0.195 (0.190) 0.352* (0.208) 0.214 (0.192) 0.308 (0.208) P.Sterkens et al. 16 Page 20 of 40 Table 3 (continued) Interview probability (1) (2) (3) (4) (5)   Female-dominated job 0.035 (0.193) − 0.000 (0.191) 0.038 (0.210) 0.045 (0.193) 0.038 (0.211)  Customer contact: high 0.168 (0.157) 0.169 (0.156) 0.203 (0.173) 0.202 (0.156) 0.236 (0.175)  Diversity statement: included 0.132 (0.154) 0.111 (0.153) 0.068 (0.168) 0.130 (0.153) 0.062 (0.169) C. Participant characteristics  Gender (ref. = man)   Woman − 0.048 (0.163) − 0.084 (0.160) − 0.044 (0.163) − 0.066 (0.174) − 0.066 (0.175)  Sexual orientation (ref. = heterosexual)   Not heterosexual − 0.549** (0.244) − 0.533** (0.239) − 0.565** (0.246) − 0.450* (0.254) − 0.440* (0.255)  Age (cont.) 0.000 (0.006) − 0.001 (0.007) − 0.000 (0.007) 0.004 (0.007) 0.004 (0.007)  Educational degree (ref. = lower than tertiary)   Tertiary education − 0.382** (0.182) − 0.396** (0.182) − 0.387** (0.182) − 0.293 (0.201) − 0.305 (0.204)  Nationality (ref. = USA)   UK − 0.039 (0.157) − 0.031 (0.158) − 0.039 (0.157) − 0.088 (0.172) − 0.065 (0.173)  Hiring experience (ref. = less than 5years)   More than 5years 0.081 (0.160) 0.083 (0.158) 0.081 (0.160) 0.027 (0.175) 0.014 (0.176)  Contact with gay people (s.) 0.187** (0.087) 0.185** (0.085) 0.180** (0.087) 0.174* (0.095) 0.181* (0.094)  Homonegativity (s.) − 0.454*** (0.093) − 0.442*** (0.092) − 0.455*** (0.093) − 0.314*** (0.093) − 0.313*** (0.093)  Risk aversion (s.) 0.168* (0.094) 0.149 (0.094) 0.174* (0.094) 0.144 (0.111) 0.135 (0.112) D. Interactions with candidate characteristics  Man in same-sex marriage × age − 0.009 (0.035) − 0.003 (0.036)  Man in same-sex marriage × 2years − 0.048 (0.466) − 0.195 (0.487)  Man in same-sex marriage × 5years − 1.379** (0.564) − 1.533*** (0.573)  Man in same-sex marriage × French 0.349 (0.488) 0.283 (0.501) Sexual orientation stereotypes andjob candidate screening:… Page 21 of 40 16 Table 3 (continued) Interview probability (1) (2) (3) (4) (5)  Man in same-sex marriage × Spanish 0.306 (0.459) 0.341 (0.449)  Man in same-sex marriage × diversity ambassador − 0.447 (0.485) − 0.367 (0.487)  Man in same-sex marriage × employee of the month − 0.060 (0.410) − 0.014 (0.414)  Man in same-sex marriage × wrestling − 1.004 (0.629) − 0.978 (0.631)  Man in same-sex marriage × gymnastics 0.050 (0.646) 0.143 (0.654)  Man in same-sex marriage × tennis − 1.044 (0.638) − 0.987 (0.652)  Man in same-sex marriage × volunteers to distribute food for local community − 1.063 (0.648) − 1.149* (0.652)  Man in same-sex marriage × volunteers at LGBTQ rights organisation − 0.816 (0.733) − 0.688 (0.764)  Woman in same-sex marriage × age 0.021 (0.050) 0.022 (0.049)  Woman in same-sex marriage × 2years − 0.257 (0.555) − 0.211 (0.554)  Woman in same-sex marriage × 5years − 1.251** (0.500) − 1.272** (0.506)  Woman in same-sex marriage × French − 0.436 (0.512) − 0.607 (0.502)  Woman in same-sex marriage × Spanish 0.149 (0.485) 0.144 (0.496)  Woman in same-sex marriage × diversity ambassador − 1.172* (0.678) − 1.230 (0.670)  Woman in same-sex marriage × employee of the month − 0.334 (0.481) − 0.252 (0.479)  Woman in same-sex marriage × wrestling 0.205 (0.679) 0.124 (0.704)  Woman in same-sex marriage × gymnastics 0.758 (0.905) 0.966 (0.919)  Woman in same-sex marriage × tennis − 0.348 (0.655) − 0.252 (0.672)  Woman in same-sex marriage × volunteers to distribute food for local community − 0.218 (0.601) − 0.166 (0.610)  Woman in same-sex marriage × volunteers at LGBTQ rights organisation − 0.044 (0.923) − 0.009 (0.959) P.Sterkens et al. 16 Page 22 of 40 Table 3 (continued) Interview probability (1) (2) (3) (4) (5) E. Interactions with vacancy characteristics  Man in same-sex marriage × male-dominated job − 0.179 (0.392) − 0.294 (0.373)  Man in same-sex marriage × female-dominated job − 0.494 (0.351) − 0.518 (0.358)  Man in same-sex marriage × high customer contact 0.052 (0.299) − 0.088 (0.305)  Man in same-sex marriage × diversity statement included 0.020 (0.294) − 0.035 (0.292)  Woman in same-sex marriage × male-dominated job 0.220 (0.375) 0.209 (0.380)  Woman in same-sex marriage × female-dominated job − 0.210 (0.391) − 0.336 (0.375)  Woman in same-sex marriage × high customer contact − 0.276 (0.307) − 0.130 (0.310)  Woman in same-sex marriage × diversity statement included 0.370 (0.310) 0.314 (0.306) F. Interactions with recruiter characteristics  Man in same-sex marriage × woman 0.350 (0.307) 0.270 (0.317)  Man in same-sex marriage × not heterosexual − 0.283 (0.458) − 0.406 (0.461)  Man in same-sex marriage × age (cont.) − 0.018 (0.015) − 0.025 (0.016)  Man in same-sex marriage × tertiary education − 0.520 (0.326) − 0.503 (0.340)  Man in same-sex marriage × UK 0.043 (0.315) 0.127 (0.316)  Man in same-sex marriage × more than 5years of hiring experience 0.158 (0.313) 0.263 (0.316)  Man in same-sex marriage × contact with gay people (s.) − 0.154 (0.172) − 0.181 (0.183)  Man in same-sex marriage × homonegativity (s.) − 0.414** (0.194) − 0.322 (0.204)  Man in same-sex marriage × risk  aversion (s.) 0.081 (0.177) 0.046 (0.179)  Woman in same-sex marriage × woman − 0.322 (0.319) − 0.408 (0.321)  Woman in same-sex marriage × not heterosexual − 0.230 (0.480) − 0.212 (0.500)  Woman in same-sex marriage × age (cont.) − 0.008 (0.013) − 0.010 (0.013) Sexual orientation stereotypes andjob candidate screening:… Page 23 of 40 16 Table 3 (continued) Interview probability (1) (2) (3) (4) (5)  Woman in same-sex marriage × tertiary education 0.161 (0.349) 0.098 (0.367)  Woman in same-sex marriage × UK 0.296 (0.326) 0.078 (0.336)  Woman in same-sex marriage × more than 5years of hiring experience 0.139 (0.310) 0.172 (0.316)  Woman in same-sex marriage × contact with gay people (s.) 0.217 (0.183) 0.140 (0.184)  Woman in same-sex marriage × homonegativity (s.) − 0.393** (0.184) − 0.443** (0.184)  Woman in same-sex marriage × risk aversion (s.) 0.014 (0.179) 0.044 (0.181) N1616 Abbreviations used: s. (scale consisting of multiple items) and ref. (reference category). The presented statistics are coefficient estimates and their standard errors in parentheses for the mediation model outlined in SubSect.4.2. Standard errors are corrected for clustering of the observations at the participant level. *p < 0.10; **p < 0.05; ***p < 0.01 P.Sterkens et al. 16 Page 24 of 40 interview probability. Interview probabilities of other groups are not (men in samesex marriages: β = 0.251, p = 0.134) or marginally (women in different-sex marriages: β = 0.224, p = 0.072) significantly different from those of men in different-sex marriages. Hence, we find only partial evidence for Hypothesis 1. 4.2 Multiple mediation analyses Using multiple mediation analyses (Hayes 2017), we explore explanations for the effect of sexual orientation on interview probabilities. More specifically, we statistically test pathways through which independent variables influence a dependent variable. In our design, schematically represented in Fig.1, we evaluate whether the effect of sexual orientation on interview probabilities (‘c-path’ in Fig.1) can be explained by indirect pathways via candidate perceptions (‘ab-path’ in Fig.1; proposed mediation). We explore such indirect effects by multiplying the causal effect estimates of sexual orientation on the perception items (‘a-path’ in Fig.1) with the associations between those perception items and interview probability (‘b-path’ in Fig.1; Baron and Kenny1986).17 However, these multiplications should be interpreted as associations because our experimental setup is limited to the causal interpretation of relationships between sexual orientation and perceptions (‘a-path’ in Fig.1), and sexual orientation and the interview probability (‘c-path’ in Fig.1).18 Indeed, we were unable to manipulate candidate perceptions (i.e. our mediators) separately, and there could be additional unmeasured variables confounding the model. As a hypothetical example, if the hiring penalties of candidates in same-sex relations are driven by unmeasured hypothetical perceptions (e.g. ‘sexual minority candidates are offending’), any indirect relationship (‘ab-path’ in Fig.1) through measured perceptions such as ‘outspokenness’ could be biased. Applied to the data, the multiple mediation framework consists of 23 linear regressions. Twenty-two of these regress the perception items ( PI i) on the Fig. 1 Schematic representation of mediation analyses 17 ‘‘Indirect effect’’ is common terminology in the mediation literature (Hayes 2017). 18 This limitation arises because we can only exert full experimental control over the candidate characteristics that are manipulated (e.g. sexual orientation). For the other independent variables in our models, namely, the vacancy and recruiter characteristics, we cannot exclude the possibility of unmeasured confounders. We discuss this matter further in Section5. Sexual orientation stereotypes andjob candidate screening:… Page 31 of 40 16 Three notable trends regarding the interaction between candidates married to a same-sex partner and recruiters’ homonegative attitudes emerge from the summarising results presented in Table6. We find that, compared to the general sample, harbouring homonegative attitudes is associated with (i) additional stereotyping functions of same-sex marriages and (ii) relatively less positive stereotypes but also (iii) agreement on sexual orientation’s stereotyping function for various perceptions— regardless of attitude. First, recruiters with homonegative attitudes are significantly more likely to interpret same-sex marriages as negative stereotypes for leadership skills (men: β = − 0.462; women: β = − 0.400), professionalism (men: β = − 0.580; women: β = − 0.416), career orientations (men: β = − 0.300; women: β = − 0.350), and current health (men: β = − 0.474; women: β = − 0.371), whereas, in the full sample, recruiters generally do not perceive same-sex marriages as stereotyping for these Table 6 Moderation analyses between candidates’ same-sex orientations and participants’ homonegative attitudes, with perception items as outcomes and men in different-sex marriages as the reference category Abbreviation used: s. (scale consisting of multiple items). The regression model’s predictors are identical to those in the final column of Table3in the Appendix. Standard errors are corrected for clustering of the observations at the participant level. *p < 0.10; **p < 0.05; ***p < 0.01 Mediators Man in same-sex marriage × homonegativity (s.) Woman in same-sex marriage × homonegativity (s.) β p β p Perceived collaboration (s.) − 0.157 0.462 − 0.289* 0.078 Perceived social skills − 0.368*** 0.005 − 0.248** 0.049 Perceived assertiveness − 0.274** 0.049 − 0.264* 0.077 Perceived outspokenness − 0.056 0.675 − 0.103 0.449 Perceived dominance − 0.333** 0.010 − 0.185 0.201 Perceived independence − 0.179 0.165 − 0.299** 0.028 Perceived competitiveness − 0.367*** 0.002 − 0.273* 0.066 Perceived leadership − 0.462*** 0.001 − 0.400*** 0.002 Perceived team orientation − 0.335** 0.010 − 0.324** 0.011 Perceived empathy − 0.167 0.162 − 0.265* 0.069 Perceived softness of personality − 0.042 0.747 − 0.142 0.253 Perceived emotional sensitivity 0.077 0.624 − 0.005 0.977 Perceived neatness 0.004 0.970 − 0.177 0.196 Perceived intelligence − 0.408*** 0.001 − 0.483*** 0.001 Perceived open-mindedness 0.050 0.730 0.108 0.448 Perceived creativity − 0.276** 0.043 − 0.020 0.863 Perceived talkativeness − 0.131 0.327 0.001 0.996 Perceived honesty − 0.458*** 0.001 − 0.312** 0.036 Perceived professionalism − 0.580*** 0.001 − 0.416*** 0.005 Perceived self-awareness − 0.424*** 0.003 − 0.095 0.533 Perceived career orientation − 0.300** 0.016 − 0.350*** 0.005 Perceived current health − 0.474*** 0.001 − 0.371** 0.010 P.Sterkens et al. 16 Page 32 of 40 characteristics at all (Subsection4.3.1). The same trend applies to the competitiveness of men (β = − 0.367) but not to women in same-sex marriages. Second, whereas recruiters generally derive positive stereotypes from same-sex marriages, the recruiters expressing more homonegative attitudes derive relatively fewer positive stereotypes from such marriages. More concretely, the latter think more negatively about the intelligence (men: β = − 0.408; women: β = − 0.483), social skills (men: β = − 0.368; women: β = − 0.248), honesty (men: β = − 0.458; women: β = − 0.312), and team orientation (men: β = − 0.335; women: β = − 0.324) of candidates in same-sex marriages. This is also the case for the perceived dominance (β = − 0.333), creativity (β = − 0.276), and self-awareness (β = − 0.424) of men married to a same-sex partner, and for the perceived independence (β = − 0.229) of women married to a same-sex partner. Third, recruiters with homonegative attitudes appear to have similar perceptions as other recruiters concerning candidates in a same-sex marriage collaboration with others, empathy, soft personality, emotional sensitivity, neatness, talkativeness, open-mindedness, and outspokenness. In conclusion, our data suggest that the association between homonegative attitudes and stereotyping is of a rather complex nature as we establish several points of convergence and divergence between the perceptions of recruiters which vary in terms of homonegative attitudes. 4.4 Robustness analyses In response to vignette experiments’ known susceptibility to socially desirable responding, we follow Steenkamp and colleagues’ (2010) guidelines in further analysing the data for potential social desirability bias. First and foremost, separate moderation analyses suggest that the interview and hiring probabilities of men and women married to a same-sex partner are unrelated to the social desirability scores of recruiters.20 Second, we calculate that social desirability has a limited association with recruiters’ responses to the modern homonegativity scale (correlation coefficient = 0.133). Finally, we checked whether participants from the United Kingdom and the United States held the same perceptions regarding candidates in a same-sex marriage. A first robustness check indicated that the results of the United States sample strongly matched the results of the full sample.21 A second robustness check which included two-way interactions between participants from the United Kingdom and the sexual orientations revealed a limited number of small perception differences. More concretely, as shown Appendix Table10, UK participants perceived male candidates in a same-sex marriage more negatively in terms of open-mindedness (β = − 0.659, p = 0.008), softness of personality (β = − 0.503, p = 0.027), neatness (β = − 0.426, p = 0.047), and emotional sensitivity (β = − 0.425, p = 0.094) compared to US participants. In contrast, female candidates in a same-sex marriage 20 The complete results of our social desirability analyses are depicted in Table8 in the Appendix. 21 The full results of the first robustness check with exclusively participants from the United States are presented in Table9 in the Appendix. Sexual orientation stereotypes andjob candidate screening:… Page 33 of 40 16 received higher scores in areas of self-awareness (β = 0.579, p = 0.015), leadership (β = 0.554, p = 0.018), dominance (β = 0.440, p = 0,062), and team orientation (β = 0.423, p = 0.071) from UK participants than US participants. These results indicate that while the vast body of perceptions is robust across countries, further research is warranted establishing cross-cultural nuances in stereotyping. 5 Conclusion To investigate the interview probabilities of candidates in same-sex marriages through stereotyping, we conducted a vignette experiment in which genuine recruiters evaluated job candidates who disclosed their marital status. The recruiters evaluated four candidates for one out of twelve job vacancies and shared their perceptions through 24 systematically selected items distilled from the literature and pre-studied among recruiters. In addition to providing causal evidence for a selection of stereotypes recruiters infer from gay men and lesbian women, we tested these stereotypes’ role in explaining interview probabilities. Moreover, we advanced our understanding of the literature’s contradictory findings related to sexual minority candidates’ hiring probabilities by analysing moderators of hiring discrimination. We find evidence that same-sex marriages generally emit positive and distinct stereotypes for men and women in a hiring context. Specifically, compared to men in different-sex marriages, recruiters perceive men in same-sex marriages as being more outspoken, open-minded, self-aware, emotionally sensitive, neat, intelligent, creative, talkative, and honest, showing more empathy in collaborations, and having more advanced social skills, more of a team orientation, and a more loving and soft personality—but being less pleasant to collaborate with. In contrast, women in same-sex marriages are seen as being more pleasant to collaborate with than their counterparts in different-sex marriages. However, they are also perceived as being more outspoken, open-minded, and self-aware than women in different-sex marriages. In addition, women in same-sex marriages are viewed as being more assertive, independent, and dominant compared to their counterparts in different-sex marriages, whereas this is not the case for men. The aforementioned effects are modest given their range between 1.9 and 7.6 percentage points. Although same-sex marriages activate many different stereotypes for men and women, only two are significantly associated with their interview probabilities: outspokenness and collaborations with employers, other employees, and customers. These perceptions might strengthen the interview probabilities of women in samesex marriages, whereas the opposite could be true for men. In addition, the stereotype that candidates married to a same-sex partner are more outspoken is negatively associated with their interview probabilities. Our moderation analyses provide additional insights regarding the circumstances under which hiring discrimination is more likely to occur. More specifically, we find tendencies that more experience yields a relatively lower hiring premium for candidates in same-sex marriages and that the effects of extracurricular activities (e.g. volunteering) and professional achievements (e.g. diversity ambassadorship) are also dependent on a candidate’s sexual orientation. The generally positive P.Sterkens et al. 16 Page 34 of 40 reception of candidates in same-sex marriages and a significant interaction effect with recruiters’ homonegative attitudes suggest that our data align well with a concentrated discrimination account (Campbell and Brauer 2021), whereby a minority of employers are responsible for most instances of hiring discrimination. Indeed, the generally positive perception patterns of candidates in same-sex marriages were frequently inversed among recruiters who privately held negative attitudes towards gay individuals. As suggested in Fric’s review (2017), equal opportunity policies could benefit from de-stigmatisation programmes. The current study’s findings complement this by calling for an efficient and targeted approach to such programmes as hiring discrimination might be centred around the negative attitudes of a limited proportion of recruiters. In addition, the perceptual patterns we evidenced for both groups of same-sex marriages could guide the development of targeted interventions. An example of such targeted approach could be to first monitor hiring discrimination via correspondence experiments and, subsequently, intervene in units (e.g. sectors) where discrimination against sexual minorities is prevalent. Policymakers could then enact legislation that increases the ‘accountability’ of individual recruiters. For instance, they could require (discriminating) organisations to adopt panel recruitment in which a team instead of a single recruiter screens resumes (Derous and Ryan 2019). Being held accountable for discriminatory decision-making would discourage homonegative individuals from acting on their prejudice. Moreover, similar interventions that structure communication, procedures, and interactions have been shown to effectively address workplace bias and discrimination of sexual minorities (Treffers etal.2024). From a candidate perspective, in particular, men in same-sex marriages could anticipate recruiters’ negative attitudes by implementing stigma-countering strategies. For example, Singletary and Hebl (2009) found that candidates presenting themselves as gay experienced fewer negative interactions with potential employers when they displayed more positivity. Applied to our findings, men in same-sex marriages would want to counter perceptions of outspokenness and unpleasant collaborations in the hiring process. An important caveat is that in anticipation of discrimination, men in same-sex marriages in particular should carefully consider the timing of their orientation’s disclosure. One frequently revisited limitation in this paper is the risk of socially desirable responses. In acknowledgment of this risk, we took measures to limit the impact of such bias. On the one hand, we simultaneously varied several candidate characteristics. In doing so, our experiment mimicked the trade-offs made in hiring decisions. On the other hand, we investigated the associations between recruiters’ evaluations and social desirable responding (Steenkamp etal.2010). Our operationalisation of the candidates’ sexual orientation leads to two other limitations of this study. More specifically, sexual orientation was revealed through the candidate’s marital status, namely being married to a same-sex partner. First, although this is a strong manipulation of sexual orientation, it could also come with perceptions associated with being married, such as the attributed health, reliability, and ambition. Moreover, certain stereotypes about single or unmarried candidates who are attracted to the same sex may not apply to those Sexual orientation stereotypes andjob candidate screening:… Page 35 of 40 16 who are married and vice versa. Consequently, the generalisability of our results is limited. Nevertheless, alternative manipulations of sexual orientation come with similar threats in terms of validity. For example, revealing sexual orientation by stating a candidate joined a LGBTQ + organisations could be ambiguous and generate stereotypes related to activism. Second, while we suspect that employers perceive candidates from a same-sex marriage as gay men or lesbian women, these candidates may equally belong to other sexual minority groups such as bisexual, queer, or pansexual individuals. Future research could focus on a more specific disclosure of sexual orientation as a treatment. A fourth limitation relates to intersectionality of discrimination based on sexual orientation with other discrimination grounds. For example, Pedulla (2014) showed that race and sexual orientation interact in a complex way to moderate discrimination in the United States. However, our study lacks those insights as no specific race or ethnicity was mentioned by which respondents may believe all candidates belong to the majority group (i.e. White European or American depending on the experimental context). This omission further limits the external validity of our experiment for ethnic minorities. Therefore, we recommend other researchers to explore the perceptions of such intersectional discrimination. Not only the interaction with race or ethnicity can be examined, but also those with other dominant or related discrimination grounds such as age and religion (Lippens etal. 2023). The perception measures employed in this study contain a fifth limitation. From the literature, numerous perceptions appeared to be related to homosexuality, which made it impossible to include all of them in the experiment. To resolve this issue, a preliminary study was conducted to reduce the items to a limited but validated set of perceptions. Although the delineation of these perceptions is based on empirical survey data, there is a risk of researcher-instilled subjectivity to item selection. Moreover, cultural differences can be explored, as our results suggest slight perception differences between participants from the United Kingdom and the United States. A final limitation of the study’s design is its inability to draw unbiased and causal inferences from the conducted mediation analyses (see Gerber and Green (2012), chapter 10 for a thorough discussion of mediation analyses). The associations between perceptions and candidate evaluations could be confounded as perceptions and evaluations were not experimentally manipulated. We cannot exclude the possibility that additional, unobserved perceptions were in play. Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s0014802501071-w. Acknowledgements The authors would like to thank the anonymous referees for helpful comments and suggestions. We thank the editor, Alfonso Flores-Lagunes, as well as the two anonymous reviewers for their valuable and constructive feedback. Data availability The dataset generated and analysed during the current study are available from the corresponding author on reasonable request. P.Sterkens et al. 16 Page 36 of 40 Declarations Conflict of interest The authors declare no competing interests. General data protection Data processing is organised in line with Ghent University’s code of conduct and therefore adheres to the General Data Protection Regulation (GDPR) standards. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/. References Acquisti A, Fong C (2020) An experiment in hiring discrimination via online social networks. Manage Sci 66(3):1005–1024. https:// doi. org/ 10. 1287/ mnsc. 2018. 3269 Adamczyk A, Pitt C (2009) Shaping attitudes about homosexuality: the role of religion and cultural context. Soc Sci Res 38(2):338–351. https:// doi. org/ 10. 1016/j. ssres earch. 2009. 01. 002 Ahmed AM, Andersson L, Hammarstedt M (2013) Are gay men and lesbians discriminated against in the hiring process? South Econ J 79(3):565–585. https:// doi. org/ 10. 4284/ 003840382011. 317 Arena DF, Jones KP (2017) To “B” or not to “B”: assessing the disclosure of dilemma of bisexual individuals at work. J Vocat Behav 103(A):86–98. https:// doi. org/ 10. 1016/j. jvb. 2017. 08. 009 Arrow KJ (1973) The theory of discrimination. Princeton University, Princeton Auspurg K, Hinz T (2014) Factorial survey experiments. SAGE Publications Ltd., New York Badgett MV (1995) The wage effects of sexual orientation discrimination. ILR Rev 48(4):726–739. https:// doi. org/ 10. 1177/ 00197 93995 04800 408 Badgett MV, Carpenter CS, Sansone D (2021) LGBTQ economics. J Econ Perspect 35(2):141–170. https:// doi. org/ 10. 1257/ jep. 35.2. 141 Badgett ML, Carpenter CS, Lee MJ, Sansone D (2024) A review of the economics of sexual orientation and gender identity. J Econ Lit 62(3):948–994 Baert S (2014) Career lesbians. Getting hired for not having kids? Ind Relat J 45(6):543–561. https:// doi. org/ 10. 1111/ irj. 12078 Baert S (2018a) Hiring a gay man, taking a risk? A lab experiment on employment discrimination and risk aversion. J Homosex 65(8):1015–1031. https:// doi. org/ 10. 1080/ 00918 369. 2017. 13649 50 Baert S (2018b) Facebook profile picture appearance affects recruiters’ first hiring decisions. New Media Soc 20(3):1220–1239. https:// doi. org/ 10. 1177/ 14614 44816 687294 Baert S, De Pauw AS (2014) Is ethnic discrimination due to distaste or statistics? Econ Lett 125(2):270– 273. https:// doi. org/ 10. 1016/j. econl et. 2014. 09. 020 Bailey J, Wallace M, Wright B (2013) Are gay men and lesbians discriminated against when applying for jobs? A four-city, internet-based field experiment. J Homosex 60(6):873–894. https:// doi. org/ 10. 1080/ 00918 369. 2013. 774860 Banaji M (2002) The social psychology of stereotypes. In: Smelser N, Baltes P (eds) International Encyclopedia of the Social and Behavioral Sciences. Pergamon, New York, pp 15100–15104 Baron RM, Kenny DA (1986) The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol 51(6):1173–1182. https:// doi. org/ 10. 1037/ 00223514. 51.6. 117n Sexual orientation stereotypes andjob candidate screening:… Page 37 of 40 16 Barron LG (2009) Sexual orientation employment anti-discrimination legislation and hiring discrimination and prejudice. Acad Manag Proc 1:1–6. https:// doi. org/ 10. 5465/ ambpp. 2009. 44243 452Accessed 15 November 2020 Barron LG, Hebl M (2013) The force of law: the effects of sexual orientation antidiscrimination legislation on interpersonal discrimination in employment. Psychol Public Policy Law 19(2):191–205. https:// doi. org/ 10. 1037/ a0028 350 Baumle AK, Compton DL, Poston DL Jr (2009) Same-sex partners: the social demography of sexual orientation. Suny Press, Albany Becker GS (1957) The economics of discrimination. The University of Chicago Press, Chicago Bello JD (2004) Attractiveness as hiring criteria: savvy business practice or racial discrimination. J Gend Race Justice 8:483–496 Beretvas SN, Meyers JL, Leite WL (2002) A reliability generalization study of the Marlowe Crowne social desirability scale. Educ Psychol Measur 62(4):570–589. https:// doi. org/ 10. 1177/ 00131 64402 06200 4003 Berger RM, Kelly JJ (1981) Do social work agencies discriminate against homosexual job applicants? Soc Work 26(3):193–198. https:// doi. org/ 10. 1093/ sw/ 26.3. 193 Bertrand M, Duflo E (2017) Field experiments on discrimination. Handb Econ Field Exp 1:309–393. https:// doi. org/ 10. 1016/ bs. hefe. 2016. 08. 004 Bethlehem J, Biffignandi S (2012) Handbook of web surveys. Wiley & Sons, Hoboken Binder K, Ward LM (2016) Can heterosexist music cause hiring discrimination against sexual minority men? Testing the effects of prejudicial media messages. J Sex Res 53(6):666–677. https:// doi. org/ 10. 1080/ 00224 499. 2015. 10136 02 Blais AR, Weber EU (2006) A domain-specific risk-taking (DOSPERT) scale for adult populations. Judgm Decis Mak 1(1):33–47. https:// doi. org/ 10. 1037/ t13084000 Bodvarsson ÖB, Partridge MD (2001) A supply and demand model of co-worker, employer and customer discrimination. Labour Econ 8(3):389–416. https:// doi. org/ 10. 1016/ S09275371(00) 000294Accessed 15 November 2020 Bohren JA, Haggag K, Imas A, Pope DG (2019) Inaccurate statistical discrimination: an identification problem. NBER Working papers w25935. Burn I (2018) Not all laws are created equal: legal differences in state non-discrimination laws and the impact of LGBT employment protections. J Lab Res 39(4):462–497. https:// doi. org/ 10. 1007/ s121220189272-0 Burn I (2020) The relationship between prejudice and wage penalties for gay men in the United States. ILR Rev 73(3):650–675. https:// doi. org/ 10. 1177/ 00197 93919 864891 Burn I, Martell M (2020) The role of work values and characteristics in the human capital investment by gays and lesbians. Educ Econ 28(4):351–369. https:// doi. org/ 10. 1080/ 09645 292. 2020. 17580 39 Campbell MR, Brauer M (2021) Is discrimination widespread? Testing assumptions about bias on a university campus. J Exp Psychol Gen 150(4):756–777. https:// doi. org/ 10. 1037/ xge00 00983Accessed 16 November 2020 Campello M, Kankanhalli G, Muthukrishnan P (2020) Corporate hiring under COVID-19: labor market concentration, downskilling, and income inequality. NBER Working papers w27208. https:// doi. org/ 10. 3386/ w27208 Census Bureau (2022) Frequently occurring surnames from the Census 2000. https:// www. census. gov/ topics/ popul ation/ genea logy/ data/ 2000_ surna mes. html. Accessed 10 Aug 2023 Chonody JM (2013) Measuring sexual prejudice against gay men and lesbian women: development of the Sexual Prejudice Scale (SPS). J Homosex 60(6):895–926. https:// doi. org/ 10. 1080/ 00918 369. 2013. 774863 Comrey AL, Lee HB (1992) A first course in factor analysis, 2nd edn. Lawrence Erlbaum, Hillsdale, NJ Crandall CS, Eshleman A (2003) A justification-suppression model of the expression and experience of prejudice. Psychol Bull 129(3):414–446. https:// doi. org/ 10. 1037/ 00332909. 129.3. 414 Derous E, Ryan AM (2019) When your resume is (not) turning you down: modelling ethnic bias in resume screening. Hum Resour Manag J 29(2):113–130. https:// doi. org/ 10. 1111/ 17488583. 12217 Di Stasio V (2014) Education as a signal of trainability: results from a vignette study with Italian employers. Eur Sociol Rev 30(6):796–809. https:// doi. org/ 10. 1093/ esr/ jcu074 Dokko G, Wilk SL, Rothbard NP (2009) Unpacking prior experience: how career history affects job performance. Organ Sci 20(1):51–68. https:// doi. org/ 10. 1287/ orsc. 1080. 0357 Drydakis N (2009) Sexual orientation discrimination in the labour market. Labour Econ 16(4):364–372. https:// doi. org/ 10. 1016/j. labeco. 2008. 12. 003 P.Sterkens et al. 16 Page 38 of 40 Drydakis N (2015) Sexual orientation discrimination in the United Kingdom’s labour market: a field experiment. Human Relations 68(11):1769–1796. https:// doi. org/ 10. 1177/ 00187 26715 569855 Drydakis N (2021) Sexual orientation and earnings: a meta-analysis 2012–2020. J Popul Econ 35(2):409– 440. https:// doi. org/ 10. 1007/ s0014802100862-1 Etaugh C, Malstrom J (1981) The effect of marital status on person perception. J Marriage Fam 43(4):801–805. https:// doi. org/ 10. 2307/ 351337 Everly BA, Unzueta MM, Shih MJ (2016) Can being gay provide a boost in the hiring process? Maybe if the boss is female. J Bus Psychol 31(2):293–306. https:// doi. org/ 10. 1007/ s108690159412-y Fiske ST (1998) Stereotyping, prejudice, and discrimination. In: Gilbert DT, Fiske ST, Lindzey G (eds) The handbook of social psychology. McGraw-Hill, Boston, pp 357–414 Fiske ST, Cuddy AJ, Glick P (2007) Universal dimensions of social cognition: warmth and competence. Trends Cogn Sci 11(2):77–83. https:// doi. org/ 10. 1016/j. tics. 2006. 11. 005 Flage A (2020) Discrimination against gays and lesbians in hiring decisions: a meta-analysis. Int J Manpower 41(6):671–691. https:// doi. org/ 10. 1108/ IJM0820180239 Fric K (2017) Access to the labour market for gays and lesbians: research review. J Gay Lesbian Soc Serv 29(4):319–361. https:// doi. org/ 10. 1080/ 10538 720. 2017. 13656 71 Gerber AS, Green DP (2012) Field experiments: design, analysis, and interpretation. WW Norton, New York Hammarstedt M, Ahmed A, Andersson L (2015) Sexual prejudice and labor market outcomes for gays and lesbians: evidence from Sweden. Fem Econ 21(1):90–109. https:// doi. org/ 10. 1080/ 13545 701. 2014. 937727 Hayes AF (2017) Introduction to mediation, moderation, and conditional process analysis: a regressionbased approach. Guilford publications, New York Hebl MR, Foster JB, Mannix LM, Dovidio JF (2002) Formal and interpersonal discrimination: a field study of bias toward homosexual applicants. Pers Soc Psychol Bull 28(6):815–825. https:// doi. org/ 10. 1177/ 01461 67202 289010 Horvath M, Ryan AM (2003) Antecedents and potential moderators of the relationship between attitudes and hiring discrimination on the basis of sexual orientation. Sex Roles 48(3–4):115–130. https:// doi. org/ 10. 1023/a: 10224 99121 222 Jepsen C, Jepsen LK (2015) Labor-market specialization within same-sex and difference sex couples. J Econ Soc 54(1):109–130. https:// doi. org/ 10. 1111/ irel. 12078 Jones JM (2024) LGBTQ+ Identification in U.S. Now at 7.6%. https:// news. gallup. com/ poll/ 611864/ lgbtqident ifica tion. aspx. Accessed 1 June 2024 Judd CM, Park B (1993) Definition and assessment of accuracy in social stereotypes. Psychol Rev 100(1):109–128. https:// doi. org/ 10. 1037/ 0033295X. 100.1. 109 Katsarova I (2019) Gender equality in sport: getting closer every day. Eur Parliam Res Serv Rep, PE 635:560 Laurent T, Mihoubi F (2017) Sexual orientation, unemployment and participation: are gays less employable than straights? J Lab Res 38(1):1–44. https:// doi. org/ 10. 1007/ s121220169237-0 Lippens L, Vermeiren S, Baert S (2023) The state of hiring discrimination: a meta-analysis of (almost) all recent correspondence experiments. Eur Econ Rev 151:104315. https:// doi. org/ 10. 1016/j. euroe corev. 2022. 104315 Liu H, Wilkinson L (2017) Marital status and perceived discrimination among transgender people. J Marriage Fam 79(5):1295–1313. https:// doi. org/ 10. 1111/ jomf. 12424 Looney D, Lusin N (2019) Enrollments in languages other than English in United States institutions of higher education, Summer 2016 and Fall 2016: final report. Modern Language Association. https:// www. mla. org/ conte nt/ downl oad/ 110154/ 24069 32/ 2016Enrol lmentsFinalReport. pdf. Accessed 26 July 2021 McCutcheon J, Morrison MA (2015) The effect of parental gender roles on st’dents’ attitudes toward lesbian, gay, and heterosexual adoptive couples. Adopt Q 18(2):138–167. https:// doi. org/ 10. 1080/ 10926 755. 2014. 945702 Mishel E (2020) Contextual prejudice: how occupational context and stereotypes shape bias against gay and lesbian employees. Social Currents 7(4):371–391. https:// doi. org/ 10. 1177/ 23294 96520 919912 Mize TD (2016) Sexual orientation in the labor market. Am Sociol Rev 81(6):1132–1160. https:// doi. org/ 10. 1177/ 23294 96518 761999 Morrison TG, Parriag AV, Morrison MA (1999) The psychometric properties of the homonegativity scale. J Homosex 37(4):111–126. https:// doi. org/ 10. 1300/ j082v 37n04_ 07 Nadler JT, Kufahl KM (2014) Marital status, gender, and sexual orientation: implications for employment hiring decisions. Psychol Sex Orientat Gend Divers 1(3):270–278. https:// doi. org/ 10. 1037/ sgd00 00050 Sexual orientation stereotypes andjob candidate screening:… Page 39 of 40 16 Neumark D (2018) Experimental research on labor market discrimination. J Econ Lit 56(3):799–866. https:// doi. org/ 10. 1257/ jel. 20161 309 Niedlich C, Steffens MC (2015) On the interplay of (positive) stereotypes and prejudice: impressions of lesbian and gay applicants for leadership positions. Sensoria: J Mind, Brain Cult 11(1):70–80. https:// doi. org/ 10. 7790/ sa. v11i1. 408 Niedlich C, Kachel S, Steffens MC (2022) Sexual orientation information and hiring: can individualizing information lead to negative stereotyping of sexual minority group members. J Appl Soc Psychol 52(5):287–304. https:// doi. org/ 10. 1111/ jasp. 12859 OECD (2020) Over the rainbow? The road to LGBTI inclusion. https:// www. oecd. org/ social/ overtherainb owtheroadtolgbtiinclu sion8d2fd 1a8en. htm. Accessed 26 July 2021 Office for National Statistics (2021) Occupations of those in employment, by age and sex, England and Wales, Census 2021. https:// www. ons. gov. uk/ emplo yment andla bourm arket/ peopl einwo rk/ emplo yment andem ploye etypes/ datas ets/ occup ation soft h osein emplo yment byage andse xengl andan dwale scens us2021. Accessed 10 Aug 2023 Office for National Statistics (2022) Baby names in England and Wales: from 1996. https:// www. ons. gov. uk/ peopl epopu latio nandc ommun ity/ birth sdeat hsand marri ages/ liveb irths/ datas ets/ babyn amesi nengl andan dwale sfrom 1996. Accessed 10 Aug 2023 Office for National Statistics (2023) Sexual orientation, UK: 2021 and 2022. https:// www. ons. gov. uk/ peopl epopu latio nandc ommun ity/ cultu ralid entity/ sexua lity/ bulle tins/ sexua liden tityuk/ 2021a nd2022. Accessed 1 June 2024 Oreopoulos P (2011) Why do skilled immigrants struggle in the labor market? A field experiment with thirteen thousand resumes. Am Econ J Econ Pol 3(4):148–171. https:// doi. org/ 10. 1257/ pol.3. 4. 148 Parnell MK, Lease SH, Green ML (2012) Perceived career barriers for gay, lesbian, and bisexual individuals. J Career Dev 39(3):248–268. https:// doi. org/ 10. 1177/ 08948 45310 386730 Patacchini E, Ragusa G, Zenou Y (2015) Unexplored dimensions of discrimination in Europe: homosexuality and physical appearance. J Popul Econ 28(4):1045–1073. https:// doi. org/ 10. 1007/ s001480140533-9 Pedulla DS (2014) The positive consequences of negative stereotypes: race, sexual orientation, and the job application process. Soc Psychol Q 77(1):75–94. https:// doi. org/ 10. 1177/ 01902 72513 506229 Pellegrini V, De Cristofaro V, Giacomantonio M, Salvati M (2020) Why are gay leaders perceived as ineffective? The role of the type of organization, sexual prejudice and gender stereotypes. Personality Individ Differ 157:109817. https:// doi. org/ 10. 1016/j. paid. 2020. 109817 Peracchi F, Rossetti C (2012) Heterogeneity in health responses and anchoring vignettes. Empir Econ 42:513–538. https:// doi. org/ 10. 1007/ s001810110530-8 Phelps ES (1972) The statistical theory of racism and sexism. Am Econ Rev 62(4):659–661. https:// www. jstor. org/ stable/ 18061 07. Accessed 26 November 2020 Pichler S, Varma A, Bruce T (2010) Heterosexism in employment decisions: the role of job misfit. J Appl Soc Psychol 40(10):2527–2555. https:// doi. org/ 10. 1111/j. 15591816. 2010. 00669.x Reynolds WM (1982) Development of reliable and valid short forms of the Marlowe-Crowne social desirability scale. J Clin Psychol 38(1):119–125. https:// doi. org/ 10. 1002/ 10974679(198201) 38: 1< 119:: aidjclp2 27038 0118>3. 0. co;2-i Rowthorn R (2002) Marriage as a signal. The law and economics of marriage and divorce 132–156. https:// doi. org/ 10. 1017/ CBO97 80511 495328. 008 Schulte-Hostedde AI, Eys MA, Emond M, Buzdon M (2012) Sport participation influences perceptions of mate characteristics. Evol Psychol 10(1):78–94. https:// doi. org/ 10. 1177/ 14747 04912 01000 109 Singletary SL, Hebl MR (2009) Compensatory strategies for reducing interpersonal discrimination: the effectiveness of acknowledgments, increased positivity, and individuating information. J Appl Psychol 94(3):797–805. https:// doi. org/ 10. 1037/ a0014 185 Social Security Administration (2023) Popularity of a name https:// www. ssa. gov/ oact/ babyn ames/. Accessed 10 Aug 2023 Steenkamp JBE, De Jong MG, Baumgartner H (2010) Socially desirable response tendencies in survey research. J Mark Res 47(2):199–214. https:// doi. org/ 10. 1509/ jmkr. 47.2. 199 Sterkens P, Baert S, Rooman C, Derous E (2021) As if it weren’t hard enough already: breaking down hiring discrimination following burnout. Econ Hum Biol 43:101050. https:// doi. org/ 10. 1016/j. ehb. 2021. 101050 Tilcsik A (2011) Pride and prejudice: employment discrimination against openly gay men in the United States. Am J Sociol 117(2):586–626. https:// doi. org/ 10. 1086/ 661653 P.Sterkens et al. 16 Page 40 of 40 Tilcsik A, Anteby M, Knight CR (2015) Concealable stigma and occupational segregation: toward a theory of gay and lesbian occupations. Adm Sci Q 60(3):446–481. https:// doi. org/ 10. 1177/ 00018 39215 576401 Treffers T, Ritter AC, Born N, Welpe I (2024) A systematic review of experimental evidence on interventions against bias and discrimination in organizations. Hum Resour Manag Rev 34(3):101029. https:// doi. org/ 10. 1016/j. hrmr. 2024. 101029 Umphress EE, Simmons AL, Boswell WR, Triana MDC (2008) Managing discrimination in selection: the influence of directives from an authority and social dominance orientation. J Appl Psychol 93(5):982–993. https:// doi. org/ 10. 1037/ 00219010. 93.5. 982 United States Department of Labor (2021) Labor force statistics from the current population survey. https:// www. bls. gov/ cps/ cpsaa t11. htm. Accessed 26 July 2021 Van Borm H, Baert S (2018) What drives hiring discrimination against transgenders? Int J Manpow 39(4):581–599. https:// doi. org/ 10. 1108/ ijm0920170233 Van Hoye G, Lievens F (2003) The effects of sexual orientation on hirability ratings: an experimental study. J Bus Psychol 18(1):15–30. https:// doi. org/ 10. 1023/a: 10250 78819 951 Van Belle E, Di Stasio V, Caers R, De Couck M, Baert S (2018) Why are employers put off by long spells of unemployment? Eur Sociol Rev 34(6):694–710. https:// doi. org/ 10. 1093/ esr/ jcy039 Van Borm H, Burn I, Baert S (2021) What does a job candidate’s age signal to employers? Labour Econ 71:102003. https:// doi. org/ 10. 1016/j. labeco. 2021. 102003 Weichselbaumer D (2015) Testing for discrimination against status: a field experiment. Ind Relat 54(1):131–161. https:// doi. org/ 10. 1111/ irel. 12079 West K, Hewstone M (2012) Culture and contact in the promotion and reduction of anti-gay prejudice: evidence from Jamaica and Britain. J Homosex 59(1):44–66. https:// doi. org/ 10. 1080/ 00918 369. 2011. 614907 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Authors and Affiliations PhilippeSterkens1 · AxanaDalle1 · JoeyWuyts1 · InesPauwels1· HellenDurinck1· StijnBaert1,2,3,4,5 * Axana Dalle [email protected] Philippe Sterkens Philippe.Sterk[email protected] Joey Wuyts Joey.W[email protected] Ines Pauwels Ines.Pauw[email protected] Hellen Durinck [email protected] Stijn Baert Stijn.Baer[email protected] 1 Ghent University, Sint-Pietersplein 6, 9000Ghent, Belgium 2 University ofAntwerp, Antwerp, Belgium 3 Université Catholique de Louvain, Louvain, Belgium 4 Institute forLabor Economics, Bonn, Germany 5 Global Labor Office, Essen, Germany