Excerpt
Table of Content
1. Introduction
2. Theoretical Background
Differences in earnings
Experiments on discrimination against homosexual men and lesbians
3. Experimental Design
Method
Hypothesis
Cities
Occupations
Application documents
Procedure
Evaluation of the results
4. Conclusion
5. References
1. Introduction
Since the first of October 2017, same-sex couples in Germany have the right to marry. Until this point in time, homosexual couples could only enter into a registered partnership with the same duties, but not the same rights like heterosexual married couples. This significant development in law may introduce new problems and challenges. Ann Ferguson (2007), for one, criticises that homosexual marriage could shift social hierarchies, which in return creates a new differentiation between homosexual people who live in a single household and the more socially accepted married homosexual-couples. Butler (2004) argued that homosexual marriage could be understood as a way to push homosexuals into a normalized lifestyle – for example marriage, a monogamous family life, and also successful careers, which could increase social acceptance.
In this sense, this paper is building on the work by Weichselbaumer (2013) who examined the difference in economic outcome between single lesbians and lesbians in a registered partnership in Munich and Berlin. She found no differences between single and partnered lesbians in Munich – both groups are treated unfavourable in comparison to heterosexual women. In Berlin, she found out that there is no discrimination based on sexual orientation. But some studies showed that negative attitudes towards homosexual men are more present than towards lesbians1, and that employers may have preferences for male employees with one steady married partner2. If this is the case, homosexual men may benefit more from signalling that they live in a marriage than lesbian women would.
Therefore, this paper sets forth the study of Weichselbaumer (2013) by focusing on homosexual men instead of lesbians, and on the other hand, by measuring the effect after the legislation of same-sex marriage in Germany. The research question is if homosexual married men are favoured in the hiring process over homosexual single men in the German labour market, and secondly if this “marriage-effect” is stronger for homosexual men than for heterosexual men. To be able to answer this question, a Correspondence Test has to be conducted. Four fictional applications of males who vary in their sexual orientation (heterosexual versus homosexual) and marital status (married versus single) will be sent to vacancies in the sixteen German provincial capitals and sixteen rural regions in Germany.
2. Theoretical Background
Discrimination in connection with work means that “individual workers who have identical productive characteristics are treated differently because of the demographic groups to which they belong” (Ehrenberg and Smith, 1994, p. 402). Different forms of discrimination can be distinguished: statistical discrimination means that “decisions are based on group belonging and not on the individual’s skills” (Arai et. al. 2000: p. 9). “Preference-based discrimination means that the (potential) employer or others in the employing organization simply dislike a particular population subgroup and therefore try to avoid it” (Brekke, I., & Mastekaasa, A. 2008, p. 4). Finally, implicit discrimination is defined as “unconscious mental associations between a target and a given attribute” (Bertrand, M., Chugh, D., & Mullainathan 2005, p. 1).
Differences in earnings
In studies that analysed available data without controlling for explanatory factors, lesbians earn more than heterosexual women and homosexual men earn less than heterosexual men (e.g. Cushing-Daniels & Yeung, 2009). This could be interpreted in a way that homosexual men are subjected to discrimination while lesbians are not. The household incomes of couples showed that lesbian couples have lower incomes than married heterosexual couples, while couples of homosexual men have similar household incomes to that of male married heterosexual couples. In contrast, the individual income of homosexual men is lower than that of heterosexual married men (cf. Klawitter, 2011). Possible central statements of the above mentioned results is that homosexual single men are more subjected to discrimination than married homosexual men, and that married homosexual men may profit more from signalling that they live in a marriage compared to heterosexual men. Other studies used multivariate data analysis to indicate differences in wages for different control factors (e.g. education, experience, occupation, industry, geographic location). All of these studies suggest that homosexual men earn less than equally qualified heterosexual men. For lesbians, they found out that they earn more than heterosexual women. (e.g. Badgett, 1995; Klawitter & Flatt, 1998; Allegretto & Arthur, 2001; Berg & Lien, 2002; Black et al., 2003; Blandford, 2003; Kroh et al., 2017).
Experiments on discrimination against homosexual men and lesbians
This part of the empirical background conducts correspondence tests to examine sexual orientation discrimination in the hiring process. For this, applications from individuals with identical qualifications but different demographic characteristics were sent out to multiple firms. They defined discrimination as the fact that one applicant was invited to interviews more often than the other. The results of these correspondence tests showed that homosexual people are less successful in the hiring process than heterosexual people (e.g. Adams, 1981; Weichselbaumer, 2003; Drydakis, 2009; Tilcsik, 2011). For example, Tilcsik (2011) found a statistically significant difference in response rates in the states of Florida, Ohio and Texas favouring the heterosexual applicant (11.5% compared to 7.2%). He found no differences in California, Nevada, Pennsylvania and New York. Ahmed, Andersson and Hammarstedt (2013) examined discrimination against homosexual men and lesbians in Sweden. Weichselbaumer (2013) discovered an equal effect between single and married lesbians in Germany. Both studies sent only one random selected application to one job offering of firms. Weichselbaumer (2013) found statistically significant differences between the call-back rates of lesbians (single and partnered) and heterosexual women in Munich but not in Berlin (cf. Weichselbaumer, 2013, pp. 16-19). Ahmed et al. (2013) found no statistically significant differences in the response rate between homosexual and heterosexual applicants within ten different professional groups. The only exception was for lesbians in cleaning jobs. Furthermore, they observed the effect of a less favourable treatment for lesbians in female-dominated jobs and of homosexual men in male-dominated jobs.
To explain the pay gaps and the findings of discrimination in the hiring process, various possible reasons were pointed out. The studies by LaMar and Kite (1998) and by Herek (2002) showed that negative attitudes towards homosexual men are more present than towards lesbians. One prejudice against homosexual men (but not lesbians) is their promiscuous sex life, which leads to the interpretation that gay employees are unsteady and irresponsible. Researches who examined the effect of male marriage premium suggest that employers may have preferences for male employees with a long-term married partner (e.g. Cohen, 2002; Hersch and Stratton, 2000). The term “marriage premium” describes “the earning advantages that married men enjoy over never-married men” (Cohen, 2002, p. 1). This means that single homosexual men don’t profit of this marriage premium and earn less than married heterosexual men, while homosexual men in a same-sex households gain the marriage premium. Lesbians in contrast cannot earn a marriage premium or an income transfer from their husbands (cf. Becker, 1991), and therefore it is necessary to put more effort into their work. This explains their higher earnings, contrary to heterosexual women.
3. Experimental Design
Method
Similar to Weichselbaumer (2013), I will use an experimental methodology called “Correspondence Testing”. For this, four different applications should be created: an application for heterosexual single men, an application for heterosexual married men, an application for homosexual single men, and an application for homosexual married men. All these applications match in all characteristics such as age, schooling and job experience, and differ only in sexual orientation and marital status. If an applicant receives an answer and gets invited to an interview, while the other is not, this is defined as discrimination. There is no differentiation between statistical, preference-based and implicit discrimination, because with this correspondence test we can only test if there is discrimination and not which kind of discrimination. For this, further studies are needed.
The applications are constructed as follows: both, the single heterosexual and the single homosexual man, communicate their single life under the information “family status” in the CV as “single”. To signal the sexual orientation, the homosexual man worked in the Lesbian and Gay Federation in Germany (LSVD) as a volunteer and performed accounting tasks. The LSVD is the largest homosexual organisation in Germany and is located in the political mainstream (cf. Weichselbaumer, 2013, p. 10). Choosing this organization may minimize criticism like “political activism or radicalism with sexual orientation” (cf. Weichselbaumer, 2013, p. 13). The single heterosexual man worked in a cultural centre with the same tasks (cf. Weichselbaumer, 2003, 2013; Patacchini et al., 2012). This activity represent the control condition for this experiment. The two married men with different sexual orientation are constructed as follows: The married heterosexual man is married to a woman and the married homosexual man is married to a man. This information is given under “family status”. Both men worked in a non-profit organisation, the homosexual man at LSVD and the heterosexual man in the cultural organisation. This is in contrast to the study by Weichselbaumer (2013). In her experiment, the two married women (hetero- and homosexual) worked in the cultural organisation for comparative purposes. My special attention is the comparison between the two homosexual men, which is why I changed this condition.
Hypothesis
The literature background shows that same-sex marriage can be understood as a signal to push homosexual people into a normalized lifestyle. Accordingly, same-sex marriage leads to an increased of social acceptance and to a minimization of discrimination against homosexual people. I deduce the hypothesis that homosexual men who live in a same-sex marriage are favoured in the hiring process over homosexual men who are not. As a result, in the following experiment they should be more successful than those who are single. Furthermore, the theoretical background shows that married homosexual men may profit more from signalling that they are in a marriage than heterosexual men. Therefore, I additionally deduce the hypothesis that heterosexual married men are favoured over single heterosexual men in the hiring process. This effect is stronger for homosexual men.
Cities
The experiment should be conducted in all sixteen federal states of Germany. On the one hand in their capitals to show if there is discrimination in the big cities of Germany, in which federal states and to what extent. Furthermore, I will perform the experiment in sixteen rural regions in Germany – one rural region in each federal state. This will identify whether there are differences in the results between cities and rural regions.
Occupations
Similar to the study by Weichselbaumer (2013), I select job opportunities for office jobs, especially as secretaries, clerical assistants and accountants.
Application documents
The application consists of the following application documents: a motivation letter, a curriculum vitae, a fake A-level report, and a fake training certificate for the apprenticeship as a office clerk (cf. Weichselbaumer, 2013, p. 17). The fictitious applicants were still employed in their first job, so that no reference letter is needed. In contrast to the study by Weichselbaumer (2013), my application documents do not include an application photo because the attractivity of an applicant has a significant direct impact on the hiring decision (Binckli (2014)).
[...]
1 Cf. LaMar and Kite (1998) and Herek (2002)
2 Cf. e.g. Cohen (2002) & Hersch and Stratton (2000)