Recent advances in rights for lesbians, gay men, and bisexual individuals (LGB) have varied substantially across the world. In the US, for example, LGB rights have increased at a rapid pace: same-sex sexual activity was fully decriminalised in 2003, legal access to same-sex marriage was granted in 2015, and employment discrimination protections were granted in 2020. Likewise, India decriminalised same-sex sexual acts in 2018, while Taiwan granted same-sex marriage in 2019. Yet, in many other parts of the world, LGB rights have advanced more slowly or not at all. As of 2019, 70 United Nations Member States (35% of all members) still criminalise same-sex sexual conduct. In six UN Member States, same-sex sexual activity is even punishable by death. Anti-LGB attitudes are particularly strong in Africa, the Middle East, and Eastern Europe, and anti-LGB policies have recently been adopted in Hungary, Poland, Russia, Tanzania, and Uganda (Mendos 2019).
In a new paper (Aksoy et al. 2022), we provide new evidence on the determinants of support for sexual minorities in Serbia, Turkey, and Ukraine – three emerging markets with some of the lowest rates of social acceptance of sexual minorities in Europe. Figure 1 shows for 33 countries the share of respondents who agree that gay men and lesbians should be free to live their lives as they wish. Serbia, Ukraine, and Turkey have the 29th, 30th, and 31st lowest shares of agreement, respectively. These countries also have highly restrictive LGBT equality laws and policies. They score just 33, 4, and 18, respectively, on a scale where zero indicates gross human rights violations and 100 represents the greatest degree of legal equality.
Figure 1 Support for sexual minorities across countries
Source: European Social Survey, 2002-2018
Simple information treatments to reduce discrimination
To investigate whether and how LGB discrimination can be reduced, even in such strongly homophobic environments, we designed an information-treatment experiment that tests several theories. First, we are interested in whether rational economic self-interest might overcome personal distaste for LGB people. Thus, in one arm of our experiment we inform people about the direct economic costs to their country from discrimination against sexual minorities, using estimates of these per-capita income changes from Badgett et al. (2019). We hypothesise that this information induces some self-interested individuals to set aside negative personal views to support LGB employment non-discrimination. Second, we want to understand whether narratives about homosexuality being a mental illness drive anti-gay sentiment. In another treatment arm we therefore try to ‘debunk’ this narrative by informing individuals that the World Health Organization (WHO) does not consider homosexuality to be a mental illness. We hypothesise that this information induces more favourable views about homosexuality. We test these hypotheses through a randomised survey experiment in which a third of respondents receive the ‘discrimination cost’ information, another third receive the ‘myth debunking’ information, and the final third receive placebo information unrelated to LGB people.
Figure 2 Effects of information treatments on support for equal employment opportunities (left panel) and LGB-related views in non-economic domains (right panel)
Source: Aksoy et al. (2022).
Notes: Odds ratios are displayed. All models include controls for age and its square; a male dummy, dummy variables for tertiary education, secondary education; a dummy for being in any kind of partnership; a dummy for living in an urban area; survey date dummies; survey country dummies; number of adults above and below 65; religion dummies (Catholic, Muslim, Orthodox, other religion) and labour market-related controls (whether the individual works at a state-owned enterprise, foreign firm or international organization, and unemployed dummy).
Our experiment yields four main results. First, providing information about the economic cost from sexual orientation discrimination significantly increases support for measures to safeguard equal employment opportunities for lesbians and gays. Individuals who received the discrimination cost treatment were 1.49 times more likely to support such equal opportunities compared with individuals randomly assigned to the control group.
Second, we find that this discrimination cost treatment spills over to support for equal employment opportunities based on ethnic origin, religious beliefs, nationality, gender, and disability (Panel A of Figure 2). However, each of the discrimination cost treatment effects in these other domains is quantitatively smaller than the effect for sexual orientation-based employment equality – although they are all statistically significant.
Third, the impact of the discrimination cost treatment does not spill over to LGB support in other aspects of life. After adjusting for false discovery rates, there are no effects on opinions concerning the moral acceptability and justifiability of homosexuality, as well as whether sexual minorities should be able to live their lives freely, or whether sexual minorities bring shame on their families (Panel B of Figure 2).
Fourth, informing people that according to the WHO homosexuality is not a mental disease does not cause more support for equal employment opportunities, but does result in improved attitudes about sexual minorities in non-economic aspects of life. Specifically, this myth-debunking treatment increases support regarding the moral acceptability and justifiability of homosexuality and the idea that sexual minorities should be able to live their lives freely. It also reduces the likelihood that individuals report that a gay or lesbian relative would bring shame on their family. Interestingly, these effects are concentrated among those individuals who actually trust the WHO.
Our results have two important implications for the expansion of LGB rights in parts of the world where anti-LGB attitudes are widely held and deeply ingrained. First, they clearly suggest that individuals in countries with strong views about the immorality of homosexuality can – when informed about the economic costs of sexual-orientation discrimination – still voice support for non-discrimination policies. This indicates that advances in LGB rights in socially conservative places may be more effective if they appeal to the economic costs of anti-LGB discrimination instead of trying to change the underlying views themselves. Second, our results also indicate that even views about the acceptability of homosexuality itself can be modestly affected by the provision of basic information, particularly when framed in the context of institutions that people trust.
Aksoy, C G, C S Carpenter, R De Haas, M Dolls and L Windsteiger (2022), “Reducing Sexual-Orientation Discrimination: Experimental Evidence from Basic Information Treatments”, CEPR Discussion Paper No. 16852.
Badgett, M V L, K Waaldijk, and Y Van der Meulen Rodgers (2019), “The Relationship between LGBT Inclusion and Economic Development: Macro-level Evidence”, World Development 120: 1-14.