Friends of Dorothy

The economics of being gay

i’m not sure what i’m going for here

November is actually pride month in Argentina¹, which means that after I already did a pride-related thing in June, about woke companies, I’ll continue promoting cultural marxism: what’s being gay like, economics wise?

Discrimination 101

Last week I already posted about this in some detail, so I’ll have to rehash (blame Argentina’s terrible elections for that), but: what is discrimination about?

Discrimination is more than just having statistically disparate outcomes, since you also have to “prove”, in some sort of way, that they occur in a systematic manner across members of a group. Well, there’s plenty of (empirical, experimental) evidence that discrimination exists:

• A study (by Nobel Laureate Claudia Goldin) found that blind auditions for orchestra positions increased the chances that women get hired. (omg Tár reference????)

• Another study found that, for similar positions and identical resumes, those with stereotypically “white” or “neutral” names got more call-backs than those with “Black” names.

• Evidence from European countries also showed that identical resumes for women with different names (“Muslim” and “local”) and different pictures (none, without a veil, and with a veil); and while “Muslim” names overall were disfavored for subsequent interviews, the discrimination was worse for Muslim women who were wearing veils in their pictures.

• All else being equal, recruiters are more likely to call back for interviews men to women, especially if the women have children.

• People rate the job performance of women as worse than that of men, and mothers as even worse - despite both groups being as productive as men.

How do we make sense of discrimination? The traditional way to make sense of it in an economics context, is what’s known as taste-based discrimination, a model designed by Nobel Laureate Gary Becker. According to this logic, people don’t like individuals from another group, and so when identifying that group is easy or possible, they tend to harm them in a variety of ways (usually labor-related). Assuming people’s preferences for racial/gender/sexual/etc segregation range from “none” to “extreme”, and supposing that each candidate for a job is equal to all others except for their identity, the decision between a (say) straight and gay applicant comes down to the sexuality preferences of the employer - but minority applicants can get an edge by changing their conditions, as in taking lower pay or showing up with a better resume. A study evaluating Becker’s claims found them adequate, but also couldn’t explain three quarters of the impact of racial prejudice. In a concrete example of racial discrimination, Black and Hispanic individuals were subject to systematic discrimination in the rental market, resulting in lower quality housing on average - and for Black prospective homebuyers/renters, they need a pretty significant income advantage versus a White candidate to actually get preference.

Another somewhat related and facially not especially useful model of discrimination is what’s known as statistical discrimination, where absent accurate information on the specific individual, decisions are made based on general prejudices about the group, resulting in biased outcomes. The clearest example of this is the unintended effects of legislation banning verifying criminal history of prospective tenants (i.e. “ban the box” laws), where landlords tend to discriminate against racial minorities assuming they are more likely to have criminal records.

Finally, you have structural discrimination, where disadvantages get baked in and persist for generations, even when the original biases don’t really play much of a role. While the debate about economic disparities is largely about “discrimination versus choice”, the real issue isn’t really that evil bosses are paying marginalized people less or whatever, it’s that there’s a different choice set for those people because they are marginalized and that has economic consequences. Surveys, for instance, show that the main reason why women attended college was to find a (successful) husband, rather than a career of their own. But, at the same time, the returns to getting a husband through an education were higher than the results of getting a job through an education - so we’re in a case of chicken and egg: either the culture led to reduced opportunities for women, or the reduced opportunities for women shaped the culture.

Whitney Chewston.png

LGBT people, as a group face multiple adversities in their economic lives. For instance, the gays earn less than straight people, though this is different for both gay men and women and for cisgender and transgender individuals.

It should be noted, however, that there’s really no way to tell whether there is a reporting bias in LGBT orientation, i.e. that certain subgroups (think, rich or white) are more likely to openly identify as LGBT and would thus bias the data. However, surveys show that the number of households with same-gender partners generally matches up with the number of people who identify as being in an LGBT relationship, meaning that there is an increase in reporting of LGBT status over time, but that this can be accounted for. In an instance of such selection bias, gay people tend to be much younger than the straight or general population, with bisexual people even younger and transgender or nonbinary people the youngest of all, which is not really about how many people are gay per generation, and more about how many people openly identify as such per age group. Gay people tend to live in more accepting or progressive areas (shocker) and have much higher rates of moving than straight people - both because people who are gay move to more accepting areas, and because more accepting areas have higher reporting rates. This dynamic could, then, partly explain why gay male couples have the highest reported earnings of all types of partnerships - without controlling for labor market qualities, gay men make a lot of money, but controlling for them, they make less than equally educated and experienced straight men, which could point to selection in disclosure status.

There is evidence of sexual orientation and gender identity discrimination in the labor market. LGBT people, for instance, have different career choices than straight people (one could see that all the women I met on dating apps have some sort of arts background), which responds both to social pressures and to personal preferences, and face several hurdles at work. Openness about sexual orientation is generally linked to better job outcomes, though experiences of harassment and discrimination are not uncommon. For instance, applicants for jobs who shared their LGBT identity on job applications were less likely to receive callbacks in both the US and the UK.

For transgender people, the situation is especially difficult, and it is difficult in ways that are different than the “LGB” side of things: firstly, outcomes are generally far worse, with much more elevated rates of poverty, unemployment, and all-around hardship than cisgender people. These results are sensitive to the degree to which someone has socially transitioned, i.e. can fully present as the preferred gender: beyond general improvements in mental health and personal experiences, greater ability to socially transition results in a smaller earnings penalty (and can explain why transgender men tend to have worse outcomes than transgender women in some metrics). Additionally, the younger someone is able to transition (which is generally associated with greater ability to “pass” as the preferred gender), the better the labor outcomes are, though a more supportive environment that can result in higher rates of disclosure and more positive experiences. Something that is kind of fascinating and slightly funny is that, across most studies, transgender men (as in, women who become male) tend to experience an earnings boost post-transition, while transgender women tend to have a wage penalty.

The trashcan is gender ideology

A very interesting extension of this is the conclusions to be drawn out of the LGBT pay earnings not regarding LGBT issues specifically, but gender norms more generally.

The results for gay men are that they earn less than equally qualified straight men, but gay male households have generally similar earnings to straight households. This is a bit puzzling, but generally could make sense because: married men tend to experience an earnings boost post-marriage (which is seen as relating to higher job stability). For instance, unmarried gay men earn 16% less than straight married men, but just 2% less than unmarried straight men. In addition, gay men work shorter hours than straight men, which relates back to the fact that men are typically advantaged in their career versus women by taking in longer hours and not taking career breaks (such as maternity leave). One final difference is that, once again, experiences and expectations of discrimination result in different labor market choices: if gay men expect discrimination in certain fields, they may end up abstaining from them, even if those fields command a significantly higher wage. One final factor is that certain types of gay male households (especially those with children) may live in different labor markets or tend to have higher educational attainment than similar straight households, making the wage gap seem artificially lower.

A surprising, opposite trend is seen among lesbians: women in relationships with other women tend to outearn women in relationships with men. This “lesbian wage premium” may seem puzzling, but it is generally straightforward: lesbians tend to work longer hours and take fewer career interruptions than straight women. This has the same explanation that lies at the heart of the gender wage gap: women are expected to do a disproportionate share of the care and domestic work in the household versus their (male) partner. Gay women share the household with their spouses much more equally than straight women, for instance, resulting in fewer penalties to marriage in their workplaces. And gay women tend to have longer commutes than straight women, resulting in higher job satisfaction and, particularly, better matched positions. This effect is actually tempered by the fact that many women married to other women used to be married to men before: among never-married gay women, the “lesbian wage premium” is even higher than in previously married lesbians, who suffered the negative impact of straight marriage on their earnings.

Both of these results point to a weird confluence of facts: that sexually normative family and household formations tend to structurally advantage men and disadvantage women, and that when the people involved are not straight, men suffer while women benefit. This “hegemonic masculinity premium” that straight men profit from can also be linked to other findings, such that women who conform less to traditional feminine expectations have better labor outcomes than more traiditionally gender conforming women in the (notoriously misogynistic) STEM fields.

Conclusion

Officially coming out as woke with this one I guess but discrimination is bad and marriage is also not very good.

1. I went to pride with some friends last Saturday, which is when the parade usually is for pretty straightforward reasons - a different historical context + nicer weather