Association for Public Policy Analysis & Management

No federal law prohibits housing discrimination based on gender identity or sexual orientation. As of 2017, only 22 U.S. states legally prohibit discrimination based on sexual orientation in the housing market. In response to state and federal inaction, over 100 municipalities have passed local housing protections prohibiting discrimination based on sexual orientation and/or gender identity. While scholars have found statistically significant rates of discrimination against same-sex couples, no scholar has yet investigated if local housing protections are effective at reducing discrimination against same-sex couples or LGBTQ-identified individuals (Murchie, 2017; Friedman et al, 2013). This paper has two aims. First, it uses data collected from an online, paired-email correspondence audit to provide the most comprehensive, nationally-representative estimate to date of the level of housing discrimination faced by same-sex couples. It then examines if state- and local-level housing protections are correlated with lower levels of discrimination against same-sex couples relative to localities that do not adopt these legal prohibitions.

Property owners and/or managers, who posted rental units on a public rental site, received two emails inquiring about rental units. In these emails, the sexual orientation of the inquiring couples is signaled via name combinations – two male (female) names and the phrase “my husband” (“my wife”) signal a same-sex male (female) couple. Opposite-sex names signal that the inquiring couple are heterosexuals. The differential response rates between same-sex and opposite-sex couples is taken as evidence of discrimination. To explore cause of discrimination, various elements of the email also signal the racial makeup (via names) of the couples, as well as their socioeconomic status. Names, racial background, and other identifiers are randomly assigned to property owners. The paired-email design allows for property-owner fixed-effects, which eliminates all unit and property-owner level unobservables. The author contacted 6,490 property owners/managers in 94 municipalities in 46 states, each of whom received two emails (12,980 email-landlord observations).

Preliminary results from the audit have found that state-level protections are correlated with lower levels of discrimination against same-sex couples relative to the rate of discrimination found in municipalities without any housing protections. Surprisingly, local-level protections are correlated with higher levels of discrimination. Consistent with previous housing audit literature, discrimination rates covary strongly with race and sexual orientation. High-income black heterosexual couples are 12% less likely to receive a response to their email inquiry than high-income white couples. However, high-income black same-sex male couples are 17% less likely to receive a response than high-income white heterosexual couples. Low-income black same-sex couples are 24% less likely to receive a response than low-income black heterosexual couples, and they are 22% less likely to receive a response than low-income white same-sex couples.