Employment Nondiscrimination Protection and Mental Health Among Sexual Minority Adults
Importance
In the 2020 Bostock v Clayton County decision, the US Supreme Court extended employment nondiscrimination protection to sexual minority adults. The health impacts of this ruling and similar policies related to sexual orientation–based discrimination are not currently known.
Objective
To estimate changes in mental health following the Bostock decision among sexual minority adults in states that gained employment nondiscrimination protection (intervention states) compared with those in states with protections already in place (control states).
Design, Setting, and Participants
This cross-sectional study used 2018-2022 data from the Behavioral Risk Factor Surveillance System and a difference-in-differences approach to evaluate changes in mental health after the Bostock decision by comparing sexual minority adults (aged ≥18 years and identifying as lesbian, gay, or bisexual) in 12 intervention states with those residing in 9 control states. Models were estimated for all participants and separately for employed participants. Data were analyzed between February and September 2024.
Exposure
Residing in a state that gained employment nondiscrimination protection after the Bostock decision.
Main Outcomes and Measures
The primary outcome was number of poor mental health days during the past 30 days, and the secondary outcome was severe mental distress (defined as 14 or more past-month poor mental health days).
Results
Of 597 462 participants (306 365 in intervention states [77.7% aged 18-64 years and 22.3% aged ≥65 years; 51.7% female] and 291 097 in control states [77.5% aged 18-64 years and 22.5% aged ≥65 years; 50.6% female]), 5.1% in intervention states and 6.0% in control states self-identified as sexual minority adults. The mean (SE) number of past-month poor mental health days was unchanged after the Bostock decision among sexual minority adults in both intervention (from 8.70 [0.27] to 9.59 [0.24] days; adjusted difference, 0.57 [95% CI, −1.02 to 2.16] days) and control (from 8.53 [0.21] to 10.15 [0.20] days; adjusted difference, 1.17 [95% CI, −0.46 to 2.79] days) states, resulting in no differential change between the 2 groups (difference-in-differences, −0.60 days; 95% CI, −1.25 to 0.06 days). Among the subset of employed sexual minority adults, the mean (SE) number of poor mental health days did not change in intervention states (from 7.99 [0.38] to 8.83 [0.30] days; adjusted difference, 0.87 [95% CI, −0.49 to 2.22] days) but increased in control states (from 7.75 [0.27] to 9.75 [0.26] days; adjusted difference, 1.84 [95% CI, 0.44-3.24] days). These findings corresponded to a significant relative reduction in poor mental health days among employed sexual minority adults in intervention vs control states (difference-in-differences, −0.97 days; 95% CI, −1.74 to −0.21 days). Mean (SE) rates of severe mental distress increased less among employed sexual minority adults in intervention (from 26.35% [1.59%] to 29.92% [1.46%]; adjusted difference, 6.81% [95% CI, 2.20%-11.42%]) vs control (from 26.53% [1.27%] to 34.26% [1.16%]; adjusted difference, 10.30% [95% CI, 5.99%-14.61%) states, also corresponding to a significant relative reduction among employed sexual minority adults (difference-in-differences, −3.49%; 95% CI, −6.71% to −0.27%).
Conclusions and Relevance
These findings show significant relative reductions in past-month poor mental health days and severe mental distress among employed sexual minority adults after the implementation of a federal ban on employment discrimination based on sexual orientation. Larger and more consistent mental health benefits observed among sexual minority adults in the workforce underscore the importance of broadening protections to other social domains.
Top-30
Journals
|
1
|
|
|
Psychiatry Research
1 publication, 50%
|
|
|
Journal of Health Politics, Policy and Law
1 publication, 50%
|
|
|
1
|
Publishers
|
1
|
|
|
Elsevier
1 publication, 50%
|
|
|
Duke University Press
1 publication, 50%
|
|
|
1
|
- We do not take into account publications without a DOI.
- Statistics recalculated weekly.