FOR RICHER OR POORER, IN SICKNESS AND IN HEALTH
DO SAME-SEX MARRIAGE BANS
NEGATIVELY IMPACT HEALTH CARE
COVERAGE?
UCLA
28 June 09
NINEZ PONCE, MPP, PHD 1
SUSAN COCHRAN, PHD, MS 2,4
VICKIE MAYS, PHD, MSPH 1,3
1.
2.
3.
4.
Department of Health Services
Department of Epidemiology
Department of Psychology
Department of Statistics
AcademyHealth Annual Meeting,
Chi IL


Employer-based health insurance (EBHI) main source of healthcare coverage

Dependent EBHI coverage generally accrues through marriage

Sexual minorities are disadvantaged

This has both individual and potential societal costs


Objective
 To determine the effect of sexual orientation on health insurance status for nonelderly adults in
California

Notably
 Study focuses on sexual orientation and health insurance consequences for a single state
 Predates the short period in California when same-sex marriage was legal
 Stratifies by gender

Exclusions
CHIS 2001, 2003, 2005
Sexual orientation Q’s
Population-based, RDD
Conducted in 5 languages
Racial/Ethnic
Oversamples
Interviews conducted by proxy
Respondents not reporting sexual orientation
Respondents 65 and older
Final Sample
N=110,254
63,719 females
46,535 males


Outcome
 Current Health Insurance Status
 Categories of: Uninsured, Privately-purchased, Public,
Employer-based own, Employer-based dependent

Predictors
 Regressor of Interest
 Sexual Orientation Status: Gay or Lesbian compared to
Heterosexual
 Other control variables
 Race/ethnicity, Age, Income, Education, Citizenship,
Partnership Status, Minor Children, Language of Interview,
Rural/Urban Labor Characteristics, Health Status


Statistical Analysis
 Weighted multivariate multinomial logit models:
 Full sample (all nonelderly adults)
 Only partnered/married adults
 All models stratified by gender
 Computed predicted probabilities and relative risks
 Gays/lesbians relative to heterosexuals
 Bootstrapped 95% CIs

48%
36%
8%
48%
3%
Women
56%
62%
36%
2%
Men
57%
36%
7% Not Married or
Partnered
Partnered
Married

Comparing Gay Males with Heterosexual Males
Sample size
All Males
46,535
Partnered/Married
29,057
Uninsured
Public
Own EBHI
Dep EBHI
Direct Purchase
Relative Risk [95% CI] Relative Risk [95% CI]
0.88 [0.59, 1.02]
1.21 [0.96, 1.55]
1.03 [0.96, 1.14]
0.78 [0.49,1.08]
1.07 [0.86, 1.25]
1.45 [1.07, 1.93]*
1.91 [1.21, 2.71]*
0.93 [0.80, 1.00]
0.51 [0.34,0 .71]*
1.16 [0.63, 1.65]
*p<0.05 CHIS 2001, 2003, 2005; bootstrapped CIs controlling for Race/ethnicity, Age, Income, Education, Citizenship, Minor Children,
Language of Interview, Rural/Urban ,Labor Market Characteristics, Health Status

Comparing Lesbians with Heterosexual
Females
Sample size
Uninsured
Public
Own EBHI
Dep EBHI
Direct Purchase
All Females
63,719
Partnered/Married
38,195
Relative Risk [95% CI] Relative Risk [95% CI]
1.39 [1.00, 1.81]* 1.99 [1.43, 2.76]*
0.95 [0.72, 1.41]
0.98 [0.78, 1.11]
0.90 [0.52, 1.33]
0.82 [0.60, 1.11]
2.06 [1.01, 3.35]*
1.27 [1.08, 1.48]*
0.32 [0.19, 0.44]*
0.98 [0.72, 1.51]
*p<0.05 ; CHIS 2001, 2003, 2005; bootstrapped CIs controlling for Race/ethnicity, Age, Income, Education, Citizenship, Minor Children,
Language of Interview, Rural/Urban ,Labor Market Characteristics, Health Status


Among cohabiting/married nonelderly adults, gay men and lesbians are more likely to be uninsured than their heterosexual counterparts
 Due to a much lower likelihood of receiving dependent EBHI

Partnered sexual minorities are also more likely to depend on public insurance than heterosexual counterparts


Low survey response rate

Telephone administration introduces nonresponse bias but addressed in weighted analysis

Wage/fringe benefit tradeoff not determined

Inability to differentiate unmarried but partnered couples into those that were registered civil unions/domestic partnerships and those that were not


Bans on same-sex marriages lead to EBHI disadvantages among gays and lesbians

Lower EBHI dependent coverage drives the disadvantage and appears to levy a toll on publically-financed health systems and raise the risk of uninsurance


Employers who discriminate are less efficient and lost efficiency threatens competitiveness and solvency
(Becker, 1971)

Disadvantage stems from how EBHI is offered and taxed –so has implications for upcoming changes in insurance
Same-sex marriage ban and its effects on
EBHI akin to “middle class” tax on
EBHI—and this study saw negative impact on coverage

UCLA
National Institute of Drug Abuse (DA 15539, DA 20826), the National
Center for Minority Health and Health Disparities (MD000508), the
National Cancer Institute, K07 CA100097