Disparities in Health Insurance Among Children With Same-Sex Parents
- Gilbert Gonzales, MHA and
- Lynn A. Blewett, PhD
OBJECTIVES: The objectives of this study were to examine disparities in health insurance coverage for children with same-sex parents and to investigate how statewide policies such as same-sex marriage and second-parent adoptions affect children’s private insurance coverage.
METHODS: We used data from the 2008–2010 American Community Survey to identify children (aged 0–17 years) with same-sex parents (n = 5081), married opposite-sex parents (n = 1 369 789), and unmarried opposite-sex parents (n = 101 678). We conducted multinomial logistic regression models to estimate the relationship between family type and type of health insurance coverage for all children and then stratified by each child’s state policy environment.
RESULTS: Although 77.5% of children with married opposite-sex parents had private health insurance, only 63.3% of children with dual fathers and 67.5% with dual mothers were covered by private health plans. Children with same-sex parents had fewer odds of private insurance after controlling for demographic characteristics but not to the extent of children with unmarried opposite-sex parents. Differences in private insurance diminished for children with dual mothers after stratifying children in states with legal same-sex marriage or civil unions. Living in a state that allowed second-parent adoptions also predicted narrower disparities in private insurance coverage for children with dual fathers or dual mothers.
CONCLUSIONS: Disparities in private health insurance for children with same-sex parents diminish when they live in states that secure their legal relationship to both parents. This study provides supporting evidence in favor of recent policy statements by the American Academy of Pediatricians endorsing same-sex marriage and second-parent adoptions.
- health insurance
- gay and lesbian parents
- same-sex households
- AAP —
- American Academy of Pediatrics
- ACS —
- American Community Survey
- aOR —
- adjusted odds ratio
- CI —
- confidence interval
What’s Known on This Subject:
Health insurance is associated with improved health for children, but gay and lesbian parents face barriers to adding their children to private health insurance. Little is known about the extent to which insurance disparities exist for children with same-sex parents.
What This Study Adds:
Children with same-sex parents are less likely to have private health insurance. When children live in states in which legal same-sex marriage, civil unions, domestic partnerships, or second-parent adoptions are available, disparities in private insurance diminish for children with same-sex parents.
A growing number of children are living in same-sex households and being raised by gay and lesbian parents. Data from the American Community Survey indicate that there are ∼125 000 same-sex couples raising 220 000 children in the United States (although these estimates do not incorporate children with single gay or lesbian parents or families not disclosing their same-sex relationship status).1,2 As a result, pediatricians should expect to encounter more children from gay and lesbian families in their practices. Although disparities in health and health care are well documented for gay and lesbian adults,3–8 very little is known about the children living within their households. The large number of studies that do investigate the health and well-being of children with same-sex parents disproportionately focus on the psychological and social development of children.9–13 Yet, gay and lesbian families face unique challenges that directly impact the health of their children, particularly health-related outcomes associated with health insurance.14,15
More than half (55%) of the US population receives health insurance through their own or a family member’s employer-sponsored health plan,16,17 but children with gay and lesbian parents can face barriers to gaining coverage through a parent’s private health plan. First, not all employers extend health benefits to same-sex partners of employees and their children like they do for married heterosexual couples. Approximately 30% of all employers offering health benefits have extended health insurance to same-sex partners and their children as of 2012.18,19 Employers offering health insurance to employees and their dependents often set provisions requiring that qualified children be related to the employee by birth, by legal marriage, or by legal adoption so that health insurance is exempted from federal taxes under standards determined by the Internal Revenue Service.20
Not all states afford children with same-sex parents the opportunity to be legally related to both parents through legal marriage or legal adoption. At the time of this writing, 13 states and the District of Columbia recognize legal marriages for same-sex couples, and an additional 6 states recognize civil unions or comprehensive domestic partnerships that include full spousal and family rights to same-sex couples (Fig 1).21 When states adopt these provisions, employers that are “fully insured” and regulated by state insurance laws are often required to extend health benefits to the dependents of gay and lesbian employees.22 Additionally, not all states allow adoption by same-sex parents. Eighteen states currently allow “second-parent” adoptions statewide, which permit both parents of a gay or lesbian couple the ability to be legally adoptive parents (Fig 1).23 State policies such as same-sex marriage and second-parent adoptions secure the child’s eligibility for private health insurance from both parents.
To the best of our knowledge, prior research has not estimated the health insurance coverage patterns and the disparities that are likely to exist among children living in same-sex households. In this article we examine the distribution of health insurance coverage for children in same-sex households compared with their counterparts living with heterosexual parents and investigate how state policies, namely same-sex marriage, civil unions, domestic partnerships, and second-parent adoption, affect the distribution of health insurance coverage.
We analyzed data from the 2008–2010 American Community Survey (ACS) 3-year public use microdata sample.24 The ACS is a general household survey conducted by the US Census Bureau and is designed to provide states and communities with timely demographic, social, economic, and housing information. The ACS maintains an annual sample size of ∼3 million housing units and a monthly sample size of ∼250 000 households. The large samples provided by the ACS make it a powerful resource for studying relatively small subpopulations, such as same-sex households.25
Like most federal surveys, the ACS does not ascertain sexual orientation. Instead, same-sex couples and their children were identified on the basis of the relationship to the primary respondent. Adults in same-sex relationships were identified when the respondent identified another person of the same sex as a husband, wife, or unmarried partner. Same-sex spouses using the husband or wife response categories were reassigned as unmarried partners in the public use files by the Census Bureau regardless of the legal status of their marriage (M. O’Connell, PhD, G. Gooding, MA, unpublished observations, 2012).26 Meanwhile, the instruction guide accompanying the survey defined an unmarried partner as “a domestic partner” or “a person who shares a close and personal relationship with the reference person.”27 Consistent with previous research using similar techniques to identify same-sex couples, we assume that these couples are lesbian, gay, or bisexual adults.4–8,28 (Our identification strategy cannot ascertain transgender populations given the binary male-female categories on gender identity included in the survey).
The subjects of this analysis were children aged 0 to 17 years related to the primary respondent as a biological child, adopted child, or stepchild. Our final sample sizes included 1 369 789 children in married opposite-sex households, 101 678 children in unmarried opposite-sex households, 1649 children in dual-father same-sex households, and 3432 children in dual-mother same-sex households. We separated same-sex households by those headed by 2 men versus those headed by 2 women to determine whether state-level policies affect gay and lesbian households differently as has been detected in previous research.29 We did not include children in single-parent households because they do not share economic experiences similar to 2-parent households.30 Furthermore, our strategy for identifying same-sex households using relationship information cannot differentiate children in single-parent households headed by gay or lesbian parents.
A question regarding health insurance was added to the ACS in 2008 and requires the respondent to report current type of health insurance coverage for all members of the household.31 We assigned each child into 1 of 3 insurance categories: (1) private coverage, (2) public coverage, or (3) uninsured. Children were assigned private health insurance if the respondent indicated that the child was covered by employer-sponsored insurance, TRICARE or other military health care, or insurance purchased directly from an insurance company. Children covered by public health insurance had insurance through Medicare, Medicaid, or the Children’s Health Insurance Program. Children were assigned uninsured if the respondent reported no type of insurance coverage or coverage through the Indian Health Service.32
Our primary independent variable of interest was the type of family each child belonged to: married opposite-sex parents, unmarried opposite-sex parents, dual-father same-sex parents, or dual-mother same-sex parents. Children were also grouped into variables previously found to be associated with children’s health insurance coverage.33,34 Child demographic variables included the following: age group in years (<1, 1–5, and 6–17 years), race and ethnicity (non-Hispanic white, Hispanic, non-Hispanic black, non-Hispanic Asian, and non-Hispanic multiple or other races), gender, citizenship status (citizen, naturalized, and noncitizen), disability status (reporting at least 1 of 6 disability types: deaf or serious difficulty hearing; blind or serious difficulty seeing; physical, mental, or emotional problems; serious difficulty walking or climbing stairs; difficulty bathing or dressing; and, because of a physical, mental, or emotional problem, difficulty doing errands), and relationship to reference parent (biological, adopted, and stepson or -daughter). Household variables included the following: age group of the reference parent in years (15–24, 25–34, 35–44, 45–54, 55–64, and ≥65 years), combined parents’ income (<100%, 100% to <200%, 200% to <400%, and ≥400% of the federal poverty guidelines for the corresponding year), work status of parents (any adult working full-time, only part-time, all unemployed, and all not in labor force), highest parental educational attainment (less than high school, high school, some college, and college), total number of children in household (≤2, ≥3), and primary language spoken at home (English, not English).
We first estimated descriptive statistics, including health insurance coverage, for children by family type. Pearson χ2 statistics were used to compare descriptive characteristics of children across family types. A multinomial logistic regression model was then used to assess the association between family type and type of health insurance while controlling for demographic and household characteristics. We report the adjusted odds ratios (aORs) for the regression model, which included all independent variables discussed in the previous section in addition to state and year fixed effects. No substantial correlations were found among the independent variables after diagnostic tests for collinearity.
We also tested whether state policies modified disparities in health insurance coverage by stratifying children on the basis of on the presence of same-sex marriage, civil unions or comprehensive domestic partnerships with spousal rights (California, District of Columbia, Massachusetts, New Jersey, and Vermont), and second-parent adoptions (California, Colorado, Connecticut, District of Columbia, Illinois, Indiana, Massachusetts, New Jersey, New York, Pennsylvania, and Vermont) in each child’s state as of January 1, 2008.21,35 We estimated the aORs by each state policy group to determine whether state policies qualitatively modified the relationship between a child’s family type and type of insurance coverage. All regression models were estimated in Stata 12 by using mlogit and svy commands with sampling weights.36 This study was exempt from approval by the institutional review board because the data were obtained from secondary sources.
Children in married opposite-sex households were more likely to have private insurance (77.5%; Table 1) compared with children in same-sex households with dual fathers (63.3%) and dual mothers (67.5%). Notably, children in unmarried opposite-sex households were much less likely to have private insurance (37.7%), but their public insurance coverage (51.1%) exceeded any other group. Children in same-sex households were also more likely to be adopted, which may affect their access to private health insurance when employers require that adopted children be adopted legally. Nearly 13% and 16% of the children in dual-father and dual-mother households, respectively, were adopted by the primary respondent. Fewer than 3% of the children in married and unmarried opposite-sex households were adopted by the primary respondent.
Results from the multinomial logistic regression model on the entire sample indicated significant differences in private health insurance coverage across family types (Table 2). After adjusting for demographic and household characteristics, children with dual fathers (aOR: 0.55; 95% CI [confidence interval]: 0.39–0.79) or dual mothers (aOR: 0.61; 95% CI: 0.47–0.79) were significantly less likely to have private insurance compared with children with married opposite-sex parents. Children with unmarried opposite-sex parents were much less likely to have private insurance (aOR: 0.46; 95% CI: 0.44–0.48), but they were significantly more likely to be covered by public insurance (aOR: 1.42; 95% CI: 1.36–1.48). A similar relationship with public health insurance was not found for children with same-sex parents. Although not the primary focus of this study, the additional independent variables predicted patterns and associations with type of health insurance in expected directions.
Our final models examined the association between family type and health insurance coverage on the basis of state policies that were available to gay and lesbian families before the first survey year. No statistical difference was detected for private insurance coverage between children with dual mothers (aOR: 0.81; 95% CI: 0.48–1.37) and children with married opposite-sex parents living in states with same-sex marriage, civil unions, or comprehensive domestic partnership laws (Table 3). Differences in the odds of private health insurance coverage remained statistically significant for children with dual fathers (aOR: 0.43; 95% CI: 0.20–0.95). When same-sex marriage, civil unions, or domestic partnership laws were absent in the state before being surveyed, children with either dual fathers (aOR: 0.62; 95% CI: 0.41–0.93) or dual mothers (aOR: 0.60; 95% CI: 0.45–0.79) experienced statistically lower odds of private health insurance coverage. The odds of having public health insurance were similar for children with same-sex parents and married opposite-sex parents across states with and without same-sex marriage policies. Differences in private and public health insurance coverage remained directionally and statistically similar across state policies for children with unmarried opposite-sex parents.
Differences in private health insurance also diminished for children with same-sex parents when these families lived in states that allowed second-parent adoptions (Table 4). Children with dual fathers (aOR: 0.71; 95% CI: 0.41–1.25) or dual mothers (aOR: 0.65; 95% CI: 0.41–1.04) were marginally less likely to have private coverage than children with married opposite-sex parents, but these differences were not statistically significant. Instead, significant differences in private health insurance coverage remained for (1) children with same-sex parents where second-parent adoptions were not available statewide and (2) children with unmarried opposite-sex parents.
To the best of our knowledge, limited research has studied the health and the provision of health care for children with gay and lesbian parents. One qualitative study conducted >15 years ago found that gay and lesbian parents in Boston were able to obtain pediatric care that was affirming, supportive, and satisfactory.37 Rather, most studies on children with same-sex parents have disproportionately focused on their psychological and social development.9–13 Whereas previous studies have examined childhood disparities in health insurance on the basis of race and ethnicity,38,39 socioeconomic status,40,41 and immigration status,42,43 this study is the first to our knowledge to evaluate health insurance coverage and disparities for children in gay and lesbian families.
Our study found that children with same-sex parents were less likely to have private health insurance than their peers with married opposite-sex parents. Disparities in private health insurance coverage diminished when children were living in states with legal same-sex marriage and civil unions or second-parent adoptions for gay and lesbian couples. State policies regarding same-sex marriage and second-parent adoptions did not affect differences in health insurance coverage for children with unmarried opposite-sex parents, probably because they were not affected by these provisions. Interestingly, we did not find any differences in public health insurance coverage rates for children living in same-sex households compared with their counterparts with married opposite-sex parents. Public programs such as Medicaid and the Children’s Health Insurance Program assess children’s eligibility for coverage on the basis of household income rather than parental marital status.
Since 2002, the American Academy of Pediatrics (AAP) has endorsed legal adoption by same-sex parents, because “children deserve to know that their relationships with both of their parents are stable and legally recognized” (p 339).44 The policy statement and accompanying technical brief45suggested that legal adoption for same-sex parents would ensure a child’s eligibility for health insurance coverage from both parents. This study adds early evidence in favor of the AAP’s policy statement. Moreover, this study adds to the growing body of evidence on the health benefits associated with legal same-sex marriage.46,47 Yet, potential gains in health that accompany legal same-sex marriage are often discussed in the context of gay and lesbian adult populations and have excluded the children raised by gay and lesbian parents and the health issues important to children and pediatricians.4–8,48
In a 2006 analysis commissioned by the AAP Board of Directors,49 expanded access to employer-based health insurance was listed among the benefits that are conferred to children when their same-sex parents are allowed to marry. More recently, pediatricians and the AAP have endorsed marriage equality for same-sex couples.50–52 Although the future of legal same-sex marriage and the benefits to health remain uncertain, this study suggests that children in same-sex households experienced increased access to private health insurance coverage when they lived in states that allowed their parents to marry, especially for children in dual-mother households. This finding is consistent with early evidence in California that revealed significant gains in health insurance coverage for lesbian couples rather than gay male couples after health insurance mandates for same-sex couples.29
Studying children in same-sex households or with gay and lesbian parents remains challenging because of data limitations similar to the ones we faced. We relied on a common strategy to identify same-sex couples on the basis of the relationship to the survey respondent and assuming that same-sex couples were gay, lesbian, or bisexual.4–8 Missing from our analysis were same-sex couples outside of the relationship with the primary respondent or same-sex partners identified as unrelated adults rather than a husband, wife, or unmarried partner. Also missing were children of same-sex couples who were living outside of the household or who were not reported as the biological, adopted, or stepchild of the primary respondent. There is some concern that our sample of same-sex households may include opposite-sex couples who misreported gender (G. Gates, M.D. Steinberger, unpublished observations). The computer-assisted telephone and personal interview versions of the ACS verified the gender of the husband, wife, and unmarried partner if it matched the primary respondent’s gender. After restricting our sample to the same-sex couples who confirmed their gender in the computer-assisted telephone and personal interview versions of the ACS and the couples most likely not using the husband/wife designation on the basis of marital status “allocation flags” in the ACS (as they are prone to error; G. Gates, PhD, M.D. Steinberger, PhD, unpublished observations, 2009),53 we estimated aORs similar in direction and significance to the results presented in Tables 2, 3 and 4.
Our study would have benefited from knowing the legal status of the same-sex couple’s relationship. We cannot distinguish whether same-sex couples are legally married, are in a state-sanctioned civil union or domestic partnership, or are unmarried cohabitating partners. The Census Bureau reassigns same-sex couples identified as husband or wife to unmarried partners without providing specific “assignment edit flags” in the public use files (M. O’Connell, PhD, G. Gooding, MA, unpublished observations, 2012).26 Making these edits or withholding edit flags in the public use files prevents researchers from examining differences between unmarried same-sex couples and married same-sex spouses. Notwithstanding the data limitations here, the ACS is the predominant resource for studying same-sex households and health insurance coverage.1,25,31,32
Conducting health research on gay and lesbian families is difficult when data are not collected. A first step toward improving our knowledge on children with gay and lesbian parents is to include questions on sexual orientation in health surveys. Beginning this year, the National Health Interview Survey included for the first time a question on sexual orientation. Pediatricians and children’s health researchers should take advantage of this improved data to build our understanding of the unique health circumstances facing gay and lesbian families and their children. Specifically, the health research agenda should move beyond determining how parental sexual orientation affects childhood psychological and social outcomes and toward advancing issues in accessing appropriate health care and tracking health behaviors of children with same-sex parents.
Findings from our study indicate that children with same-sex parents face barriers to private health insurance that translate into significant disparities in insurance coverage. Disparities in private health insurance, however, can potentially be modified and even reduced when pediatricians and states support policies that promote the well-being of all children, including same-sex marriage and second-parent adoptions. Results highlighted in this study provide supporting evidence in favor of recent policy statements by the AAP endorsing same-sex marriage and second-parent adoptions.
We thank Breanna Auringer for her technical assistance and the Pediatrics editors and 2 anonymous reviewers for their helpful comments and suggestions.
- Accepted July 11, 2013.
- Address correspondence to Gilbert Gonzales, MHA, Division of Health Policy and Management, University of Minnesota, 2221 University SE 345, Minneapolis, MN 55414. E-mail:
Mr Gonzales conceptualized and designed the study, carried out the analyses, drafted the initial manuscript, and revised the manuscript; Dr Blewett discussed the conceptualization and design of the study and critically reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships to disclose.
FUNDING: This study was funded in part by a grant from the Robert Wood Johnson Foundation to the State Health Access Data Assistance Center (SHADAC) at the Division of Health Policy and Management, School of Public Health, University of Minnesota.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Copyright © 2013 by the American Academy of Pediatrics