The nationwide legalization of same-sex marriage in 2015 through the Supreme Court decision in the landmark Obergefell v. Hodges case is one of the most significant advancements for civil rights in recent years. It is regarded as a pivotal advancement towards justice and equality. However, the majority of the conversation about same-sex marriage is dominated by ideology rather than the actual practical implications of it. As such, an element that is often overlooked is the impact of access to same-sex marriage on health insurance and health outcomes for sexual minorities, both good and bad. This paper aims to use the variation in state legal access to same-sex marriage before this decision to evaluate the causal impact of access on health insurance coverage and health outcomes through a difference-in-differences analysis. Data will be taken from the CDC’s Behavioral Risk Factor Surveillance System. (BRFSS) I expect the results of this study to indicate that legal access will increase health insurance coverage and improve health outcomes overall, but certain groups within this population may be negatively impacted. In particular, unmarried same-sex couples are likely to face negative consequences in these two metrics as a result of the nationwide legalization of same-sex marriage.
Despite lingering work to be done, in the past decade, there have been dramatic changes in legal rights and societal attitudes about sexual minorities in the United States. The first state to legalize same-sex marriage was Massachusetts in 2004 following a decision by the Massachusetts Supreme Judicial Court in the case of Goodridge vs. Department of Public Health. After this, legal access to marriage varied on a state-by-state basis with a wide degree of variation in access between them. This extended far beyond the binary of just whether or not same-sex marriage was legal in a given state as some states had it legalized, some had it banned, and others had no position, with different implications for all of those conditions. The legal landscape for same-sex marriage was incredibly complicated and inconsistent across states. Then, the first action at the federal level was taken in 2013 with the ruling in the United States v. Windsor case which gave federal marriage recognition to same-sex couples that wed in states with legal recognition but did not require other states to recognize those marriages. This began the process of a unified legal position across the country but obviously still allowed a great deal of variation. It wasn’t until the landmark case of Obergefell v. Hodges in 2015 that same-sex marriage was made legal across all 50 states.2
Marriage is associated with a vast array of social, legal, and financial benefits. One of the largest financial benefits of marriage is the ability for an individual who is covered by company-sponsored health insurance to add their spouse to the policy. With this condition in mind, it stands to reason that access to same-sex marriage would increase adoption of health insurance as well as improve health outcomes for sexual minorities. However, there are a variety factors that complicate this prediction.
Most notably, private employers are not actually required by law to provide coverage to same-sex married couples. Neither federal ruling made an explicit mandate on this issue leaving it up to employers to decide. In fact, a survey across 2,160 randomly selected employers found that in 2018, only 63% of employers offered spousal health insurance to same-sex couples. While this is more than would be covered without access to same-sex marriage, the fact that health insurance coverage is not required for same-sex couples will definitely mitigate its impact.
Another factor that could impact the number of sexual minorities covered by health insurance is that the legalization of same-sex marriage resulted in a sharp decrease in coverage for non-married same-sex couples. With the aforementioned variation in access to same-sex marriage across states, many employers offered healthcare benefits to same-sex couples in other arrangements such as a civil union or domestic partnership. However, following the Obergefell ruling a critical mass of employers decided to drop this coverage. The International Foundation of Employee Benefits Plans found that from 2014 to 2016 coverage of same-sex partners in civil unions dropped from 51% to 31% and for those in domestic partners from 59% to 48%. This creates a condition in which certain couples who perhaps never had any intention of getting married, or never considered it due to the laws at the time, either have to get married to continue to access their healthcare benefits, and many might not choose to do so for the sole reason of this financial benefit.
Finally, there are differences in health behaviors and demographic-specific health problems that may reduce the impact that access to same-sex marriage has on health insurance and health outcomes. Members of the LGBT community are at a higher risk for substance use, sexually transmitted diseases (STDs), cancers, cardiovascular diseases, and mental health issues.2 In fact, this disparity is so profound that in 2012 the U.S. Department of Health and Human Services made improving health for this community an explicit objective outlined in the national Healthy People 2020 objectives. There are a number of factors that could potentially explain this disparity but researchers at the National Institutes of Health have found that the main circumstances that likely cause this effect are, “stigma, lack of healthcare providers’ awareness, and insensitivity to the unique needs of this community.” (Hafeez et al, 2017)
This paper looks to address an important and overlooked policy implication of legalizing same-sex marriage. That is, what is the impact of access to same-sex marriage on health insurance adoption and health outcomes for sexual minorities? To do so, this study will exploit differences in state access to same-sex marriage prior to national legalization in a difference-in-differences analysis.
The first area that previous literature exists in is between the relationship between marriage and health broadly speaking. Extensive research has demonstrated positive health outcomes as a result of marriage. It has been associated with higher survival rates from fatal diseases such as cancer. (Gardner and Oswald 2011) In addition, individuals experienced better physical health and mental health overall. (Umberson 2003)There are various theories to explain why these benefits exist in addition to the benefit of being able to adopt one’s spouse’s health insurance. One such theory builds on the Grossman model for health capital, in that it may be that individuals are simply engaging in healthier lifestyles and putting more investment into their health care due to their commitment to the marriage and their spouse. (Sherbourne et al, 1991)Another theory is that there is a positive externality on health as a result of the physical and emotional support that a partner is able to provide. (Ducharme 1994)
Another body of literature that is important to my study is the connection between health insurance and health outcomes. This is another well-documented area that has been looked at extensively. Two of the most famous economic studies looked at this very concept. The first of which is the RAND Health Insurance Experiment which randomly assigned people into 14 different health insurance plans with varying degrees of coverage. The study found that amongst the 14 groups individuals on average all experienced similar health outcomes in the metrics used. (Manning et al 1988) This indicates that the exact level of coverage does not matter, but more so whether or not an individual has some level of health insurance. A separate randomized controlled trial was conducted several years later known as the Oregon Health Insurance Experiment. In this study, people were either assigned to have health insurance or not. The study found significant increases in health care usage, with increased hospitalizations, outpatient visits, and prescription drug use. There were also moderate improvements to health outcomes with improved diabetes care and reported mental health. (Finkelstein et al 2012). On the other hand, a separate study found that there is a significant increase in mortality amongst uninsured individuals due to the fact that they are less likely to get preventative care and thus are already very ill when diagnosed. (Woolhander et al 2017)
Finally, I reviewed the smaller pool of studies that exist that have looked at the topic of same-sex marriage and health outcomes in sexual minorities. One study that was conducted in Europe found that same-sex marriage resulted in a reduction in sexually transmitted infections amongst sexual minorities. (Dee 2008) However, it is important to note that both marriage laws and associated benefits are very different in other countries so this does not provide the most solid comparison. However, due to the fact that this study was evaluating a behavioral change, it is worth mentioning for the purpose of the analysis. Another study found that legal access to same-sex marriage reduced suicide attempts amongst sexual minorities. They theorized that legal same-sex marriage mitigated the effect of “minority stress,” a concept that describes the chronic stress that oppressed minority groups feel. (Raifman et al, 2017)Finally, a regression discontinuity study found that in Massachusetts there was a decrease in utilization of mental health services amongst sexual minority men, again suggesting that the existence of same-sex marriage has a positive impact on mental health. (Hatzenbuehler et al, 2013)
This literature review helps explicate the conceptual framework in which this study is based. The focus of this paper is to look at how legal access to same-sex marriage impacts adoption of health insurance and health outcomes. As shown in the literature review, health insurance has been linked to better health outcomes. Additionally, the mere presence of marriage itself has health associated health benefits. The idea is that with legal access to same-sex marriage, more same-sex couples will get married, thus, increasing health outcomes through expanded health insurance coverage from spousal coverage and the benefits derived from marriage itself. However, as previously discussed, there are a number of mitigating factors that may reduce this impact or indicate no impact at all. These factors are the lack of legal requirement for spousal coverage, drop in coverage to same-sex relationships other than marriage, and demographic specific health needs and behaviors. All to say that there is a clear gap in previous literature that this study seeks to fill.
This study will utilize the Behavioral Risk Factor Surveillance System (BRFSS) conducted by the CDC. This data set is collected from national telephone surveys that measure, “Health-related risk behaviors and events, chronic health conditions, and use of preventive services.” With several hundred thousand people interviewed each year, it is one of the world’s largest health survey’s.
The data will have to be adjusted in a number of ways for the purposes of our study. The first is setting a timeframe in which to look at for the study. The first state to legalize same-sex marriage in the United States was Massachusetts in 2003 and spread to nationwide legalization in 2015. Therefore, the study will use BRFSS data from 2002-2014 to disregard inputs from before any states had legal access or after they all did. Next, is determining households that have same-sex couples that live in them as sexual orientation was not one of the questions asked on the survey until the most recent years it was conducted. However, one of the question’s asked is about the composition of adults in the household. In order to reduce the likelihood that the adults are living together for reasons other than a romantic relationship (roommates, students, etc.) we will limit the sample to adults 25 years or older reporting exactly two adults in a home. The simple assumption being that if there are two adults this age or older living together in a household it is very likely that they are in a same-sex relationship, whether this is two women or two men. Once the sample is limited to the population that we are looking to evaluate the remainder of the outcome variables are asked very directly in the survey. There are binary yes or no questions asked about marital status and health insurance coverage as well as a self-reported examination of health status, all of the key variables our study is looking to examine.
My study will exploit the variation in the timing of legal access to same-sex marriage across states to evaluate the impact of access on health insurance and health outcomes. I will employ a difference-in-differences analysis, meaning, I will be comparing the outcomes for sexual minorities living in states with legal access to same-sex marriage against people living states without them in terms of health insurance and health outcomes with the hypothesis being that it will increase the number of people insured and improve health outcomes. It is a necessary assumption for this kind of analysis that outcomes would have been the same if not for the treatment applied to the different groups. In this study, the sheer number of states included within the different treatment group is assumed to cancel out any state fixed effects that could have otherwise influenced outcomes.
My sample will be adults above the age of 25 living in a household with exactly two adults one as indicated by the survey utilized as my data set. Thus, by comparing across the entire United States and taking these factors into account this should isolate the difference in legal access to same-sex marriage as the causal impact of differences that are accounted for between outcomes in sexual minorities across states. The major assumption being here that there are not other factors that could account for these differences. By nature of the size of the study across all the different US states the study would mitigate any fixed effects of individual state conditions, and in isolating other aspects of identity through race and gender also controls for other health-related variables.
For my difference-in-differences analysis I will utilize the following regression:
- Hist = α + β1*Xist + β2*(access to same-sex marriage)ist + eist
Where H is the outcome variable looking at health insurance and health outcomes, I is the individual, s is the state, t is the time, and “access to same-sex marriage” is a dummy variable equal to one if the state has legal access to same-sex marriage. This regression will be ran twice, once with two males and once with two females to see if there is a difference in the gender composition of the couples.
I predict this study will find that sexual minorities who live in states with legal access to same-sex marriage will have a greater prevalence of health insurance coverage and improved health outcomes. However, the caveat to this conclusion is that I also expect unmarried same-sex couples in other arrangements to see a drop in their coverage and health outcomes. On net, legal access to same-sex marriage will likely yield a positive result for these metrics in sexual minorities. However, it is important to not overlook the couples that will be negatively impacted as well.
The major limitation to the study as it is designed is a threat to internal validity. This threat exists as a result of the way that the study determines same-sex households by making a reasonable assumption derived from the composition of the number of same-sex adults living in that household. Using this method will almost definitely incorporate some households that include people living together for reasons other than a relationship. However, it is impossible to determine how large this group of people is. With this being said, a study that relied on self-reported survey results about sexuality this would also likely yield a slightly biased result due to the fact that people who are not open about their sexuality may not want to accurately disclose it on the phone to an unknown interviewer. So, despite its limitations, this is still likely the best method to determine same-sex households.
Should the study yield the predicted results, this would have some clear and robust implications. The first of which is that employers should continue to expand coverage to same-sex spouses in order to improve health outcomes for sexual minorities. In the same vain, employers should not drop coverage for same-sex couples in domestic partnerships or civil unions as this will likely have a negative impact on their health outcomes.
After meeting with my writing fellow I made some changes to the pose and grammar of the paper. I also put my references into the correct format and included a section on my calculations including the regression that I would do for my study. But beyond that I have made a number of significant revisions to my paper and study itself following my last draft. As I continued to do research I learned more about the legal nuance of access to same-sex marriage that I elaborated on in the paper. I also learned more about the types of same-sex couples in arrangements such as domestic partnerships or civil unions and wanted to make sure significant attention was given to this sub-group.
- Dee, Thomas. “Forsaking All Others? The Effects of Same-Sex Partnership Laws on Risky Sex.” 2008, Economic Journal 118, 530
- Ducharme, F. “Conjugal Support, Coping Behaviors, and Economic Well-Being of the Spouse” 1994, Economica, 16(2), 167–190.
- Finkelstein, Amy et al. “The Oregon Health Insurance Experiment: Evidence from the First Year”, 2012, The Quarterly Journal of Economics
- Gardner J., and A. Oswald. “How is Mortality Affected by Money, Marriage and Stress?” 2011, Journal of Health Economics
- Hafeez, Hudaisa et al. “Health Care Disparities Among Lesbian, Gay, Bisexual, and Transgender Youth: A Literature Review” Cureus vol. 9,4 e1184. 20 Apr. 2017, doi:10.7759/cureus.1184
- Hatzenbuehler, Mark L et al. “Structural stigma and all-cause mortality in sexual minority populations” 2012, Journal of Behavioral and Experimental Economics vol. 103 (2013): 33-4
- Manning WG, Newhouse JP, Duan N, Keeler EB, Benjamin B, Liebowitz A, et al. “Health insurance and the demand for medical care. Evidence from a randomized experiment.” RAND Corporation, 1988. The American Economic Review
- Raifman, Julia et al. “Difference-in-Differences Analysis of the Association Between State Same-Sex Marriage Policies and Adolescent Suicide Attempts” 2017, American Economic Revue vol. 171,4 (2017): 350-356.
- Sherbourne, Donald et al.“Marital Status, Social Support, and Health Transitions in Chronic Disease Patients.”, 1991 Journal of Behavioral Economics. 31. 328-43. 10.2307/2136817.
- Umberson, Debra. “Family Status and Health Behaviors: Social Control as a Dimension of Social Integration.” 2003, Economic Inquiry
- Woolhandler S, Himmelstein DU. The Relationship of Health Insurance and Mortality: Is Lack of Insurance Deadly?. 2017, Applied Health economics and Health Policy;167:424–431. doi: 10.7326/M17-1403
 Investopedia, https://www.investopedia.com/financial-edge/0412/why-marriage-makes-financial-sense.aspx#