Sexual Orientation Data Collection Policy in the United States: Public Health Malpractice

The greatest threat to the health of lesbian, gay, and bisexual (LGB) Americans is the lack of scientific information about their health. Researchers studying LGB health recognized the importance of having data to understand the health of sexual minorities over a century ago using the motto “Justice through Science.”¹ Researchers in the United States, many of them lesbian, gay, or bisexual themselves, beginning in the late 1980s and through the 2000s, started to argue for the collection of sexual orientation data in publicly funded data sets and, in particular, surveillance systems at the Centers for Disease Control and Prevention (CDC). Surveys targeted included the National Health Interview Survey (NHIS) and the National Health and Nutrition Examination Survey (NHANES), and surveillance systems including The Behavioral Risk Factor Surveillance System (BRFSS) and the Youth Risk Behavior Surveillance System (YRBSS). Arguments for sexual orientation data collection largely centered on the need for data concerning the spread of HIV/AIDS, but arguments also recognized other sexual minority health concerns including drug, alcohol, and tobacco use; mental health; cancer; aging; and violence.²

Some researchers even argued for sexual orientation data collection in all data collection activities directly or indirectly supported by the US Department of Health and Human Services (HHS) including all research funded through the National Institutes of Health (NIH).³ The arguments and actions on the part of researchers sometimes focused on individual variables that could be added to specific surveys, but they also sometimes argued for recognizing sexual orientation as a demographic variable like race, ethnicity or age, and providing guidance for standardized sexual orientation data collection in all HHS activities in the same way the Office of Management and Budget’s Directive 15 provided instructions for the collection of race and ethnicity data.⁴ Outside of some extremely modest steps to collect sexual orientation data in relationship to understanding the spread of HIV, HHS largely ignored these calls for greater data collection.

In fact, we would argue that the handful of data sets and surveillance systems that have added sexual orientation variables have done so despite leadership at HHS through numerous political administrations beginning in the 1980s, including those of Ronald Reagan, George H. W. Bush, Bill Clinton, George W. Bush, and Barack Obama. Individual managers of surveys have sometimes on their own made the decision to add sexual orientation variables to the survey they manage because they recognized the importance of this data in the context of their specific survey. For example, three variables that can be used to measure sexual orientation were added to The National Epidemiological Survey on Alcohol and Related Conditions (NESARC) in the early 2000s after the director of the survey held public hearings in which researchers made the case for sexual orientation data collection. Having well-documented requests from outside experts shielded the survey to some degree from potential criticisms. But the managers of many data sets actually feared collecting sexual orientation data based on their understanding of what happened with The National Health and Social Life Survey, which lost funding before its completion because of its collection of data on sexuality, including sexual orientation.⁵ Many other studies that directly or indirectly studied LGB people and received funding from the US government were called into question by antigay organizations that monitored NIH funding. Studies of LGB people were often consequently hidden within other studies, and the studies used terms such as “diverse” populations instead of the terms “lesbian,” “gay,” or “bisexual” to remain invisible to those opposing such research. This tactic served to keep researchers studying sexual minorities as well as the resultant research in a scientific closet. This closet still exists, and its impact remains unrecognized and undocumented.

The lack of sexual orientation data collection can almost entirely be credited to a lack of leadership at HHS. However, having worked with many survey administrators, other reasons for not collecting sexual orientation data have been voiced, including the belief that sexual orientation data are not relevant to their survey; that adding a variable to their survey would be costly or require another question to be deleted (because of survey length); that a valid and reliable measure of sexual orientation does not exist; that respondents would refuse to answer sexual orientation questions or break off interviews when asked their sexual orientation; that lesbians, gays, and bisexuals would be so rare there would not be enough power to analyze data related to sexual orientation; that a sexual orientation variable is really a proxy for other variables that should be measured instead (e.g., stigma and discrimination); and finally, that if LGB people are found to be at greater risk for certain health concerns, this could be used to further stigmatize these populations or even be used to argue for the curing of homosexuality. Space does not permit here, but these concerns are largely unwarranted and easily addressed. For example, we know that sexual orientation plays a role in such a broad spectrum of health outcomes that rarely, like race or ethnicity, is it irrelevant.⁶ Furthermore, excellent measures of sexual orientation exist, and respondents rarely refuse to answer sexual orientation questions or break off interviews because a sexual orientation question is asked of them.⁷

HHS’s failure to act in the face of the documented health disparities faced by LGB people is a clear example of public health malpractice. The few actions that have been taken by HHS leadership include its most recent support for the development of a sexual orientation question that was added to NHIS in 2013. Listening sessions were also held to discuss the collection of gender identity data to allow for the study of transgender health. Little else of significance can be credited to HHS leadership over the past several decades related to sexual orientation data collection. In 1999 the office of the Assistant Secretary for Planning and Evaluation funded a white paper on sexual orientation data collection. Recommendations from this paper were completed in 2000 and published here in the American Journal of Public Health in 2001.⁸

These recommendations, which are summarized here, included recommendations that HHS (1) should examine the collection of sexual orientation data in HHS data collection and reporting activities; (2) create a process for the development and selection of standard definitions and measures of sexual orientations; (3) should assist in the selection of a minimum set of standard sexual orientation measures, including questions and response categories, and examine the implications of these categories on data tabulation and analysis; and (4) should develop a long-range strategic plan for the collection of sexual orientation data that includes periodic oversampling or screening sampling of LGB people.

It is worth repeating these recommendations here because of HHS’s decade of inaction. Implicit in these recommendations was a call for leadership that never materialized. The YRBSS, which is the focus of this issue of the Journal, is a perfect example of how NIH’s benign neglect produced chaos.

Since 1991 the YRBSS has been collecting data both through a national survey conducted by the CDC as well as through education and health agencies in urban school districts and at local, state, territorial, and tribal levels. The primary aim of the YRBSS is to monitor trends in the leading health-risk behaviors among high school students (grades 9–12) in the United States.⁹ The YRBSS also helps measure progress toward national health objectives. Although Healthy People 2010 and Healthy People 2020 both directed the YRBSS to collect sexual orientation data to monitor specific objectives, and despite well-documented evidence showing LGB youths to be at greater risk than other youths for more than half of the risk behaviors measured by the survey, the CDC does not require and has only recently begun to recommend the collection of sexual orientation data on these surveys.¹⁰

A core set of survey variables are required on local, state, territorial, and tribal survey instruments to receive CDC funding and data. At the local level survey administrators can add variables in addition to the core required variables and some survey administrators have chosen to use this opportunity to collect sexual orientation–related data. Sexual orientation data collection first occurred in Massachusetts in 1993, with other states cautiously following suit. As of 2011, 14 states plus the District of Columbia had collected sexual orientation data by variously assessing sexual orientation identity, sexual behavior or sexual attraction, as well as bullying because of perceived sexual orientation.¹¹

Although it is encouraging that a number of states have collected sexual orientation data, there has been tremendous variation not only between these states and localities in how sexual orientation is assessed, but within individual states and in the types of questions chosen and their wording. This variation makes comparing data across states and localities difficult at best. Currently the CDC provides sites with a list of optional YRBS questions, among which one question regarding sexual orientation identity and one question assessing sex of sexual contacts are included. For the 2015 edition of the survey the CDC is providing a $5000 incentive to any state that asks at least one of these questions.¹² Still, fewer than half of all states collect sexual orientation data and it is unclear whether states will be required to report this data back to the CDC.

Even if the national YRBSS never includes a sexual orientation variable in its core, the survey should be modified to be more sensitive to sexual minorities. For example, LGB youths may respond to questions that ask about “sexual intercourse” in ways that were unintended or are hard to interpret when analyzing data. For example, the question “Have you ever been physically forced to have sexual intercourse when you did not want to?” may exclude from surveillance women forced to have sex with other women, or men forced to have sex with other men. Additional questions asking about sexual intercourse including “Have you ever had sexual intercourse?” and “How old were you when you had sexual intercourse for the first time?” are particularly exclusionary to sexual minority youths. LGB youths who are sexually active may appear to be sexually inactive or even appear to have never had sex. Important information is lost, and overall data reporting percentages of youths who are sexually active may be underestimates. How might a lesbian interpret and respond to the YRBSS question asking “The last time you had sexual intercourse, did you or your partner use a condom?” or the question asking “The last time you had sexual intercourse, what one method did you or your partner use to prevent pregnancy?” One has to wonder how a young woman having sex with other women might feel after reading this series of questions.

It is time for the CDC to add a standard sexual orientation question to the core set of questions all states and localities must collect. It is not enough to provide a monitary incentive. A decade after the recommendations of Sell and Becker (2001)⁸ the CDC continues to exclude sexual minority youths from data collection and reporting processes that could ultimately save lives. But more importantly, it is time for HHS to demonstrate leadership and make sure the current disarray in sexual orientation data collection does not continue. We can no longer allow any publicly funded data sets, not just those in the YRBSS, to exclude sexual minorities. We need to recognize how findings may be biased and skewed if we continue to omit sexual orientation data collection, and we need to review surveys and data collection instruments to make sure questions that are insensitive to LGB health are not included.