Williams Institute 10th Anniversary Conference on LGBT Law and Public Policy – Day Two, Part Two

This is my last blog from the Williams Institute Update Conference on the last session of the meeting.  US Census Director Robert Groves was scheduled to speak on LGBT data collection, but unfortunately had to back out due to anticipated government shutdown, even though the shutdown didn’t (yet) happen.  Fortunately, Gary Gates stepped in to talk about “LGBT data: politics, policy, and practice”–a subject I know readers of the Network’s blog are keenly interested in.  Gates is an expert in using data to estimate the population of LGBTs and same sex couples and describe their characteristics.  He’s frequently used Census and American Community Survey data (also from the Census bureau) in his work.

Gates started by characterizing the political challenges in collecting LGBT data: it’s seen as too hard or complicated to measure who LGBT people are, questions are seen as too sensitive and producing non-response or survey termination, and finally it’s seen as too politically volatile, particularly at the federal level.  He then outlined a case for collecting LGBT data:  to inform scholarship and public policy, to increase visibility, because a solid body of evidence indicates questions measuring LGBT status can be constructed with clarity.  Gates has published an op-ed on these issues in tomorrow’s Washington Post  of 4/10/2011.

Gates then reviewed transformations over time with how relationships were defined in the US Census.  In the 1800s, people could be single, married, widowed or divorce.  In 1950, the category of “separated” was added.  In 1980, “partner/roommate” was added, but there was no way to disentangle romantic relationships from the roommates.  In 1990, “partner/roommate” became “roommate” and “unmarried partner.”  While never the intention, this allowed analysts to identify same sex unmarried couples and made the Census the largest data source for examining this one slice of the LGB community.  He went on discuss challenges in using Census data to examine such couples – particularly the problem of error when different sex couples miscode the gender of one of the partners.  Gates and colleagues have studied the problem of error in the American Community Survey  and estimated that 30% of the same sex couples identified were actually miscoded different sex couples.  They successfully persuaded the Census to search for such errors and account for them through a variety of techniques.

Gates presented analysis of American Community Survey data (conducted every year) that shows a steady growth from 1990 in households identifying themselves as same sex unmarried partners.  Gates notes that conservative states have experienced the highest rates of growth.  For Gates, the increase could indicate a greater likelihood of people to create same sex cohabiting households, but more likely indicates a decrease in the size of the closet.

Gates has been a tireless advocate for increased LGBT data collection.  He reported on some recent progress at the federal level.  For example, he reported that the National Center for Health Statistics is cognitive testing questions on gender identity for possible inclusion on the National Health Interview Survey.  Further, Census is conducting a large-scale project to assess possible changes in the ACS household roster and marital status—potentially including terms like domestic partnership.  The challenges there are that heterosexual couples tend not to understand such terms.

Finally, Gates overviewed data from recent surveys.  He recently analyzed data from several national and state-based surveys and released a controversial estimate of the size of the LGB and T population at 9 million Americans or 3.8 percent of the population.  Gates further estimates that there are 1.16 million individuals in same sex couples, with 80,000 in legally married relationships. The analysis that Gates conducted found several persistent demographic patterns in the data: bisexuals are more likely to be women, younger people are more likely to identify as LGB, lesbian and gays report higher levels of education, LGBs are racially and ethnically diverse, and lesbians and gays are more out than bisexuals–particularly in the workplace.  Some patterns that Gates found starkly challenge stereotypes.  For example, he found that many same sex couples are disadvantaged –living in poverty or are on public assistance.  He also found that LGB child-rearing is more common in racial/ethnic minorities and more comon in the South.  He closed his talk with a recent story on CNN that highlighted a same sex family that is typical in his data –southern, black, and raising children—but particularly atypical in media representations of LGBT people.  Gates suggested that researchers and LGBT movement advocates should pay a lot more attention to families like the one profiled in the CNN story.

It’s been great to be out here in LA for this conference.  Thanks to the Network for the opportunity to guest blog while I’m here!  Hope to do it again soon.