Knowledge is power. True, yet collecting accurate knowledge on our LGBT communities is often a struggle. This is not because we cannot collect such information but usually because those shaping surveys and marshaling clip-boards are either not interested in learning about LGBT people, ignorant of the needs of our communities, or plain hostile to us.
In this installment on "The Recession and LGBT Communities," I address the issue of information available in order to merely pose questions such as, "How are our lesbian, gay, bisexual, and trans people faring in the recession?" The blog comments on my first entry in this series on the recession and LGBT communities introduced the thorny problem of researching LGBT persons and basically made the point that trying to understand the recession's impact on our communities is a necessary and complex task.
After the jump, I will discuss some of the research problems with all of this. If you are not interested in this go to the concluding section of this blog called: "This Is Not Sexy."
We're here, we're queer, count US!
A blog commentator named Michael brought up some good points on the first blog in this series:
"I respectfully submit that because of the ongoing reluctance of many to self-identify as LGBT that there is NO SUCH THING as a survey on ANY topic that can possibly meet the minimum criteria for application to ALL LGBTs. Put in research speak: there is simply no way to determine a credible 'margin of error.'
Though I disagree that we cannot create reliable and valid surveys for lesbian, gay, bisexual and/or transgender folks, his point here brings up two important considerations. First, not everyone wants to identify based on sexual orientation and some trans folks simply identify as their gender or are "stealth." We should be sympathetic to these urges and researchers also need to adjust our methods accordingly.
This leads us to our second consideration: hidden populations. A population is said to be "hidden" if, "there is no public listing of its members, such as a voter registration roll, or a telephone directory. Sampling such populations is difficult because the standard procedures that ensure samples will be representative are not applicable to these populations," report Drs. Douglas and Broadhead in their 1996 publication on sampling advice for hidden populations.
Reasons abound for populations to be uninterested in being counted. These usually have to do with social stigma. Some purposefully avoid researchers because of past "research" abuses, like those done to African Americans in the US (see the book Medical Apartheid: The Dark History of Medical Experimentation on Black Americans from Colonial Times to the Present by Harriet Washington). Other groups are so dispersed or nomadic that they are not picked up in generalized sampling schemes, like migrant field workers.
Research techniques for identifying these folks use smaller, respondent-driven sampling techniques. For example, I interview you then ask you to refer your friends who are also L, G, B, or T, then they refer their friends, and the study "snowballs" into a valid sample size. Many are "incentivized" with little tokens like gas cards or small amounts of cash given for their participation. Some such projects are called, "community-based participatory research," which takes a collaborative approach to constructing and conducting research with a given target group. There are many articles by statistical whizzes showing that these and similar approaches are scientifically valid and rigorous.
Once many of these smaller studies are achieved, they are sometimes written up together in a meta-analysis to examine trends and summarize the areas where they agree and disagree. There is usually great care given to understanding how the methodological differences between studies impact the reliability of findings to emphasize what can and cannot be taken as a generalizable finding in the meta-study. When a social problem is identified and shows a need for information on hidden populations, such smaller studies are normally conducted to gather basic information. Larger studies are then attempted to answer these basic questions with many, many people involved. This is primary research. Some researchers use data that was collected for a different reason but has enough information on LGBTs available to meaningfully examine these issues.
The main problem with using secondary sources is the problem of data quality. Who asked the questions? Were they particularly partisan or homophobic/transphobic in their question construction? What categories were used for people to identify their sexuality and/or gender? Should we put people into categories? If categories were used, were people recoded or reclassified into another category? Who was kept in and who was "cleaned out" of the data? Where LGBTQ participants statistically "weighted" to increase their representation? What types of sensitivity testing was conducted to ensure this data was good enough to analyze?
This Is Not Sexy
I am intending to highlight some of the sticking points in getting at accurate data on our communities. The problem of hidden populations combined with social and political resistance for including such data variables as "sexual orientation" or "sexuality" or "romantic partner" or "gender identity" rather than, or in addition to, "gender" and "sex" give this a complex but not insurmountable set of circumstances that must be addressed.
We must demand better, more rigorous statistics on our communities. Asking for accurate statistics is not sexy: You say you're an activist--what do you do? "I ask universities, NGOs and government agencies to count me in their surveys," rather than "I fight hate!" "I am destroying the gender binary!"
Indeed, this might be a bean-counter approach to activism but it is something that is more than crucial. Like any of the other more sexy activist activities, this work takes persistence, fortitude, and hard-headed will over time. I asked one person with such fortitude about this vexing problem, economist M.V. Lee Badgett. As the Research Director at the Williams Institute, Dr. Badgett and her colleague Dr. Gary Gates are prominent researchers mining such data sources for information on our communities. I asked her whether there were any good data on non-coupled LGBT folks and her simple answer was "unfortunately no." Dr. Badgett did point out a useful source that I also recently stumbled upon: http://gaydata.org/. It is the brain child of Randall L. Sell, Sc.D. and is hosted through Drexel University's School of Public Health. The website "encourages the collection of sexual orientation data and the analysis of data sources that have already collected such data."
Because of the marketplace focus that began this series, I asked Dr. Badgett about the problem with considering all LGBTs as "DINKS" (Dual-income, no kids). To that she simply said, "They're not all in couples, for one thing." She says that the problem with characterizing LGBT people who are in couples is the suggestion that, "DINKs are just frivolously spending all of their money on themselves. But they also save for retirement, support LGBT organizations, supply lots of mainstream community organizations, and live in the economic shadow of the threat of discrimination." On my question of whether or not there is good data on the economic situation of LGBT folks, Dr. Badgett said, "I've looked really hard for data on how LGBTs are doing, including using the Witeck-Combs/Harris Interactive data and the current population-based survey from the U.S. Bureau of Labor Statistics. We can see unemployment rates rising among LGBs and people in same-sex couples, but the sample sizes of unemployed folks is just really small, mainly because the sample sizes of LGBs and 'same-sex' couples is pretty small. So it's been hard to do anything much in the way of analysis of those data."
Added to this problem of counting individual LGB persons, the studies on trans populations are sparse at this point. As an optimist on the subject, I anticipate that the queer-friendly departments in the academy understand this problem. I do understand that there are a good number of qualitative and some quantitative studies underway because of the fortitude of trans activists and their allies. Many of the studies on trans folks are qualitative (for many of the reasons discussed above) and focus on the effects on health, work and well-being from pervasive anti-trans discrimination.
One of my favorite researchers is Dr. Kristen Clements-Noelle, a public health researcher at the University of Nevada at Reno. Dr. Clements-Noelle is doing solid, trans-positive research. She is not the only researcher on trans topics out there but I would contend that she is too rare a bird for the amount of work we have ahead of us. There have also been needs assessment-type studies or health studies done on trans issues usually through social services organizations serving trans populations. In general, researchers need to be especially sensitive when designing such research to avoid creating false distinctions based on bias: trans folks can identify as lesbigay, just as LGBs can be people of color (and visa versa all the way around!). These are not exclusive terms but overlap in ways that impact who is counted and why.
So what do we know then? There is some data on LGBT people but economic data is particularly sparse. Data on couples are available from the Census and there have been a few larger studies on lesbians and gays. Smaller studies on trans and bisexual people also exist. To my knowledge, no population-based study has yet accurately captured individuals in these communities in their totality.
I want to ask readers to provide links to any studies on lesbian, gay, bisexual and trans people in the comment section of this blog. Perhaps we can start our own resource similar to "gaydata.org" and call it the more apropos "queerdata.org"?
I conclude with a no-brainer, all in all, there is much work to be done. And it's up to us to do it or push for it. This point brings me to a teaser for the next installment of this blog: the 2010 U.S. Census.