Introduction

Abstract

Standardizing populations is one of the central tasks of modern states. But a map of a dominion's territory is of little state use if not accompanied by a detailed inventory of its peoples. To that end, states seek to enumerate populations with some specificity: number of people per household, age, sex, race, ethnicity, birthplace, work status, income, education, health status, languages spoken, and so on. Cataloguing the population according to its properties not only facilitates the exercise of the state's broad police powers to regulate the population, often to the detriment of minoritized groups, but can also, in the best-case scenarios, help ensure the safety, health, and welfare of the people.¹Enumeration can be a somewhat misleading term for this process to the degree that it suggests one is merely counting objects that already exist, for the enumerative process in fact has the capacity to create what it purports only to name, causing new kinds of people to appear on the social map—sometimes drawing attention to their needs and redirecting resources to address those needs, sometimes concentrating recalcitrant, wayward, messy, incoherent, and noncompliant types of bodies and subjects deemed inconvenient to or obstructive of ends in categories that facilitate containment and control. Transforming “a social hieroglyph” of “exceptionally complex, illegible, and local social practices” into “a legible and administratively convenient format” requires the creation of standardized classifications and metrics for measuring them (Scott 1998: 3). To regularize a population is to flatten its zoetic confusions of movement and form, of time and space, of doing and being, into neat two-dimensional axes specifying static properties and numbers. This process requires, in short, fixing people and their purported attributes in place—“making up people,” in Ian Hacking's now famous phrase (1986). But which properties should matter? Precisely what and who should be counted? Or should the counting practices themselves be resisted?

For people who share some perceived commonality, the decision to enable official recognition of the numerical sort can mark a group's entry into “statistical citizenship” (Hannah 2001: 516). That is why debates about population enumeration are often so politically charged. In the United States, for example, while the Census Bureau has collected data on the number of households with same-sex couples for some time, that information was released to the public for the first time only after the 2010 census. Previous administrations declined to make it available on the premise that doing so would violate the Defense of Marriage Act (Sherman 2009). In reversing this policy and releasing the numbers of same-sex households, the Census Bureau director appointed by President Obama proclaimed, “We understand how important it is for all groups to have accurate statistics that reflect who we are as a nation” (US Census Bureau 2011). While the 2010 decennial census did not yet ask about sexual orientation directly—if you were not partnered and living together, you were not counted—questions on sexual orientation were added to one of the federal government's largest data collection instruments, the US Centers for Disease Control's National Health Statistics Survey, in 2013 (Ward et al. 2014). Unfortunately, despite the urging of advocates, federal policy makers declined to add measures to identify trans respondents on that survey, continuing the exclusion of trans people from the official national imaginary.

In the absence of research conducted by government agencies, trans activists have been producing surveys and needs assessments of their communities for decades, documenting the characteristics of these populations, including their demographics (race, ethnicity, gender, education, income, and so on) and their experiences (access to transition and non-transition-related health care and the frequency and type of violence and discrimination they experience) (e.g., Singer, Cochran, and Adamec 1997; Wilchins, Lombardi, and Priesing 1997). The largest survey of trans people to date in the United States—the National Transgender Discrimination Survey (NTDS), conducted largely over the Internet by the National Center for Transgender Equality and the National Gay and Lesbian Task Force—generated a convenience sample of 6,456 transgender people (Grant, Mottet, and Tanis 2011: 15). (To learn more about how the survey identified transgender people as well as other aspects of it, see Reisner et al., Labuski and Keo-Meier, and Harrison-Quintana, Grant, and Rivera, in this issue.) It found that 47 percent had experienced an “adverse job outcome,” 41 percent lacked an identity document that matched their gender identity, and 20 percent had served at one time in the armed forces (Grant, Mottet, and Tanis 2011: 3, 5, 30). Working off the NTDS findings, the Williams Institute, an LGBT think tank, claimed that ten states with strict voter identification laws could disenfranchise 24,000 transgender people in the November 2014 elections and that 149,800 transgender people have served or are currently serving in the military (Herman 2014: 1; Gates and Herman 2014: 1). These findings suggest that more resources should be directed to health care for transgender veterans and that policies for changing identity documents should be changed.

But using the NTDS results to make larger inferences about the number of transgender people in the United States affected by particular state policies or decisions about resource allocation, as these reports did, requires having credible statistics generated by random sampling about trans prevalence in large population-based surveys. Unfortunately, there are no large-scale, high-quality data on these questions. For the Williams Institute to produce the numbers cited above, demographer Gary Gates had to extrapolate from two state-level surveys and a handful of meta-level analyses to come up with an estimated transgender prevalence rate of 0.3 percent of the adult US population (about 700,000 transgender people, by this reckoning) (Gates 2011: 6). Gates himself notes that this methodology has “substantial limitations” (Chalabi 2014). Moreover, as Scout, a researcher and health policy analyst who works on LGBT issues, notes, findings based on community and regional surveys “just don't carry the same authority as federal data in grant applications, policy decisions, and resource allocation”; as a result, LGBT, trans, and health advocates have been “beating this drum for data collection” for years (Scout 2013). Regardless of how nonconforming people name themselves or are named by others, they are not yet represented in the official numbers of any state. Only Nepal and India have attempted to measure these subpopulations in national survey instruments (see Knight, Flores, and Nezhad's article in this issue; Nagarajan 2014).

We chose “Making Transgender Count” as the title of this special issue of TSQ because it contains a number of telling puns. On one hand, one makes trans count (in the sense of having its importance recognized) by counting it (making it visible through quantification). On the other hand, one makes (i.e., compels) trans count by forcing atypical configurations of identity into categories into which they do not quite fit—the proverbial square peg in a round hole. In this way, the imperative to be counted becomes another kind of normativizing violence that trans subjects can encounter and hence another problematic to be critically interrogated by the field of transgender studies. The tensions among what to count, whom to count, how to count, why to count, or whether to count or be counted at all are explored in the articles that follow. What makes the notion of trans* such a fecund point of departure for work in transgender studies is that the definitional lines of the concept are “moving targets” (Stryker, Currah, and Moore 2008: 13). That very instability frustrates the project of fixing embodied identities in time and space—a requisite operation for the potentially life-enhancing project of counting trans populations and better addressing their needs as well as for the necropolitical project of selecting certain members of the population for categorical exclusion as dysgenic. The essays in this issue do not resolve the tension between efforts to refine techniques of governmental reason and strategies of resistance, between attempts to sedentarize trans identities and movements that refuse such settling, or between universalizing imperatives to classify and local demands to reject incorporation into a global schematics of gender difference organized by male/female, man/woman, cis-/trans-, trans-/homo-, or white/color dichotomies. Some attempt to do both, while all ultimately fall on one side or the other of various problematics. Our goal in curating this issue has been less to gather a collection of articles that definitively settle these vexed questions in ways with which we might both agree than to stage a conversation in which the stakes of the game are made visible.

It is usually the case that people who conceptualize transgender as a fluid state of being resist efforts to quantify it. In this issue's lead article, “The (Mis)Measure of Trans,” Christine Labuski and Colton Keo-Meier argue that “asking questions from within this instability” can generate more useful findings than bracketing that fluidity, the usual practice in quantitative analysis. The authors advise us to reject a number of a priori ideas that usually accompany empirical and quantitative research on this topic. By jettisoning the assumptions that transgender is a “stable and measurable ‘thing,’ ” that trans and nontrans are mutually exclusive categories, that there are “aspects of trans identity … incommensurate with nontrans experience,” researchers will be better able to identify the mechanisms at work in producing hierarchical social difference in a given milieu. Indeed, reifying transgender by attempting to measure the effects “it” produces can obscure situations in which other axes of identity and experience might matter more, such as race, gender, or sexual orientation. In other contexts, the authors point out, transgender and other such broad categories should be disassembled because they can function as “a proxy for something far more specific and amenable to measurement,” something that could be better grasped by a more precise question about “what is to be learned from a specific … population.” For example, rather than setting FTM as the independent variable, one might choose “ ‘people assigned female at birth taking exogenous testosterone at X dose’ or ‘trans men who live in cities with large transgender populations.’ ” Drawing on the work of feminist science studies scholar Karen Barad and other “new materialist” approaches, the authors remind us that the conceptual apparatus of any particular trans research project (e.g., the definition of transgender, the tools for measuring phenomena, and the cultural, anatomical, or psychological frame of reference) is not external to causes and effects it observes but part of the phenomena it purports to measure (Barad 2007: 143).

In most publicly funded data collection, “sex/gender” is most definitely treated as an a priori. (Researchers not interested in trans populations typically use both terms interchangeably, making no distinction between the sex one is assigned at birth and one's gender identity.) It is generally measured by self-reports on survey instruments (circle M or F) or by telephone interviewers who decide (circle M or F) based on the interviewee's voice. Trans advocates can demand that birth sex and gender identity be disaggregated so that trans people can better be counted, but those demands need to be accompanied by assurances that there are reliable measures to identify this population. Members of the Gender Identity in US Surveillance (GenIUSS) Group at the Williams Institute, a multidisciplinary group of researchers interested in technical, political, and epistemological issues related to measuring trans populations, have been testing various questions to find out which ones work best. In “ ‘Counting’ Transgender and Gender-Nonconforming Adults in Health Research,” some members of the GenIUSS Group (Sari L. Reisner, Kerith J. Conron, Scout, Kellan Baker, Jody L. Herman, Emilia Lombardi, Emily A. Greytak, Alison M. Gill, and Alicia K. Matthews) present their findings regarding what they consider to be the most effective questions for distinguishing “gender minorities” in population-based surveys. Interestingly, the group finds that general-population survey questions designed to capture data about gender minorities must, in order to prevent false-positive responses that mistakenly inflate transgender prevalence, be “geared toward the majority group” (cisgendered respondents) such that they are easy for them to answer correctly.²

The GenIUSS Group report also highlights another obstacle—the problem of transgender collectivities' being “too small” to count in the US context. They point out that government agencies tend to resist adding questions to general-population surveys that would glean information from any group thought to comprise less than 0.5 percent of the population. That means that individuals who share some potentially important commonality but number fewer than 1.6 million are not likely to be counted. In fact, the authors point out, “federally funded national surveys collect and report data on subpopulations and health conditions that affect far fewer than this number of citizens.” For example, those who fall under the category Native Hawaiians and other Pacific Islanders have been counted in national surveys, and those data help health agencies direct the appropriate resources and information to those communities (AHRQ 2010; CDC 2014). Given that transgender people, especially trans women of color, have some of the highest known rates of HIV infection, according to reports from community groups and public health departments (CDC 2013), it would seem only logical to include “gender minorities” on general-population surveys.³ Counting this population and deploying resources to those most in need within it—trans women of color—could do much to lower their morbidity and mortality rates. But those lives are not objects of the biopolitical project of measuring the population to foster life. Overlooking them in the state statistics suggests that these lives fall on the letting-die side of the biopolitical project.⁴ Federal neglect of the health crisis among trans women of color epitomizes Ruth Wilson Gilmore's construction of racism as the “production and exploitation of group-differentiated vulnerability to premature death” (Gilmore 2007: 28).

In his contribution to this issue, T. Benjamin Singer observes how gender-variant and trans-identified people continue to be erased by medical knowledges and practices still organized around a naturalized gender binary. He makes a convincing case, however, that neither the violence of categorical exclusion nor the flattening effects of a unitary and reductive identity category (transgender) can account for what he actually found during fieldwork in community health settings and in analyzing trans-specific health care needs-assessment studies: a rapidly emerging, seemingly endless stream of emergent “embodiments and identities that exceed familiar sex and gender categorization,” which Singer characterizes as constituting a proliferative matrix structured by the “generative capacity of the category transgender itself.” His research refuses the facile dichotomy that would have us believe that activism, resistance, or liberatory critique can come only from outside institutional frameworks and that institutions can only homogenize and consolidate identities in reactionary ways. He goes so far as to suggest that even within the apparently immobilizing forces of institutions, even as institutional strategies attempt to contain unruly bodies and subjects within administratively useful categories, “bodies can and do exert an insurrectionary pressure upon the enumerative practices designed to produce and define them.”

Given the growing call of trans advocates to develop metrics for measuring trans populations, Kristen Schilt and Jenifer Bratter set themselves the task of asking two questions of trans people: Do they want a transgender option on the census? If there were such an option, would they choose it? In their article, “From Multiracial to Transgender? Assessing Attitudes toward Expanding Gender Options on the US Census,” they frame these questions with social-science findings on identity validation, which has been used to explain people's choices concerning whether or not to identify as multiracial. (A multiracial option was added to the US Census in 2000 when individuals were allowed to choose more than one racial category.) Their findings, based on a convenience sample of 167 individuals attending transgender or genderqueer conferences, produced some surprising results: 41 percent of the respondents supported adding a transgender option to the US census, but only 29 percent actually reported identifying as transgender on forms when given the option. The fact that forms (from doctors' offices, etc.) are not anonymous while the census is may account for part of that discrepancy. But digging deeper into these responses, the authors note that identity validation played an important role, as it had with individuals choosing the multiracial option. If one's gender identity was socially validated (by others), they were “far less likely to endorse adding a category to the US Census or to use such a category on official forms.” But those whose gender identity was not validated socially, or those whose transgender identity was validated, were more likely. These findings suggest that a significant proportion of people identified by social scientists and advocates as “transgender” might not want to be seen occupying a position outside the gender binary.

In 2007, the Blue Diamond Society petitioned the Nepalese government to repeal all laws and legal and institutional arrangements that discriminated on the basis of sexual orientation and gender identity. Responding to Blue Diamond's claim that “even the state has ignored us,” the Supreme Court directed the government to recognize “third gender” (tesro lingi) people by, among other things, including that option on the census form (Pant v. Nepal, Writ No. 917 of the Year 2064 BS [AD 2007], translated in NJA Law Journal; see “Decision of the [Nepal] Supreme Court” 2008). Accomplishing that task, however, proved easier said than done. Nepal's attempt to add a third gender category to the national census—the first such attempt anywhere—was largely deemed a failure: the category was not clearly defined, the enumerators lacked training, and the census software used to tabulate the data was not retrofitted to include the third gender option. In “Surveying Nepal's Third Gender,” Kyle G. Knight, Andrew R. Flores, and Sheila J. Nezhad discuss their research on how the census might be improved in the future. Working with the Blue Diamond Society, they conducted focus groups to develop and refine a survey instrument that was used on 1,178 members of sexual and gender minorities in Nepal. One of their most significant findings was that tesro lingi, conceptualized as a catch-all category similar to transgender, may not be an adequate term for accurately representing all of the seven sex/sexuality/gender identity descriptors they found that people applied to themselves. Moreover, the neat analytic distinctions between sexual orientation and gender identity that have characterized much Western scholarship do not hold in the Nepali context. Instead, they point out, any locally relevant survey instrument has to account for “the fluidity and hybridity of sexual and gender terms used in Nepal (as elsewhere).”

Anna M. Kłonkowska's contribution to this issue, “Making Transgender Count in Poland: Disciplined Individuals and Circumscribed Populations,” could also be titled “Making Transgender People Disappear in Poland.” In this article, she documents how a medical model of transsexuality governs not just individual transgender lives but also their numerical representation in the population. As she notes, “transposed onto the regulatory realm, the norms governing whether, how, and when individuals in Poland may transition from one gender to another also become metrics for estimating the number of transsexuals in the population.” As a result, only physically transitioned bodies come to matter, rather than gender identifications, behaviors, or self-presentation. For those who desire to transition physically, very few are able to overcome the many legal, medical, and bureaucratic obstacles. The inevitable result is that the vast majority of trans people in Poland simply are not “legible as trans,” as the authors put it; out of a population of 38.5 million, only 203 people received a court order allowing them to change their gender marker between the beginning of 2009 and the end of 2012. Kłonkowska's work, based on interviews with trans people in Poland, documents their strategies of resistance and assimilation and provides a vital counterpoint to the official ignorance with respect to trans lives.

Large entities such as government agencies and private insurance companies have a stake in smoothing out differences between members of a population such that individuals become more or less interchangeable units, represented and manipulated in standardized ways. In contrast, smaller, community-based health and social-service organizations want and need to know who their patients/constituents are in order to give them the particular kind of care they need. But because delivering that care requires reporting data about their patients to outside agencies and corporations, LGBTQ health centers must come up with ways to balance the need for specificity in one context with the need for homogenization in another. In “Counting Trans* Patients: A Community Health Center Case Study,” Natalie Ingraham, Vanessa Pratt, and Nick Gorton discuss how Lyon-Martin Health Services in San Francisco has responded to this dilemma. In addition to relying on the two-step method (asking about the sex assigned at birth and current gender identity), clinicians at Lyon-Martin have developed other strategies that allow them to respect their patients' gender identities and to translate the data in ways that are legible to outside entities. Any solutions, however, are only provisional. Seemingly mundane yet often intractable technical matters, such as codes in electronic medical records or insurance company billing software, can thwart service-provider attempts to recognize and validate the self-understanding and self-presentation of their patients or constituents.

The violence of global and globalizing categories is made evident in Hale Thompson and Lisa King's contribution to this issue. In “Who Counts as ‘Transgender’? Epidemiological Methods and a Critical Intervention,” they show how collapsing a wide range of differently gendered subjects under the transgender category makes invisible the economic, political, and social processes that “contribute to the marginalization and invisibility of trans and gender-nonconforming individuals within the HIV prevention and treatment complex.” In relation to HIV, fixing “transgender” in place as a discrete category reinforces the idea that it is being “transgender” that causes vulnerability to premature death, rather than the many structural forces that increase trans people's “risk of risk” and create conditions for what Lauren Berlant has called “slow death” (Berlant 2007: 754). Disability rights activists have long rejected the idea that disability originates in individual bodies and have favored instead a model that understands disability as an effect of legal, social, medical, and physical landscapes that privilege the normatively able-bodied. Drawing on the same logic, Thompson and King argue that the target of interventions should be the normalizing institutions, discourses, and apparatuses rather than trans people themselves.

The social norms embedded in information systems comprise a crucial but often overlooked area of contestation. In a research note, “Information Systems and the Translation of Transgender,” Jeffrey Alan Johnson expands on the discussion begun in Ingraham, Pratt, and Gorton's article and examines how the complexity of gender nonconformity gets erased in the sociotechnical processes of the “translation regime” through which data systems interpret and construct the world. He uses Utah Valley University's data systems as a case study. At one point, the system included “U” (for “unspecified”) as a value in its gender field, in addition to “M” and “F.” After state legislators became aware of an effort to install unisex restrooms in support of gender-nonconforming students at Utah Valley University, the U was “deprecated” (removed). In this political context, U values were no longer seen as a way to represent individuals who did not see themselves as M or F; the U values were instead seen as “bad data.” Although there are no technical barriers to expanding gender possibilities, that the U did not mesh with the Utah System for Higher Education's data standards (M or F only) also provided a technical justification for the political decision policing the gender binary.

In the “Manifestos” section that closes the special issue–themed articles, we present two appraisals of efforts to make transgender count. In “Boxes of Our Own Creation: A Trans Data Collection Wo/Manifesto,” Jack Harrison-Quintana, Jaime M. Grant, and Ignacio G. Rivera reflect on their experiences working on the National Transgender Discrimination Survey, the largest survey of trans people to date in the United States. Most significantly, they discuss the qualifying questions through which the NTDS constituted the “transgender” population. With its survey instrument based on a grassroots research process, the questions were meant to serve as a “radical welcome” to participate in what would turn out to be an important tool for trans advocacy in the United States. The 6,456 trans and gender-nonconforming people who decided to respond helped to create “the largest quantitative data set on trans experience anywhere in the world.”

Switching attention from the output to the input side of the survey process, from the utility of the findings to the interests of those contributing the data, the Reverend Megan M. Rohrer, a longtime San Francisco–based activist, offers an alternative view. In “The Ethical Case for Undercounting Trans Individuals,” they point out that asking vulnerable individuals to answer surveys about their bodies, their identities, and their behaviors can leave them “feeling pathologized, overexposed, and abnormal.” Moreover, because needs-assessment data collection often takes place in social-service settings where trans people are seeking food, shelter, or medical care, they might perceive requests to answer questions as a necessary part of a transaction in which they risk being denied a necessity. Rohrer argues that vulnerable people, people who are low- or nondisclosers, and people who might be classified by others as trans but who do not identify as such should have the right not to be counted. Finally, because the attempt to quantify trans communities requires radically minimizing or erasing our many differences by imposing a static and neat order upon them, Rohrer suggests we forgo trying to “capture in statistics” a “community that cannot be generalized.”

We are deeply grateful to those who contributed to the editorial process of “Making Transgender Count”: the anonymous reviewers whose comments and suggestions were invaluable; Somjen Fraser, Jeff Mathias, and Amaya Perez-Brumer, whose timely help at crucial stages in the editorial process was invaluable; and our diligent and ever-patient editorial assistant, Abraham Weil, whose title should actually be TSQ Whip.

References