More Information

Published Year: 2017
Survey/Program(s): Current Population Survey (CPS)
Catalog: Statistical Survey Paper

Jessica Holzberg, Renee Ellis, Matt Virgile, Dawn Nelson, Jennifer Edgar, Polly A. Phipps, and Robin L. Kaplan

Assessing the Feasibility of Asking About Gender Identity in the Current Population Survey: Results From Focus Groups With Members of the Transgender Population

PDF version

by Jessica Holzberg, Renee Ellis, Matt Virgile, Dawn Nelson, Center for Survey Measurement, U.S. Census Bureau
Jennifer Edgar, Polly Phipps, Robin Kaplan, Office of Survey Methods Research, Bureau of Labor Statistics

Disclaimer: This report is released to inform interested parties of research and to encourage discussion. The views expressed are those of the authors and not necessarily those of the U.S. Census Bureau or Bureau of Labor Statistics.

Executive Summary

This report presents findings from exploratory focus groups with self-identified members of the transgender¹ population that were conducted by the U.S. Census Bureau and the Bureau of Labor Statistics (BLS). Focus groups were held in four cities to explore the feasibility of asking about gender identity in the Current Population Survey (CPS). We gathered feedback from 29 transgender respondents in focus groups conducted between September 2016 and June 2017 in order to answer four research questions:

1. How do transgender respondents define “transgender” and “gender identity”, and relate these terms to their self-identity?
   
   

   Focus group respondents generally agreed that gender identity involves how someone feels and sees themselves, with some respondents commenting that the term “gender identity” makes them think of “anyone who is not cisgender” or “cis.”²

   They were also generally able to agree that to be transgender means to be the opposite of cisgender. While some respondents said that they would not use the word “transgender” to describe themselves personally, most agreed that it could be used as an umbrella term to describe members of a diverse community.

   Respondents in the focus groups self-identified in a variety of ways, for example, man, woman, transgender, queer, gender-fluid, non-binary, and genderqueer. Some respondents indicated that their self-identification has changed over time or that it may change in the future, a process one respondent described as “fine tuning their own self-description.”

2. How do transgender respondents feel about the collection of gender identity information by the Federal government, both generally and in the context of the CPS?
   
   

   Most respondents thought that government surveys should include gender identity questions for reasons including getting a count of the transgender population, using this count to advocate and allocate funding, and understanding potential discrimination, amongst other things. However, there were a few comments in each group from respondents expressing skepticism that the data would be used for these positive purposes or general distrust of the government. The question of who in the government would be using the information was raised in two of the groups.

   A later explanation of the reasons for possible collection in the CPS was reassuring for respondents, and the idea of collecting gender identity information for the purposes of the CPS received wide support across the focus groups. A few respondents talked about it being useful to see how unemployment differs between people who are transgender and not transgender, but most had just general reactions to the idea as a whole. Nonetheless, most of the respondents did have some sort of concern about how this data would be collected by the Federal government and implemented in the CPS and how the information would be used in this context. Concerns voiced by respondents in all four groups included leak of information outside of the government, general malicious use, and confidentiality/anonymity.

   Some respondents were concerned about having their gender identity tied to their name or other identifying information, and some respondents said that they do not always disclose their gender identity in order to protect their safety, avoid inappropriate questions, or prevent harm to their career prospects. Concerns about general malicious use and confidentiality/anonymity were amplified in the three focus groups that were conducted in March, April, and June 2017. In Portland, there were also concerns about interviewer misconduct resulting in personal information being shared in non-official ways.

   Across all four groups, several of the respondents who shared concerns about the government collecting information about gender identity also explicitly stated that they support the idea of it, suggesting there may be internal tension between understanding the value of having statistics on the transgender population and hesitating to provide their own personal information.

3. How do transgender respondents feel about gender identity information being collected via proxy response?
   
   

   The CPS uses proxy response, in which one person generally responds for all eligible members of the household. If implemented, gender identity questions would be asked by proxy. The immediate reaction to the idea of proxy response was negative in three out of the four focus groups. Some respondents explicitly asked why proxy response had to be used. Overall, the two respondents in the Nashville focus group were okay with the use of proxy response, and there were a few respondents in each of the other groups who felt similarly.

   Respondents were concerned about issues with accuracy and the inappropriateness of answering gender identity questions on someone else’s behalf. While not all of these concerns applied to respondents personally, they reported that they knew people for whom proxy response would be problematic.

   A handful of respondents in each of the DC, Portland, and Fargo groups said household members would refuse to report gender identity on their behalf. Very few respondents thought that members of their household would answer accurately. They were concerned about the CPS respondent being unaware of household members’ correct gender identity or refusing to accept it. Parent-child and unrelated roommate relationships were cited as being particularly difficult.

   Most respondents felt proxy response of gender identity was sensitive, citing concerns about general confidentiality, potential for sharing person information that the person would not want shared, and safety.

   Respondents were split on whether it would matter who in the household was answering on their behalf. Most of those who said it did matter commented on the inability of family (particularly parents) to correctly answer questions about their gender identity; however, the answer was likely to be inaccurate regardless of who in the household was giving the proxy response.

4. What feedback do transgender respondents have on wording of questions about gender identity?
   
   

   Respondents in all four groups gave strong reactions to the presented survey question wording. The presented wording included a one-step question on current gender identity, as well as four variations of a two-step question on both sex assigned at birth and current gender identity. The latter four questions differed in the number and specificity of response options offered for current gender identity.

   Some respondents explicitly said that they thought it would be difficult for researchers to create questions with adequate response options, given the diversity of terms used and debate within in the transgender community itself about terminology. Respondents saw shortcomings with all five versions of the questions that were presented, but the question with the largest number of current gender identity response options was seen as the most promising. Not all respondents were clear on how some of the questions would be used to identify transgender respondents.

   There were some respondents who were uncomfortable selecting a “transgender” response option to describe themselves. There were also respondents who were unsure, unwilling, or uncomfortable identifying as transgender instead of male or female. General question criticisms and comments included the inability to mark all that apply, lack of adequate response options, use of “cisgender,” and the language of the question stems.

Conclusion

Overall, feedback from focus group respondents highlighted a number of obstacles to accurate collection of gender identity in the CPS through proxy response, for which there is no clear solution. Beyond just finding question wording that is understandable and relatable for the heterogeneous transgender population, there are also concerns about the accuracy of reporting and sensitivity of the task. It is very important to consider these obstacles carefully, as they have the potential to create serious classification errors. That said, the transgender respondents we spoke to agreed that gender identity would be valuable information to have available on a large-scale government survey.

We find it important to emphasize that these focus groups are exploratory, and we were limited in our ability to fully explain the CPS to respondents and the context in which gender identity questions would appear. Respondents were not administered the CPS interview, and thus we were unable to see how transgender respondents would react to being asked to answer these questions in context. Therefore, opinions expressed in the focus groups were more hypothetical in nature. Opinions may or may not be predictive of behaviors (Fazio, 1986; Horwitz & Finamore, 2017). Additionally, the group dynamic of focus groups shapes respondents’ comments and perspectives (Krueger, 1998). Respondents are influenced by each other, and respondents who are quiet may or may not agree with other opinions voiced. This is further discussed along with other study limitations in Section 5.2.

These focus groups were just one part of a larger study on the feasibility of asking about gender identity on the CPS. A decision on overall feasibility of collecting SOGI information in the CPS should consider the findings of the focus groups as well as those of the cognitive test on gender identity questions (Ellis et al., 2017).

1. Introduction

In 2015, the Office of Management and Budget (OMB) convened a working group of more than 100 representatives from 35 federal agencies across 14 departments and 7 independent Federal agencies (Park, 2016). The purpose of the group is to share knowledge on the development and testing of questions on sexual orientation and gender identity (SOGI) in both Federal and non-Federal surveys in the United States (Federal Interagency Working Group, 2016a). The goal of including SOGI data in Federal surveys is to allow researchers to estimate the size and distribution of the sexual and gender minority populations in the U.S., and to identify disparities between people who identify as LGBT (lesbian, gay, bisexual, and transgender) and those who do not in domains such as health, crime, or employment. Currently, there are 11 Federal surveys that collect data on sexual orientation, and of these, seven also ask about gender identity (Federal Interagency Working Group, 2016a). These surveys vary on features such as question wording, mode of survey response, and population being surveyed. In terms of context (i.e., primary topic of the survey), most are health surveys; the exceptions are the National Inmate Survey (NIS), the Survey of Prison Inmates (SPI), and the National Crime Victimization Survey (NCVS). All of these surveys ask respondents to report SOGI for only themselves—not other household members.

The OMB working group has so far issued three working papers on the landscape of SOGI questions in Federal surveys: one on current SOGI measurement in federal surveys, a second on evaluations of these Federal SOGI measures, and a third on research priorities moving forward (Federal Interagency Working Group, 2016a, 2016b, 2016c). In this latter paper, the group identified proxy reporting and question wording as primary research priorities for SOGI questions and survey context as a secondary research priority (Federal Interagency Working Group, 2016c). The working group has also recently taken an interest in asking about SOGI on the Current Population Survey (CPS), which serves as the primary source of labor force statistics for the U.S. population. The CPS is sponsored jointly by the Bureau of Labor Statistics (BLS) and the Census Bureau. The CPS differs from the surveys currently collecting SOGI information in two important ways. First, the context is employment, as opposed to health, which may affect respondents’ perceptions of the relevance of SOGI questions. Second, a single household respondent answers questions about all other household members. Thus, in households with two or more members, household respondents provide self-reports for themselves, and also proxy report for other household members. It is unknown whether respondents are able to report SOGI accurately by proxy and whether they feel comfortable doing so.

Because of these differences, the U.S. Department of Labor’s Office of the Assistant Secretary for Policy sponsored research to explore the feasibility of asking about SOGI in the CPS. The overall goal was to assess the feasibility of collecting SOGI data in the CPS setting—that is, an employment survey context with proxy reporting. More specifically, we wanted to examine: (1) the sensitivity of the questions in general; (2) whether household members have the knowledge about each other with regard to SOGI questions and are willing to provide those answers, and (3) reactions to the inclusion of SOGI questions in an employment survey. To address these goals, cognitive interviews were conducted with both LGBT and non-LBGT populations and focus groups were conducted with individuals who identified as transgender. Results of the cognitive interviews are provided in a separate report (Ellis et al., 2017); the purpose of this report is to document results of the focus groups.

The purpose of focus groups is to gather feedback from respondents on a given topic. Focus group respondents participate in a group discussion led by a moderator, who asks respondents open-ended questions and probes the respondents on their answers. A total of four exploratory focus groups were conducted with 29 self-identified members of the transgender population. The focus groups were designed to collect respondent feedback on the Federal collection of gender identity information specifically, and were used to answer four research questions:

1. How do transgender respondents define “transgender” and “gender identity,” and relate these terms to their self-identity?
2. How do transgender respondents feel about the collection of gender identity information by the Federal government, both generally and in the context of the CPS?
3. How do transgender respondents feel about gender identity information being collected via proxy response?
4. What feedback do transgender respondents have on wording of questions about gender identity?

2. Background

2.1 Current Population Survey

The CPS is conducted monthly by the Census Bureau on behalf of BLS with an annually-selected probability sample of about 60,000 occupied households in all 50 states and the District of Columbia.

The survey excludes those living in institutions, such as prisons, long-term care facilities, and nursing homes. One person (age 15 or older), known as the “household respondent,” generally responds for all household members and is usually a person who owns or rents the housing unit. Households are in the survey for four consecutive months, out of sample for eight months, and then return for another four months before leaving the sample permanently, for a total of eight interviews. In terms of content, the CPS consists of a basic monthly survey, to which a supplemental or topical module is added in most months. The basic monthly survey is divided into two parts: basic household and demographic information (e.g., date of birth, marital status) and labor force information (e.g., employment during the past week). Supplements cover a range of topic areas including annual work activity and income, veteran status, school enrollment, and volunteerism, among other topics (U.S. Census Bureau, 2017a).

Most of the household and demographic questions are asked only once, in the first interview. In subsequent interviews, respondents are asked to confirm that members of the household previously reported are still living there. If someone has left the household, they are removed from the household roster; if someone has moved into the household, demographic questions are asked for that person. Respondents are also re-asked demographic questions if they previously reported they did not know the answer (but if they refused initially, the question is not re-asked). Some questions (such as educational attainment and disability) are re-asked of all respondents in select subsequent interviews regardless of prior responses because the answers could change over time. The CPS is administered through both computer-assisted personal interviewing (CAPI) and computer-assisted telephone interviewing (CATI). The majority of the first and fifth CPS interviews are conducted through personal visits, with other interviews often being conducted over the telephone.

Currently there are no questions on gender identity in the CPS. There is a question on sex, in the demographics section of the basic monthly survey. Household respondents are asked to respond for themselves and all other household members, regardless of age. The question is worded as shown below (note that for this and all items tested, responses of “don’t know” and “refused” are not explicitly displayed or read to the respondent, but they do exist as response categories):

CPS Current Question on Sex
What is [NAME’s] sex?

• Male
• Female

Interviewers are instructed to ask this question “only if necessary” (U.S. Census Bureau, 2017b). If implemented in the CPS, gender identity questions would likely replace the current sex question. Because of this placement at the beginning of the interview, there is concern that respondents may refuse to answer gender identity questions and breakoff from the rest of the CPS.

2.2 Gender Identity

According to a multidisciplinary expert panel convened by the Williams Institute at the University of California, Los Angeles, known as the Sexual Minority Assessment Research Team (SMART), gender identity is “A person’s internal sense of gender (e.g. being a man, a woman, or genderqueer) and potential affiliation with a gender community (e.g. women, trans women, genderqueer)”³ (Federal Interagency Working Group, 2016a). Throughout this report, we use “transgender” as an umbrella term to refer to “anyone whose gender identity differs from their sex assigned at birth” (GLAAD, 2017).

2.3 Measurement of Gender Identity in Surveys

There has been limited testing of question wording on gender identity, but studies thus far have demonstrated that transgender respondents are generally able to understand and answer questions on the subject (Baker & Hughes, 2016; Lombardi & Banik, 2016; Reisner et al., 2014; Cahill et al., 2014). In 2014, the Williams Institute issued a report addressing best practices for asking about gender identity in a survey based on research by a multidisciplinary expert panel known as the Gender Identity in U.S. Surveillance (GenIUSS) Group. The group recommended using a two-question approach by first asking for a respondent’s assigned sex at birth, and then asking for their current gender identity (GenIUSS group, 2014). Other organizations, such as the Center of Excellence for Transgender Health (CoE) and the World Professional Association for Transgender Health (WPATH), also recommend using a two-question approach, but they suggest asking about current gender identity before sex assigned at birth (Federal Interagency Working Group, 2016b). Of the seven federal surveys that ask about gender identity, three of them—the National Crime Victimization Survey (NCVS), the Survey of Prison Inmates (SPI), and the National Adult Tobacco Survey (NATS)—use the two-question approach. All ask the sex-at-birth question first, and all use slightly different wording and response categories, as shown:

Question 1: Sex at Birth
NCVS and SPI: What sex were you assigned at birth, on your original birth certificate?
NATS: What sex were you at birth? [Field Interviewer Note: “Did they tell you that you were born male or female?”]

• Male
• Female

Question 2: Current Gender Identity
NCVS: Do you currently describe yourself as male, female, or transgender?

• Male
• Female
• Transgender⁴

SPI: How do you describe yourself (select one)?

• Male
• Female
• Transgender
• Do not identify as male, female or transgender

NATS: Do you currently consider yourself to be:

• Male
• Female

In terms of measuring prevalence, under a two-question approach, respondents would be identifiable as transgender if they selected different options for their sex at birth and for their current gender identity. In the NCVS and SPI, those who chose “transgender” would also be added to the tally. The NCVS also includes a follow-up question that has interviewers verify they are recording the correct answer if there is a mismatch between the two questions:

[IF Q1?Q2] Just to confirm, you were assigned [male/female] at birth and now describe yourself as [male/female]. Is that correct?

• Yes
• No

The other four federal surveys—the Behavioral Risk Factor Surveillance System (BRFSS), the National Inmate Survey (NIS), the Population Assessment of Tobacco and Health (PATH), and the Health Care Patient Survey (HCPS)—use a one-question approach that asks directly about transgender status (though the PATH survey first asks a yes/no on transgender, then a follow-up on type of transgender). The HCPS also has a follow-up question for those who identify as genderqueer or “other” to determine skip patterns for subsequent survey questions. As with the sex-at-birth question, all three surveys use different question wording for gender identity, as shown:

BRFSS: Do you consider yourself to be transgender?

• Yes, Transgender, male-to-female
• Yes, Transgender, female-to-male
• Yes, Transgender, gender non-conforming
• No

NIS: Are you male, female, or transgender?

• Male
• Female
• Transgender

PATH: (a) Some people describe themselves as transgender when they experience a different gender identity from their sex at birth. For example, a person born into a male body, but who feels female or lives as a woman, would be transgender. Do you consider yourself to be transgender?

• Yes » ask (b)
• No
• Not sure

[if R answers YES]
(b) Do you consider yourself to be male-to-female, female-to-male, or non-conforming?

• Yes, Transgender, male to female
• Yes, Transgender, female to male
• Yes, Transgender, gender nonconforming
• No
• Not sure

HCPS: What is your gender?

• Male
• Female

[If age >13]:

• Female to male transgender male/trans male/female to male
• Male to female transgender female/trans woman/male to female
• Genderqueer
• OTHER, specify

[If Genderqueer or other]
We have entered your gender as [Genderqueer OR other: FILL]. In this interview, questions will appear based on gender. For example, we only ask questions about mammograms to females of a specific age. Since this is a research study collecting medical-related data, could you tell us your biological sex at birth?

• Male
• Female

Under the one-question approach, respondents would be identifiable as transgender only if they selected that answer category explicitly, which could risk under-reporting compared to the two-question approach (Tate et al., 2012).

Because these focus groups are exploratory, we presented five different questions to respondents for feedback. This included a one-step question and four versions of two-step questions, all of which were based on GenIUSS recommendations with slight modifications. When discussing proxy, we gave respondents a modified two-step question with the precursor “To the best of your knowledge.”

1. What is your current gender?

• Male
• Female
• Transgender
• Do not identify as male, female or transgender

Two-step questions

1. Was your sex recorded as male or female at birth? [Proxy reporting: To the best of your knowledge, was [NAME]’s sex recorded as male or female at birth?]

• Male
• Female

2A. Do you describe yourself as male, female, or transgender? [Proxy reporting: To the best of your knowledge, does [NAME] describe themselves as male, female, or transgender?]

• Male
• Female
• Transgender

2B. Do you describe yourself as male, female, or transgender or something else?

• Male
• Female
• Transgender
• Do not identify as male, female, or transgender

2C. How do you describe yourself? (Check one)

• Male
• Female
• Transgender
• Genderqueer/gender non-conforming/non-binary

2D. How do you describe yourself? (Check one)

• Male
• Female
• Trans male/ Trans man
• Trans female/ Trans woman
• Genderqueer/gender non-conforming/non-binary
• Different identity (please state): ___________________

In addition to issues of question wording, the OMB working group expressed concerns about cultural and non-binary⁵ inclusivity, noting that, “Some individuals who vary in age, cultural and linguistic groups, etc., may not endorse terms such as ‘transgender’ when responding to federal gender identity questions because they do not identify with this term” (Federal Interagency Working Group, 2016c). To address this, we recruited on a wide range of demographic characteristics and aimed for adequate representation from various minority and non-minority groups, on characteristics such as race, ethnicity, and urban or rural residency.

Another measurement concern is how surveys should use pronouns such as “he” or “she” in subsequent questions. Transgender respondents vary in their preferred pronouns; for example, some respondents prefer the gender neutral “ze” (University of California, Berkeley, 2017). Surveys collecting gender identity need to decide whether to use the preferred pronoun based on gender identity, the pronoun aligning with sex at birth, or “they” as a singular, non-gendered pronoun (Federal Interagency Working Group, 2016c). After deciding how to use pronouns, surveys then need to implement this change throughout the entire survey, a task that is operationally complex and time-consuming.

Finally, the working group expressed concerns about terminology evolving over time. We acknowledge this, and below we discuss findings on the fluidity of some of these categorizations even in a static context (that is, even from day to day, self-identities may change).

Respondents’ feedback in the present study on the one-step question and four variations on the two-step question provides insight into some of these unresolved question terminology issues (Section 4.4).

2.3.1 Proxy Reporting

Many federal surveys use proxy response, in which one person generally responds for all eligible household members, primarily to reduce costs and nonresponse (Tamborini and Kim, 2013). It can be very time-consuming and difficult to collect survey responses when all members of the household are required to answer for themselves (Pierce et al., 1993; Park, 2015). However, proxy response involves other tradeoffs, as proxy answers may differ than those provided by other household members. Across survey topics, evidence on the quality of proxy response in surveys is mixed. Respondents may use less precise question answering strategies such as estimation when answering about other people in the household (Bickart, Blair, Menon, and Sudman, 1990). Data quality can depend on the question topic and the relationship between the respondent and others in the household. For example, studies have found that respondents most familiar with other household members, such as spouses, tend to be better proxies (Kojetin and Mullin, 1995; Tamborini and Kim 2013; Grieco and Armstrong, 2014; Pascale 2016). Small differences in agreement between proxy and self-response answers are more common than large differences (Mellow and Sider, 1983; Boehm, 1989; Moore, 1988; Tamborini and Kim 2013).

With regard to gender identity questions in particular, very little is known about whether respondents have the knowledge necessary to report for other household members, and whether they would be willing to report the information if they do have it. Surveys in New Zealand and the United Kingdom do not permit proxy reporting for sexual orientation due to concerns about accuracy and confidentiality (Joloza et al., 2010; Park, 2015). One of the only studies conducted on this issue employed an online nonprobability panel to test questions similar in wording to the questions we use in this study (Ortman et al., 2017). That study, while unable to get at accuracy of the responses, found overall low rates of nonresponse for the gender identity questions. Item nonresponse for these questions was lower than for income, which is also considered to be a sensitive question. However, nonresponse to the gender identity questions was also significantly higher for proxy reports compared to self-reports. Nonresponse rates also varied by relationship categories; rates of nonresponse were higher when respondents were reporting for children of respondents (age 16 or older) and roommates than for other household members, such as spouses and unmarried partners.

In the present study, we spent time discussing proxy response of gender identity with respondents, in terms of both accuracy and sensitivity (Section 4.3). Their feedback provides insight into whether proxy response is appropriate for gender identity.

2.3.2 Survey Context

As noted above, the OMB working group raised the issue of survey context—that is, asking gender identity questions in the context of an employment survey, versus the more common context of health surveys. The concern was that if respondents view gender identity questions as irrelevant to the subject matter of the survey, they may refuse to answer the question or breakoff from the survey entirely. Thus far, there is very little research available in this area. In one study, after gender identity questions were added to the NCVS, researchers conducted a debriefing survey with interviewers about issues they experienced when administering gender identity and other questions. A relatively small percentage of interviewers reported that at least one respondent asked why gender identity was relevant to crime (Truman et al., 2017). Nonresponse to these questions was low and only a few respondents broke off from the survey after being asked these questions.

In the present study, we introduced the CPS and described it as a survey about employment to assess whether respondents felt gender identity questions are appropriate in this setting (Section 4.2). Their feedback provides insight into whether gender identity questions are seen as irrelevant in surveys about topics other than health.

3. Study Methodology

We conducted four focus groups with 29 respondents in Washington, DC; Portland, OR; Nashville, TN; and Fargo, ND. The Washington focus group was conducted in September 2016, while the Portland, Nashville, and Fargo groups were held in March, April, and June 2017, respectively. These cities were selected to represent different geographic regions of the country, with the assumption that these regions would also vary on attitudes, political experiences, and other factors that would impact respondent experiences and opinions. The Washington, DC; Portland, OR; and Fargo, ND focus groups lasted about two hours, while the Nashville, TN focus group lasted about one hour due to the small number of respondents (see Section 3.1.2). Focus groups were moderated by staff from the Center for Survey Measurement (CSM) at the U.S. Census Bureau. All focus groups were conducted in person, and respondents received $75 each to compensate them for their time. Following standard Federal research procedures, OMB provided clearance for this study prior to the start of recruitment.

3.1 Data Collection

3.1.1 Recruitment

For the research carried out in Washington, DC, recruitment methods consisted of posted flyers, advertisements through Craigslist.com, a broadcast message sent to all U.S. Census Bureau employees who work in the Suitland, Maryland headquarters, and posts on Facebook pages for LGBT groups that featured a telephone number and email address.

For the other sites, the Census Bureau and Bureau of Labor Statistics (BLS) established a contract with Community Marketing, Inc. (CMI) to handle recruiting and onsite logistics. CMI maintains a nationwide research panel of LGBT individuals who were recruited for this study. CMI also recruited new respondents using targeted Facebook advertisements to transgender Facebook users, flyers, and Craigslist.com advertisements. Print recruiting materials featured a telephone number, while digital advertisements directed participants to a CMI intake survey first. Intake survey responses were used to evaluate whether the respondent was potentially appropriate for the study.

All prospective respondents were screened via telephone prior to being scheduled for focus groups. Screening included questions on respondents’ age, race, ethnicity, employment, geographic area (urban versus rural⁶), household composition, and gender identity. On gender identity, we used screener questions that differed from the questions being tested.⁷ We aimed to recruit respondents who were diverse on these characteristics to the maximum extent possible; however, given the relatively small size of the transgender population and thus difficulty in recruiting respondents, we were not always successful in doing so. The next section provides details on the characteristics of those successfully recruited.

3.1.2 Respondents

In total, four focus groups were conducted with 29 transgender respondents. Table 1 below shows the number of focus group respondents per site, while Table 2 describes respondent and household characteristics overall.

3.1.3 The Moderator's Guide

All respondents were provided with a consent form before beginning the focus group. Respondents were also told that information they provided would be confidential and video- and audio-recorded,¹² and were notified that there were observers present. After respondents signed the consent forms, focus group moderator(s) began the group by introducing the topic, setting ground rules for the discussion, and facilitating the introductions of respondents to each other and to the moderator(s). Respondents were told that these focus groups were one of the first steps being taken as a part of research to help the Census Bureau and BLS understand how to ask questions about gender identity on one of our national surveys. Focus group moderators followed a pre-scripted guide for the remainder of the discussion, but had the freedom to spontaneously add or eliminate probes and adjust focus group timing as needed to encourage discussion and elicit feedback from respondents. Changes were also made to the moderator’s guide as needed following each of the first three focus groups.

Following the introductions, we spent about 30 minutes discussing gender self-identification. Respondents were asked what comes to mind when the term “gender identity” is used, how they self-identify, how they came to use their chosen term(s), and how they would define the term “transgender.” While not an explicit open research question in the research on measurement of gender identity in Federal surveys, we opened the focus groups with this discussion to allow the focus group moderators to introduce the topic and because we expected responses to be informative for purposes of question wording.

Next, we slowly introduced the idea of questions about gender identity being asked in the CPS.¹³ We first introduced the topic of surveys generally,¹⁴ followed by Federal government surveys, and finally the CPS. In this section we asked respondents why the Federal government might be interested in collecting this information, how they thought the data would be used, what concerns the respondents had, if any, and how the Census Bureau and BLS could make collection of gender identity information as easy and comfortable as possible for respondents. This section was allotted approximately 25 minutes.

We then discussed wording of questions about gender identity, aided by several example questions, including a one-step question and variants of the two-step question (See Section 4.4). Respondents offered opinions and suggestions on these sample questions. While wording of gender identity survey questions is an important, relatively unstudied area, the primary goal of this project was to assess the feasibility of asking about gender identity on the CPS more generally. Thus, we only allotted 15 minutes for this section.

Finally, we concluded the groups with 25 minutes of discussion about measurement of gender identity via proxy. We asked respondents how others in their household would answer these questions about them, whether their answers would be accurate, if they would be comfortable answering, and whether this depended on who in the household answered, amongst other things. The group concluded with a few minutes for respondents to offer additional comments and feedback.

3.2 Analysis

Following data collection, each focus group was transcribed verbatim. Personally identifiable information (PII)—such as names, places of employment, etc.—were not included in the transcription, and respondents were referred to using respondent ID numbers. The introduction, off-topic comments, and moderator probes were summarized rather than transcribed.

We then created a summary document organized by the research questions and sections of the focus group moderator’s guide. Each section of this document corresponded to probes or groups of probes. We answered each of these research questions individually for each focus group by summarizing the discussion and including relevant quotes from respondents for these probe groups. When possible, we counted the number of respondents who voiced a particular sentiment during the group discussion. However, we caution against overreliance on these numbers, as the group dynamic of focus groups shapes respondents’ comments and perspectives (Krueger, 1998; Horwitz & Finamore, 2017). Respondents are influenced by each other, and respondents who are quiet may or may not agree with other opinions voiced. It is difficult to adequately capture how many people specifically agree with a sentiment voiced during the group.

This summary document formed the basis of the report, as it allowed researchers to look across focus groups and identify common themes as well as differences across sites. Notes from observers at the four focus groups were also used to supplement this analysis.

In order to protect the confidentiality of focus group respondents, some of whom prefer identifying pronouns, we refer to all focus group respondents in this report as “they.”

4. Focus Group Findings

The focus groups were used to gather feedback from transgender respondents to answer four research questions:

1. How do transgender respondents define “transgender” and “gender identity”, and relate these terms to their self-identity? (Section 4.1)
2. How do transgender respondents feel about the collection of gender identity information by the Federal government, both generally and in the context of the CPS? (Section 4.2)
3. How do transgender respondents feel about gender identity information being collected via proxy response? (Section 4.3)
4. What feedback do transgender respondents have on wording of questions about gender identity? (Section 4.4)

We discuss findings corresponding to each of these research questions in Sections 4.1, 4.2, 4.3, and 4.4, respectively, and respondents’ concluding remarks are discussed in Section 4.5. Findings are organized with separate subsections for cross-group similarities and differences.

4.1 Self-Identification and Definitions

While not an explicit open research question in the research on measurement of gender identity in Federal surveys, we introduced the focus groups with a discussion on self-identification and definitions. This was for two purposes. First, opening the focus groups with this discussion allowed the focus group moderators to introduce the topic. Second, collecting responses to these probes was expected to be informative for purposes of question wording. This first section of focus group findings discusses what comes to respondents’ minds when they hear the term “gender identity,” their process of self-identification, and how they would define the term “transgender.”

4.1.1 Gender Identity

We first asked respondents what comes to mind when they hear the term “gender identity.”

4.1.1.1 Similarities across Groups

Respondents in all of the focus groups generally agreed that one’s gender identity involves how someone feels and sees themselves.

“Who we are, or who I am. How I identify myself.”

“Internal perception of ‘I.’”

“How you see your identity in your head and how you feel.”

“Matching my soul with the way I feel.”

“I cannot quantify it, I cannot measure it, it cannot be observed physically by my anatomy. It is what I know my gender to be.”

4.1.1.2 Differences across Groups

In the Portland and Nashville focus groups, respondents said “whoa” and “yikes” in response to the question. Respondents in these groups felt like “gender identity” was a broad concept and difficult to unpack.

Additionally, respondents in the DC, Portland, and Nashville focus groups discussed that the term “gender identity” makes them think of “anyone who is not cisgender” or “cis.”¹⁵

“Everything that’s not the two things that are cis people, which is like a billion things.”

“We have gender identities but cis people get just a gender.”

“When I hear the term ‘gender identity,’ I think of trans people.”

In DC, three respondents made comments indicating that someone’s gender identity also involves a component of how you want other people to perceive you, but the majority of the group disagreed with this.

In Fargo, two respondents thought that societal expectations of gender roles are a component of gender identity; two respondents disagreed with this.

“I would say that it’s a psychosocial construct that is developed both by our internal sense of who we are as well as the social interactions and social expressions that we engage in—I think that it’s an interaction.”

“I think also there’s a sense of how you view your gender in relation to others, at least for me.”

“I’ve always considered gender identity to be completely separate from the societal construct. The societal construct I think is the gender role, the gender norm, whereas the identity is what’s internalized. What society expects us to do in our gender identity is the role. Whether you decide to adopt that role is not relevant to your identity.”

One respondent in Portland also mentioned gender fluidity, saying, “gender identity is fluid and does not remain static, although some people can place it in a static zone and keep it.”

Respondents in three of the focus groups (DC, Portland, and Fargo) also commented on the role of other people’s perceptions and expectations in their identity, as well as differences between sex¹⁶ and gender.¹⁷ There was not a clear consensus from respondents on how these concepts interact.

In Portland, respondents discussed that sex and gender are often conflated by society:

4.1.2 Self-Identification

Next, we asked respondents to describe their process of self-identification. This included probes on how they identify themselves, how easy or difficult it was to pick a term or terms, and whether they used different terms at other points in their lives.

4.1.2.1 Similarities across Groups

There was a wide variety of identities in our focus groups. The vast majority of respondents shared a term or terms they used to self-identify, including:

• Transgender
• Male and/or man
• Transmasculine
• Masculine
• Woman
• Trans woman
• Woman and trans woman
• Agender fem
• Fem androgynous
• Aporagender
• Queer
• Gender-fluid
• Genderqueer
• Gender-fluid/genderqueer
• Non-binary
• Gender-fluid, genderqueer, non-binary trans woman
• Non-binary genderqueer trans guy
• Bi-bi kid (bisexual, bigendered)

Two respondents who provided a term or terms to explain how they self-identify also made comments that they were still unsure or “working [things] out in their head.”

Some respondents used metaphors or other types of descriptions to communicate their gender identity, either in place of or in addition to term(s). This was most common in the Portland and Fargo focus groups.

“I often say that if you picture gender as a color wheel, and let’s stick with the western notion of pink being girl and blue [being] boy, I identify as mauve, because it’s an invented color that’s not really blue or pink, but it’s more on the pink side. ...I just say that I’m like basically a gender nesting doll because no matter how many layers you go there’s always something else. So like trans, then non-binary, then I use agender as an umbrella term, which a lot of people don’t, so eventually you just get to me.”

“If I have to break that down for people, usually non-binary, genderqueer, date whoever.”

“I like to tell them that people can be anywhere on or outside of the [gender] spectrum, including more than one place at once or none of the places, and that I just move. I identify as different things at different times.”

“We kind of have the main labels: the non-binary, agender, trans, bigender, and so on and so forth. We have a good amount of labels that fit most people mostly. You think of them like colors, like red, blue, and green are primary. There are an infinite amount of things between these colors, hue, and saturation. The same is true of how we understand ourselves and our gender, the complexity of it...I’m fuchsia, but it’s kind of like, I’m just going to say I’m reddish.”

Respondents were split in whether it was easy or difficult for them to select a term or terms. In Portland and Fargo, more respondents said it was difficult; in DC, more respondents said it was easy; in Nashville, both respondents said it was easy. A common sentiment across the groups was that everyone has their own, unique, self-identification journey that drives the selection of the terms they use to describe themselves.

Many respondents (4 in DC, 5 in Portland, and 5 in Fargo) said that they used another term at some point in their lives. Some respondents who previously identified differently than they do now did so because they had not yet been introduced to the language that best fits them. For example, three respondents in the DC focus group first identified as part of the lesbian, gay, and bisexual (LGB) community, but then realized that “transgender” was more appropriate once they learned of it.

Others changed descriptions for different reasons. One respondent referred to it as a process of “fine tuning their own self-description.” In three of the focus groups (DC, Portland, and Fargo), respondents discussed that gender identity can be fluid and change over time.

“When I first [was] transitioning,¹⁸ I identified as genderqueer, they/he and she, and I went by ‘he and she’ for quite a bit of time until I had top surgery...and then I slowly transitioned to just he/him as I started to look more masculine all the time. So that’s what I feel comfortable with [now], masculine pronouns, but it’s a process for everyone. Pronouns may change over time; how you identify may change over time.”

“I think it’s just sometimes where you’re at along that journey in that transition in how you identify. I identify as just generally non-binary, but it’s one of those things. Would I identify as more of a trans woman later on, 10 years down the round? I don’t know; I have no clue.”

“I think it can change...I started transitioning a long time ago and I think the way I felt about myself at that time has shifted now.”

“I came out as trans almost 20 years ago and my gender identity has certainly evolved over time. I used to be much more trans identified, genderqueer, and [recently] I’ve definitely developed a more binary gender identity.”

“Especially when you’re trying to capture it on a form, it’s really difficult because it’s gonna change over time. In some cases from one hour to the next depending on the context, depending on how that person is feeling.”

While the majority of respondents in these focus groups were younger, some older respondents mentioned difficulty self-identifying in a time when the word “transgender” was not frequently used or well known. For example, one respondent cited lack of Internet and a sheltered, religious upbringing as reasons why self-identification was difficult for them.

“When ‘transgender’ came out as a term it was commonly used—I latched onto that right away because it immediately set me apart from the LGB.”

4.1.2.2 Differences across Groups

In DC, most (6 of 8) respondents’ identities were binary,¹⁹ with the majority of the respondents identifying as male or a man. In Nashville, one respondent had a binary identity and the other did not; in Portland, most (6 of 8) respondents had non-binary identities; in Fargo, it was fairly evenly split (5 of 10 were binary, three were non-binary, one identified as both binary and non-binary, and one did not explicitly state their gender identity).

In Portland and Fargo, there seemed to be more emphasis on outward gender expression²⁰ and commentary on whether respondents met or violated expectations for their gender.

“Inside I’m totally genderqueer, and outside I’m just pretty much a guy.”

“I don’t do a lot of the typical things [that a man would do]—I’ll wear whatever the hell I feel like wearing.”

Four respondents in DC, one respondent in Portland, and one respondent in Fargo commented that their description of their gender and gender identity depends on the setting. Some respondents in DC said they avoid disclosing they are transgender in order to protect their safety, avoid inappropriate questions, or prevent harm to their career prospects. One respondent in the DC focus group also said that they let misidentification slide because they have trouble “passing”²¹ as male. In Fargo, a respondent said they offer a simpler description because “a lot of people raise their eyebrow and say, ‘what?’” when they use their preferred terms.

4.1.3 “Transgender”

The last probe in this section asked respondents how they would define the term “transgender.”

4.1.3.1 Similarities across Groups

Respondents across all groups agreed that to be transgender means being “the opposite of cisgender.” Other definitions offered included “not identifying as the gender of your birth” and “if you question your gender, you are probably some flavor of transgender.”

Respondents generally agreed that “transgender” could be used as an umbrella term to describe members of a diverse community.

“An umbrella term that encompasses any identity that challenges the social norms as we know them.”

“I think that people can for themselves individually decide that—trans isn’t a term they use, but I have always sort of seen it as an umbrella that includes people who undergo medical transition and people who don’t, people who identify as non-binary or have fluid identities, things like that. In that sense for me it’s...a way of grouping a community together.”

However, some respondents expressed that they did not feel the term applied to them personally or to others that they know. One respondent who is non-binary said they consider themselves to be part of the transgender community, but that they knew others who are non-binary who would not.

“I personally feel like I don’t really identify as being trans even though I’m non-binary, because I feel like there’s a really long intensive experience that is taking hormones, and transitioning, and physically being that in the world. It is different than because there’s social repercussions to that. Whereas I feel like just internally, being non-binary is my own thing, and sure I express that in the world and it has its own consequences...but it’s like they’re different experiences to me.”

“You know I used the word transgender to describe myself for many, many, many years. Two or three years ago I stopped using the word transgender to describe what my situation is because the prefix ‘trans’ means to go from point A to point B...And I said, ‘wait a minute, I’m not going from point A to point B.’”

4.1.3.2 Differences across Groups

In the DC focus group, there was also some discussion of transsexual²² as a term. Four respondents explicitly commented that they did not like this term and that they thought it had a negative connotation. One respondent stated that “transgender” seems to be a word that is more encompassing of other changes beyond physical ones.

4.2 Government Collection of Gender Identity Information

Of principal interest to this project was how transgender respondents feel about the collection of gender identity information by the Federal government, both generally and in the context of the CPS, a survey about employment.

We first introduced the topic of surveys generally, followed by Federal government surveys, and finally the CPS. In this section, we asked respondents why the Federal government might be interested in collecting this information, how they thought the data would be used, what concerns the respondents had, if any, and how the Census Bureau and BLS could make collection of gender identity information as easy and comfortable as possible for respondents.

4.2.1 General and Government Survey Experience

Respondents were first asked about surveys they have taken, with optional probes about the survey sponsor, reason for responding, and mode of response. This line of probing was primarily included to orient respondents to what we meant by surveys.

Next, we gave a high-level overview of government surveys, and asked if respondents had had any experience with them in particular.

4.2.1.1 Similarities across Groups

Nearly all respondents (6 of 8 respondents in Portland, 7 of 10 respondents in Fargo, and both respondents in Nashville) reported having some kind of survey experience.²³ Survey topics that respondents mentioned included:

• Politics
• Marketing
• Health care experience
• School (course evaluations, surveys for classmates’ projects or departmental research)

Survey sponsors included news organizations, theatre organizations, and LGBT support organizations.

Most of respondents’ experience seemed to be with surveys that are conducted online. One respondent in Nashville thought they had done a telephone survey before for a business and a paper survey after a theater performance.

Few respondents had prior experience with government surveys. Of those who did have this experience, the vast majority mentioned the decennial census. Three respondents in Portland, one respondent in Nashville, and one respondent in Portland reported completing the census. One respondent worked as an enumerator for the census. Other government survey experience included a survey for the Veterans Administration (VA) and a survey for the Social Security Administration (SSA) after a name change.

4.2.1.2 Differences across Groups

Respondents’ answers to these first two probes were consistent across groups.

4.2.2 Whether the Government Should Collect Gender Identity Information

After introducing the topic of government surveys to respondents, we asked them if they thought government surveys should include gender identity questions.

4.2.2.1 Similarities across Groups

Most respondents said that they thought government surveys should include gender identity questions. In DC,²⁴ four respondents explicitly voiced their support immediately; in Portland, two respondents voiced their support; in Nashville, one of the two respondents voiced their support; in Fargo, all respondents voiced their support. Comments made during this section of the guide received wide agreement from each of the groups. No respondents said they thought gender identity information should not be collected during this line of probing. Respondents supported the collection of gender identity information for a variety of reasons, including:

• Learning what the overall U.S. population looks like in terms of gender identity
• Getting an accurate count of the transgender population
• Using an accurate count to advocate and allocate funding
• Helping medical facilities and insurance companies adequately staff and train employees
• Preventing discrimination and moving towards equality
• Making the transgender population more visible
• Allowing transgender respondents to express themselves accurately on gender questions

Respondents commented:

“Doesn’t the government want to know who their citizens are? I would think you would.”

“I really think it’s important to describe our population, to be able to describe disparities that exist, to be able to say more definitive things about some of the things we’ve heard—that transgender people experience more suicide risks, or smoking, or these types of things, and there aren’t going to be any government funded services that help trans people without data.”

“The more that government understands what the trans word is like, what our specific experiences are...the more likely we are to be treated better, treated like real people.”

“Since it’s how government dollars are used to get services, it needs to be known that we are out there so that we can be included in those dollars, getting grant money, whatever, for our causes.”

“Not so much [adding] questions, but you know, being able to express your identity honestly.”

“Accurate population statistics would be fantastic.”

“We need to arrive at a position where LGBT rights are not called LGBT rights, they’re called human rights.”

One respondent in each of the Portland and Nashville groups was initially hesitant, saying that there should be a good reason for collecting gender identity or feeling skeptical that it would be implemented.

“If it’s relevant...then it should definitely be on there.”

“It would really take a Federal adoption of accepting that there’s more than just the gender binary in order for them to even consider really putting it on every Federal survey.”

“I think that, for example, [other] surveys I’ve taken, they always ask about gender, but I don’t always see the point of that. It just seems like a thing people put on surveys, so if you have a good reason, sure.”

Others cited concerns about how data collection would work; these concerns are discussed in Section 4.2.4. There were also comments made about an accurate data collection being difficult and question wording being important; these comments are discussed in Sections 4.2.4 and 4.4, respectively.

4.2.2.2 Differences across Groups

The reaction to government collection of gender identity was more mixed in this line of probing in Portland than in other groups. Five participants said that having information on the transgender population would be valuable, but also talked more in depth about data use, confidentiality, and question wording concerns than respondents in other groups did.

One respondent in each of the Nashville and Fargo focus groups who supported the collection of this information also voiced that they wished gender identity was not something that needed to be collected.

“I would really wish that the government didn’t have to collect any gender statistics because every person on the gender spectrum would just be treated equally and the government wouldn’t have to care who is what gender.”

4.2.3 Reasons for Collecting and Use of Gender Identity Information

Respondents were then asked why the Federal government might want to collect gender identity information and what they would do with it.

4.2.3.1 Similarities across Groups

When asked why the government would want to collect this information and what they would do with it, respondents generally echoed comments made in the previous section about needing population statistics, wanting to increase visibility/provide evidence of existence, improving health care, and achieving positive policy change. The majority of comments made during this section of the guide were positive. One respondent said that they were not sure how the data would be used.

“Selfishly, I want this to get normalized in our society, and the more that it’s out there, and the more it’s visible —‘oh, it’s even on this form that the government sent me!’ — I think that will go a long way toward normalizing it.”

“To get the actual number because there’s no idea how many transgender people there are in the United States. It’s all a guess.”

“Because they’ve been missing out on it until this point in a lot of ways.”

“It might [make] our issues become important to senators or congressmen to actually start heading dollars our way, [to tell them] we are a voting block that they should pay attention to and not just dismiss. It might bring us some equal rights.”

However, there were a few comments in each group from respondents who were skeptical that the data would be used at all, that it would be used for the intended purposes, or were generally distrustful of the government. The concerns voiced here received agreement from many other members of the group. In two of the groups, respondents discussed that it is important to know who in the government would be using the information. In particular, respondents were worried that politicians who support anti-transgender legislation would have access to the data.

“[They’re] going to round us up [laughter].”

“Nefarious reasons.”

“It depends on who has decided to ask the question. On the one hand, if I were interested in silencing a percentage of the population it might be helpful to know how many of those people exist so I know how to silence them. On the other hand, if I’m interested in improving the civil rights of a minority class, I would ask those questions so I have a better understanding and have the data to use as back-up. So I’m honestly, even coming here, I was a little apprehensive. I don’t trust the government.”

“I’m happy to come here, but I was like, hopefully this is a good set of people.”

“I made a joke at the beginning: ‘to be sure, you’re not rounding us up here to gas us or something?’ That is a joke, but—”

“Definitely people I know would feel scared to give out that information to the government.”

As a counter-point, one respondent in Fargo said that they did not think current conditions could get worse for the transgender community as a result of the government collecting gender identity information.

“I don’t think information about us could hurt us, considering the points made against us by individuals who don’t want to be trans accepting: that we don’t need to [be accepted], that it’s a mental illness, it’s an insignificant population size, or problems aren’t happening because [cisgender people] accept and love everyone. Those are kinds of the main ways to erase trans individuals, and I think erasure is the biggest tool being used right now.”

Many of the respondents who stated concerns about the government collecting information about gender identity also explicitly stated that they support it.

4.2.3.2 Differences across Groups

Respondents in all four focus groups were very similar in their responses to these probes. However, respondents in the Fargo focus group spent more time discussing the role that gender identity questions could have in increasing the visibility of the transgender community in society generally. Respondents thought questions would remind survey respondents that there are people who identify as something other than male or female. They also thought that the presence of additional options might encourage others to come out.

4.2.4 Concerns about the Government Collecting Gender Identity Information

Next, we asked respondents what concerns they have about the collection of gender identity information. In most groups, respondents began voicing these concerns in the prior two sections (Sections 4.2.2 and 4.2.3).

4.2.4.1 Similarities across Groups

Very few respondents (3 in DC, 1 in Portland) volunteered that they were not concerned about government collection of gender identity information. The respondents in DC said that they were not concerned because they did not think the concerns others voiced would come to fruition, while the respondent in Portland said, “At this point in my life, I don’t care.”

Concerns voiced by respondents in all four groups included leak of information outside of the government, general malicious use, and confidentiality/anonymity. The latter two were the most frequent concerns communicated.

“If someone malicious, fearful, or hateful towards transgender and genderqueer people gets ahold of that information, it would be very frightening.”

“Just in general, people taking the statistics and using them for something that wouldn’t be helpful to the community.”

“I’ve historically filled out surveys that haven’t been anonymous, and just putting that I identified as a woman and [was] living alone gave me a sense of anxiety...I think everybody has someone to be afraid of.”

4.2.4.2 Differences across Groups

Respondents offered more unique concerns and discussed them more in-depth in the Portland, Nashville, and Fargo groups. Malicious use within the government to discriminate or implement and legitimize anti-transgender legislation was a concern voiced in all three of these later groups. Unlike the DC focus group, these groups were conducted after the 2016 Presidential election. Many respondents in these later groups discussed the election and its aftermath when communicating their concerns. There were also a few scattered comments across the groups about other transgender issues in the news, such as the ability to change the sex recorded on birth certificates²⁵ and the controversial passage of the Public Facilities Privacy and Security Act in North Carolina.²⁶ It is also possible that respondents in the DC focus group differed in their attitudes because of the area’s proximity to, and thus familiarity with, the Federal government. While we did not have any Federal employees in the focus group, a few respondents mentioned having friends and family who worked for the government, and one respondent worked for a Federal government contractor.

Appropriate question wording was also spontaneously raised as a concern of respondents in the Nashville and Portland groups, which is discussed in Section 4.4.

Those in the Portland group offered additional concerns, including inappropriate rationale for asking about gender identity and interviewer misconduct. They described specific data use scenarios that were concerning to them:

“I worry about that information being shared to, say, my insurance company somehow, and then I’m denied coverage.”

“I think my concerns would be how it would be used politically...would they use that to identify you?...now they’re going to identify me in some way, and tell me which bathroom I have to use, or identify me to my employer, or nullify my marriage...however they want to target trans people, that it could come back to bite you.”

“The government can have my information but say, the Kim Davis example.²⁷ She now knows I’m transgender, she has a problem with transgender people, she’s not going to give me a marriage license. Or Joe Schmoe found out because he took the information for the government, and he told his friend Randy, who then goes and beats me up because he found out I was trans.”

4.2.5 Concerns about Mode of Administration

Next, we asked if the mode of survey administration made a difference to respondents. We were particularly interested in differing perceptions between telephone and in-person interviews, as the CPS is conducted in these two interviewer-administered modes, with most initial interviews conducted in-person. Paper and online self-administration were not directly probed on because they are inapplicable to the CPS, but some respondents commented that these self-administered modes would be preferred without prompting.

4.2.5.1 Similarities across Groups

While mode seemed to matter, there was no clear consensus on mode preferences. Phone seemed to be preferable to in-person for the few respondents who offered an opinion.

“I might feel a little weird about face to face, because if you’re on the phone and it’s a survey, you can say whatever you want to and you don’t see the other person’s reaction to it, but in person you do.”

Two respondents (one in Nashville and one in Fargo) said that they did not care about the mode of administration.

4.2.5.2 Differences across Groups

Respondents in the DC focus group did not offer very many opinions here, while respondents in Portland generally felt uncomfortable with disclosing their gender identity to an interviewer, either in person or on the phone. Three respondents in this group were unsure what they would do, and thought they might lie to an interviewer.

“How would you have that conversation? Because then it’s like a dialogue...I just have a feeling that would be more uncomfortable. I guess I feel uncomfortable if I was taking a form and I didn’t feel comfortable with the box I was checking, but I feel like that would be fleeting ...But like to have that whole stressful conversation and probing you about identity to be taken then and done things with ...”

“People who got my phone number somehow that I’ve never met asking me questions about my gender. Am I going to lie to them? Probably. Unless it’s someone from a queer organization that I know of.... I would answer very differently than I would if someone was like I am calling from the Department of Agriculture...I’m not really feeling like being open.”

They were also concerned for respondents in vulnerable situations.

“I had a thought about people who were in abusive situations and how it’s probably not safe for them to answer questions on the phone or in person.”

Four respondents in Fargo also said that they would rather not have someone come to their house in general. Later in the group, one respondent in Portland offered a positive comment on having an interviewer ask gender identity questions.

“I think talking with someone, there’s a little back and forth that you can do, a little explaining if you need to. With a form there’s...no way to say, ‘kind of this, but not really.’”

Also in Fargo, respondents were very enthusiastic about the idea of a “public forum,” or a safe space where people could go to share their information. While impractical for the CPS, it underscored the importance of making transgender respondents feel comfortable when responding to gender identity questions.

4.2.6 Gender Identity on the Current Population Survey

At the end of this section, we introduced the CPS by name, and briefly described at a broad level how gender identity information on this survey would be used if it were to be collected. We told respondents that the CPS produces statistics on unemployment for specific populations, such as different states and different races, and that similar estimates could also be produced by gender identity. We asked respondents how they felt about this.

4.2.6.1 Similarities across Groups

The idea of collecting gender identity information for the purposes of the CPS received wide support across the focus groups. Respondents talked about it being useful to see how unemployment differs between people who are transgender and not transgender, with agreement from the larger group.

“I think that’s very important so it can illuminate it, so people can say, ‘wow, look at all these people who can work but they don’t have jobs, why?’”

“I think something coming from the government could really capture a lot of this data because you have a much wider reach.”

“I think that sounds cool because I think you would find that the unemployment rate is higher the further down on the gender totem pole you go as far as who is accepted by society; I think that would be telling.”

Some respondents also made comments about how these statistics could then be used by the government and other groups.

“If the Federal government is committed to protecting the workplace rights of trans people, then they absolutely have to know how many trans people are in the workforce.”

“These data are available to people, like nonprofits or researchers and even community organizations that are trying to affect the employability of trans folks. [They] need to be able to make their case and be able to say, ‘oh, 60% of trans folks are unemployed or underemployed’ or whatever it ends up being. That is a way for them to...identify the problem [and] address it.”

Three respondents in each of the DC, Portland, and Fargo focus groups also commented on transgender people being fired or denied a promotion because of being transgender. One of the three respondents in Portland who commented on this reported being fired from a previous job for speaking out about LGBT rights.

4.2.6.2 Differences across Groups

In Nashville, the two respondents talked about how it would be interesting to see how unemployment varies for transgender people in different regions of the country.

Respondents in Portland and Fargo were slightly more cautious in their support, echoing concerns about confidentiality/anonymity mentioned in Section 4.2.4.

“As paranoid as that idea makes me, I think it’s important to collect that data because I’m currently unemployed and I’ve been searching for a job for 4 months. I know that my public identity as a transgender individual has been hampering that job search and I know that this is not unique. So I think that having that data would finally tell the Labor Bureau what we’ve known for decades “ if you are trans and if you are outed, good luck finding a job.”

“I think it’s a great idea to include it for accurate statistics. I feel like there’s a lot of focus on fear, which is a legitimate concern because we’ve all experienced negative reactions and slurs, violence, and we are in one of the most conservative states in the U.S. We also have to focus on, [the Census Bureau and Bureau of Labor Statistics] did get people here who are trans who have an opinion and want our voices to be heard. I think it’s important to have input on this subject, so while it’s scary for people as trans people to talk about this, I feel like it’s important to get our input and feedback.”

4.2.7 Ways to Make Collection of Gender Identity Information Easy and Comfortable

As the last probe in this section, we asked respondents how we could make the collection of gender identity information as easy and comfortable as possible for respondents.

4.2.7.1 Similarities across Groups

Respondents in the DC and Portland focus groups were very focused on question wording in their suggestions. These are discussed in Section 4.4.

4.2.7.2 Differences across Groups

In Nashville, both respondents discussed that they thought that some people in the area would be offended by being asked gender identity questions and urged caution.

One respondent in Portland emphasized the importance of not making assumptions about someone’s gender from their voice or appearance. This is something that was mentioned a few other times at various points throughout the groups.

“One of the surveys that is done here in Oregon doesn’t allow interviewers to ask people what their gender is. They’re supposed to assume it based on the person’s voice...if [I were in] a situation where that happened and they were to [then] ask my gender, there’s no way in hell I would tell them. But...if there were multiple questions about lots of different things that were straightforward and matter of fact and the gender question was asked in that same fashion, then I would be comfortable answering it.”

Several respondents in Fargo suggested partnering with LGBT pride organizations for a “community-based participation research sort of approach.”

4.3 Proxy Response for Gender Identity

After discussing government collection of gender identity information, we asked a series of probes designed to capture how transgender respondents feel about gender identity information being collected via proxy response. Respondents’ thoughts on proxy response are critical since the CPS collects information by proxy.

4.3.1 General Concerns and Hesitations

The first probe in this section captured respondents’ initial reactions to the idea of proxy response. Respondents were told that one person responds for all people in the household. In later focus groups, we gave the additional example of asking one person for everyone’s names, everyone’s races, etc. We showed them a modified version of the two-step question for reference:

1. To the best of your knowledge, was [NAME]’s sex recorded as male or female at birth?

• Male
• Female
• Don’t know
• Refused

2. To the best of your knowledge, does [NAME] describe themselves as male, female, or transgender?

• Male
• Female
• Transgender
• Don’t know
• Refused

Because this section started out very open-ended, the order in which various comments were made and thus when probes were asked differed by group. They have been organized in the order of the moderator’s guide for the purposes of this report.

Additionally, though the questions were structured to elicit respondents’ thoughts on how other members of their household would respond, many respondents also commented on how they would respond if they were asked to respond by proxy on the behalf of other people.

4.3.1.1 Similarities across Groups

The immediate reaction from respondents in the DC, Portland, and Fargo focus groups was negative. It was not always clear to respondents why proxy response would be used or how the information would be used.

“The only way I envision [proxy response is with] a dependent...it would be super weird for it to be my partner...or like an adult who is capable of answering for themselves.”

Most of the initial concerns voiced were regarding possible inaccuracy of response and the inappropriateness of someone answering gender identity questions on someone else’s behalf. These concerns are discussed in Sections 4.3.3 and Sections 4.3.4, respectively.

While not all of these concerns applied to respondents personally, they reported that they knew people for whom proxy response would be problematic.

4.3.1.2 Differences across Groups

Despite the generally negative response, some respondents did not think proxy response would be problematic for them personally. The Nashville focus group in particular was very different from the other focus groups throughout this section on proxy. Both of the respondents in this focus group lived with one other person: one with a spouse and one with a roommate. They had very few concerns about proxy response.

In Fargo, three respondents said proxy response would work in their households.

“I don’t think my wife would have any issues and she would answer the same as I would, but not everyone has that same support that I do.”

“I live with my partner, and I think you’d really need to discuss that in your relationship if that would be ok. I’m openly trans, and I’m ok with her letting certain other people know without us first discussing it, like mutual friends and everything, that’s fine with me, but it’s really something you need to discuss with your partner and have a lot of trust in them. And...if someone came to the door and said, ‘does your partner identify as a man, woman, or transgender,’ I would be okay with her saying transgender, [female to male], that would be perfectly fine with me.”

“I live with my fiancé...but my experience is obviously not the norm.”

4.3.2 Whether Respondents Would Answer

Respondents were also asked whether they thought that members of their household would answer this question at all (i.e., whether members of their household would refuse to answer).

4.3.2.1 Similarities across Groups

A handful of respondents in each of the DC, Portland, and Fargo groups said they did not think members of their household would report gender identity on their behalf.

In Portland, three respondents explicitly said that people in their household would not answer for them. Later on, one of the three respondents in Portland said they thought their wife would actually answer if she were “under the gun” because “she knows me—I want to be counted.” One other respondent said they would not answer the first part of the two-step question for the people in their household. Two of these respondents asked why proxy response had to be used. One respondent in Fargo said their mother would not answer for them due to safety concerns.

“I think...especially people who are sensitive to gender issues would feel uncomfortable filling it out. Everyone at this table has told me tonight the answers to these questions. I would not feel comfortable filling this out for anybody at this table because you just, you can’t answer these questions for other people. You just can’t. Even if they’ve told you the answer, you just can’t.”

“I just have a problem outing people. I just don’t like to do that. I probably wouldn’t answer anything if I didn’t have to.”

“I also have an issue with people outing me. One hundred percent I would not want anybody to answer this for me at all.”

“I definitely wouldn’t put my realness in someone else’s hands to explain me.”

“I think my partner would be like, ‘you can [expletive] right off.’”

“I think my partner would ask me; there’s only 2 of us. It wouldn’t take long for her to ask everybody in the household.”

Four respondents in DC made comments indicating that they personally would not want someone else to report their gender identity or report the gender identity of someone else.

4.3.2.2 Differences across Groups

As mentioned previously, respondents in the Nashville focus group were much less concerned about proxy response. Both respondents thought their household members would answer for them and would not have any concerns about doing so.

4.3.3 Accuracy of Proxy Response

Respondents were asked if others in their household would be able to answer on their behalf. This line of probing was designed to elicit feedback on perceived accuracy of proxy response.

4.3.3.1 Similarities across Groups

Respondents had many comments about the perceived accuracy of proxy response. One respondent in DC, three respondents in Portland, both respondents in Nashville, and four respondents in Fargo thought that other people could answer the same way they would; all other respondents thought other people could not answer the same way. Respondents were primarily concerned about two different types of potential inaccuracy in response:

1. Respondents who are unaware of household members’ correct gender identity

“They’re going to mark you as whatever the see you as, and you’re not getting the authentic [answer] because you haven’t told them what your situation is.”

Even before proxy response was specifically discussed in the focus group, one respondent in Fargo commented:

“The only thing I could think of that could be problematic with that is if you had a trans family member that isn’t out to someone, so that’s not accurately reflected, or if the person is trans themselves but doesn’t feel comfortable revealing that identity to a stranger who’s coming to their door or calling them. So I can see how the statistics could still be skewed so it’s not as well represented as it should be and accurate.”

2. Respondents who may refuse to accept and acknowledge household members’ gender identity

Family and particularly parent-child relationships were cited as especially difficult, as parents may be “in denial” or unaware that their child is transgender. Three respondents in DC and two respondents in each of the Portland and Fargo groups discussed this.

“It’s very problematic because if you were to ask my dad, his answer would be totally different from mine, and anybody who isn’t out to their head of household, they’re not going to know any of that information or might refuse to answer because they don’t want you to know.”

“Currently I live alone, but I spent 1.5 years living with my parents, and despite all attempts in many formats, in many ways, at many times to get my parents to fully understand what’s going on with me, who I am, they do not necessarily have that information.”

“I’m just thinking about my own family that I do not talk to because they don’t accept my gender, and I don’t want those people to tell the government how I identify because clearly they’re wrong and that’s harmful.”

Another difficult situation discussed was that of roommates.

“Out of all of the roommates I’ve ever had, I’ve come out to like 10 percent of them maybe.”

“I live with 4 roommates, and 3 out of those 4 people, I don’t know if they know or not about my identity, and I don’t know if they think they know, or if they don’t think they know. I don’t know what they would say. I have literally no idea.”

Most of the respondents who felt more comfortable with members of their household answering on their behalf lived with a partner or other members of the transgender community.

“Queer households often have those conversations.”

“If I was living with my parents, absolutely not, but I live with another trans woman as a roommate so I feel like we would both be able to.”

One positive comment made in the Portland group was that the “to the best of your knowledge” and “best represents” language used in the example question is good because it allows respondents to select “what’s closest.”

4.3.3.2 Differences across Groups

As mentioned previously, respondents in the Nashville focus group were much less concerned about proxy response. Both respondents thought the other members of their households would be able to answer easily, though they commented there were a few different ways the household members might answer. Both of these respondents said that it would be more difficult for their parents to answer accurately if they were living with them.

A respondent in Portland said that accuracy is difficult due to the diversity of terminology used in the transgender community, and another respondent thought that the government should look up the answer to the first part of the two-step question for more accurate data.

4.3.4 Sensitivity of Proxy Response

Respondents were asked multiple probes gauging response sensitivity in order to gather feedback on whether respondents would feel comfortable with proxy response of gender identity. They were asked if members of their household would be comfortable answering on their behalf, if household members would have any concerns or hesitations about answering on their behalf, and if they thought household members would find the question sensitive to answer for other people.

4.3.4.1 Similarities across Groups

Along with response accuracy, response sensitivity was an area in which respondents had many comments. Four respondents in DC made comments indicating that people feeling comfortable answering on the behalf of someone who is transgender would be worrisome, and indicative that the respondent is answering incorrectly.

“I would be worried if they were comfortable to answer for me. I wouldn’t really know if I could trust them if they were comfortable to answer this for me.”

In Portland, five respondents voiced concerns about response sensitivity, and in Fargo, seven respondents indicated they were uncomfortable with proxy response. One respondent in Fargo said their parents do not like being “confronted” with the fact that the respondent is transgender.

Most of the comments made regarding response sensitivity were on:

1. General confidentiality concerns and inappropriateness of respondent answering gender identity questions on someone else’s behalf

“I think there’s a strong sense of you don’t disclose other people’s sex or gender; you don’t speak for them, whether it’s a case of accidentally outing them or giving more information than they’re comfortable with.”

“I know darn well if someone came to me with something like this, if somebody asked me about my wife, I would say ‘none of your damn business.’ Period.”

“And even if say I had [other focus group respondent] as a roommate, I do not one hundred percent know everything about how [other respondent] identifies. I should not have control over [other respondent]’s information and how [other respondent] is represented. Also, if [other respondent] is my roommate, and you don’t want that information presented I have no right to give that information for you.”

“That’s not my business to answer...that’s not for me to say.”

2. Situations involving transgender individuals who do not want to be outed

“I am totally uncomfortable outing anyone; even with their consent, it still feels kinda weird outside of queer circles to out people.”

“What happens if somebody gets outed as a result of this form? ‘To the best of your knowledge?’ That means you start thinking.”

“Let’s say I have a roommate who is cisgender female, they don’t know I’m trans, my gender has never been an issue, they get the survey, they ask me, suddenly I look uncomfortable, suddenly they press the issue further, and then that outs me in some way or another.”

“This potentially forces people to out themselves to their family, to their friends, to their roommates. I don’t want to speak for everyone, but for a number of people, stealth²⁸ is a way of survival. This questionnaire, trying to get this data by proxy eliminates stealth.”

3. Risks to transgender household members’ safety

“[My wife]’s scared of putting out too much because I might get hurt.”

Some of the comments made here were similar to those voiced in Sections 4.2.3 and 4.2.4.

Gender identity being tied to an individual’s name or other identifying information was also of concern to respondents in all of the focus groups.

“I would want to know how identifiable this is. Is this just how you’re referring to someone or [do you have] their actual name? I am much more comfortable fitting myself into the boxes that forms have than I am doing that for others....I wouldn’t feel comfortable making that choice for someone else.”

“I think it’s really the name part that really throws me off more than anything...I think that would make a difference for me.”

“I might be willing to fill out a survey like that if my name wasn’t attached...but if [there were] and there were any way to identify people, given the current political climate, I wouldn’t fill out anything.”

“[The government] already had a paper trail; I changed my passport gender and stuff like that. If they really wanted to find me and get a list, I would be on a list. But the less my name [is] attached to anything that may [result in] somebody scary knocking on my door, the better at this point, because who knows where we’re headed?”

“I love the idea of gathering this information but not necessarily keeping a record of this person, that person, other person.”

“Is that a survey that collects name, date of birth, that kind of thing?”

One respondent in DC and one respondent in Portland suggested getting a count of the number of transgender individuals in the household instead of collecting it at the person-level. In Portland, another respondent replied, “that wouldn’t let you get at were the trans people the ones who were employed?”

4.3.4.2 Differences across Groups

As mentioned previously, respondents in the Nashville focus group were much less concerned about proxy response. Both respondents thought the other members of their households would be comfortable answering in this context, though they may have concerns answering questions about gender identity in other, non-government survey contexts.

In Fargo, two respondents said they thought that reporting someone else’s gender identity would be disclosure of medical information, and thus may be a violation of the Health Insurance Portability and Accountability Act (HIPAA).²⁹ The concern about data getting into the wrong hands was reiterated in this group.

One respondent in Portland said they would feel comfortable answering the second part of the two-step question if different question wording were used, and another respondent said their comfort level depended on how the data would be used.

One respondent in Fargo said that answering these questions outs the respondent as someone who is associated with someone who is transgender, and that may bring about additional concerns. Another respondent in Fargo took particular issue with the first part of the two-step question via proxy:

“’To the best of your knowledge was BLANK’s sex recorded as male or female at birth,’ first of all, no one has the right to that information unless I have decided that they do. My roommate doesn’t need to know. My parents, they are aware, but unless I am living with my mother, my lover, or my doctor, it is not necessarily anyone’s business what I was assigned at birth and what my biological sex is. I think that you will either find a lot of push back or at the very least a lot of fear of someone being asked that question, whether they are the transgender person or the person they are living with is. Because that’s not their information to know.”

4.3.5 Differences by Household Member

Respondents who had more than one other person in their household were asked if question accuracy and sensitivity would differ depending on who in the household were answering.

4.3.5.1 Similarities across Groups

Respondents were split in their answers, with three respondents in each of the DC, Portland, and Fargo groups commenting that it would matter who in the household responded, and a few respondents in each of the groups saying it would not matter.

“If someone else were answering on my behalf, you’d either get the people who are answering and they don’t know [that] they don’t know and so they’re guessing, and you’d get the people [with whom] I’m basically stealthing, and so they think that they know but they know the wrong answer.”

Most of those who said it did matter commented on the inability of family (particularly parents) to correctly answer questions about gender identity. This included inability to acknowledge transgender status at all, as well as incorrect identification within the transgender umbrella term. For example, one respondent said that their family members were in denial about them being transgender, and that seeing them answer the question incorrectly would disappoint them and possibly end their relationship. There was a sense that some household members may be unable to answer beyond “some kind of transgender.”

“They generally know I’m transgender.”

One respondent said they would only be comfortable with having a spouse or partner answer, and another respondent talked about how their younger children would not be able to answer if asked.

On the other hand, those who said it did not matter indicated they were in supportive households, but said, “there are many, many people...for whom that is not the case.”

4.3.5.2 Differences across Groups

This question was not applicable for the two respondents in the Nashville group, who each lived with one other person in their household.

4.4 Question Wording

While not a primary focus of this study, we allotted time in the focus groups to collect feedback from transgender respondents on wording of questions about gender identity.

Even before seeing the example question wording, respondents in DC, Portland, and Fargo focus groups made references to potential question wording. Respondents also continued to comment about question wording during other sections. They felt strongly that the question wording must be appropriate and inclusive. While not of primary focus in this research, the salience of wording to respondents clearly indicates it is an area of important study.

Two respondents in DC, two respondents in Portland, and one respondent in Fargo discussed difficulties associated with asking about gender identity unprompted:

“It gets real confusing because...it’s so many categories in being trans, period. So you can never really do an accurate survey and get an accurate response because so many people identify as different things.”

“I want to add to I think it’s super important to be accounted for somehow. I know talking about the language we can’t even pin ourselves down, but it’s important that it happens.”

“The data is only going to be as good as the information that you let people choose. And especially phrasing a question on a survey like, ‘what is your gender identity,’ if you were to only offer two options then you would just...it would be kind of weird, but then how do you offer all of the options? Probably also impossible.”

One of these respondents said when there are inadequate categories, “it’s kind of like asking [you] to write down your age but you can only pick 20 or 40.” Another respondent in Fargo was concerned that the question may use terminology that is unfamiliar to non-English speakers or transgender people who are not an active part of the transgender community.

Two respondents in DC, three respondents in Fargo, and one respondent in Nashville thought that using open-ended questions was one potential solution, though they acknowledged that it would be more work for the government to code, and may encourage people to submit false answers like “Jedi” or “potato.”

“I think the only real way to ask about gender on a document is to just leave a blank, not try to do a checkbox situation. But then if it’s something that you’re trying to collect data on tons and tons of stuff that becomes a problem when you have a billion different answers. But I think that’s really the only way to do it both respectfully and in a useful way.”

“If the answer choices are male/female/transgender and then there was a blank for me to explain, I would put trans and then write ‘non-binary’ in the blank. So yes, I’m part of the trans community, but this is how I identify...that would just for me as a non-binary person encourage me to answer ‘trans’ and not just circle female and be like, ‘eh, whatever, it’s fine, it’s enough for me to just circle female.’”

Another question suggestion included using a scale, such as one with a spectrum from 1 to 10 with male on one side and female on another side, though other respondents disliked this because it may exclude some respondents.

Unintentionally closely anticipating the example questions later shown, one respondent said:

“If I have to take another study where it says ‘male/female/transgender’ I might just rip out all my hair.”

In the section of the guide specifically geared towards gathering feedback on question wording, respondents were first shown a one-step question and two-step question simultaneously on two separate cards:

One-step question

1. What is your current gender?

• Male
• Female
• Transgender
• Do not identify as male, female or transgender

Two-step question

2. Was your sex recorded as male or female at birth?

• Male
• Female

3. Do you describe yourself as male, female, or transgender?

• Male
• Female
• Transgender

Next, they were shown a card with this two-step question as well as variants on the second part of the two-step question:

Two-step question options

1. Was your sex recorded as male or female at birth?

• Male
• Female

2A. Do you describe yourself as male, female, or transgender?

• Male
• Female
• Transgender

2B. Do you describe yourself as male, female, or transgender or something else?

• Male
• Female
• Transgender
• Do not identify as male, female, or transgender

2C. How do you describe yourself? (Check one)

• Male
• Female
• Transgender
• Genderqueer/gender non-conforming/non-binary

2D. How do you describe yourself? (Check one)

• Male
• Female
• Trans male/Trans man
• Trans female/Trans woman
• Genderqueer/gender non-conforming/non-binary
• Different identity (please state): ___________________

Comments made by respondents that are not clearly tied to either the one-step question or the two-step question are discussed as part of the two-step question findings.

4.4.1 One-step Question

The one-step question was probed on simultaneously with the two-step question. Respondents were asked how they came up with their answers to both of these options. Next, they were asked what they see as the differences between the options, if one is easier to answer, and which one they preferred. Because respondents had many comments about the questions, not all follow-up probes were asked in every focus group due to time constraints.

The one-step question was worded as:

What is your current gender?

• Male
• Female
• Transgender
• Do not identify as male, female or transgender

4.4.1.1 Similarities across Groups

Respondents had fewer comments about the one-step question than the two-step question that they viewed simultaneously. Many comments about the one-step question were made comparing the two versions to each other.

Two respondents in Fargo explicitly said they really disliked this question. Respondents in Portland and Nashville felt like the options provided in the one-step question were limiting, with some respondents feeling like “transgender” as a response option was not sufficient.

“I personally [saw] the last option and well, that’s not telling of who I am as a person at all. I’m not these other things, so how do I answer it? I don’t know.”

“I picked transgender because I think of that as an umbrella term, and even though it’s really weird I was like, ‘well, I guess I’ll boost this number, of the options [available].’”

A couple of respondents described going back and forth between choosing their preferred gender term of male or female or selecting transgender instead.

“I felt like I had to choose between more than one thing and I chose male for that, but my first thought reading that was I want to be counted as trans. I don’t want that part of me to be invisible especially in a data type situation, but male feels more authentic to me.”

“It was hard to not just put down female since that’s how I would prefer to identify, but it needs to be transgender right now just because...right now I can’t just be female, for the community. I have to identify as transgender and be counted.”

“I would be tempted to circle female because you see those words used on medical forms, and if I were at the doctor’s office I would circle female because that’s what they want to know. I wouldn’t know what to do with this. I would probably just circle female because that’s appeasing to the source of this, but none of these feel right or seem correct to me.”

Two respondents each in DC and Portland disliked the language about “current gender” in the one-step question.

“[It makes it sound like] a kind of ‘gender day of the week thing.’ Which like, people totally are and that’s totally great, but it’s kind of [expletive] [for me].”

“I just feel like when they ask what your current gender is and those are the choices, I feel like I’ve been dissed...I am male but that’s not my gender, that’s my physical sex, because I still have all my parts and I probably still will.”

4.4.1.2 Differences across Groups

Three respondents in DC thought that the one-step question was easier to answer, but two respondents said it was harder to answer. Three additional respondents said they thought that the two-step question was better because it collected additional information. This may have been a more prominent line of discussion in this group because fewer respondents had non-binary identities in this group compared to the other groups.

“If I identified as a male today, [the two-step question] is going to tell them that I identify as a trans male basically, whereas [the one-step question] is not. I mean if I just put I identify as male then that’s all you got and you don’t know anything beyond that.”

“I think you will get more accurate data that way [the two-step question]...I think it would be the most accurate because it’s a way of forcing people without making them say ‘transgender’ which a lot of people don’t want to say.”

Two respondents in Fargo said the language of the one-step question seemed to be getting at respondents who are gender-fluid or have other identities that are not static, which was indeed the intention of the question.

“I crossed out the question and put instead, ‘how do you currently identify in terms of gender?’ to rephrase the question.”

“By ‘current’ do you mean my overall identification, or exactly at this second?”

4.4.2 Two-step Question

The two-step question was probed on simultaneously with the one-step question. Respondents were asked how they came up with their answers to both of these options. Next, they were asked what they see as the differences between the options, if one is easier to answer, and which one they preferred. Because respondents had many comments about the questions, not all follow-up probes were asked in every focus group due to time constraints. The two-step question was worded as:

1. Was your sex recorded as male or female at birth?

2. Do you describe yourself as male, female, or transgender?

• Male
• Female
• Transgender

4.4.2.1 Similarities across Groups

Reactions to this question were mixed. Three DC respondents said they liked the two-step question better than the one-step question generally. One additional respondent said their preference depends on the context; they would be worried if their answers were tied to their name. Respondents in Fargo did not explicitly voice a preference. Both respondents in Nashville preferred the two-step question. One respondent in Portland preferred this question over the one-step question, but the rest of the respondents said they did not like either.

“They’re all wrong; start over.”

“We don’t necessarily use those terms anymore.”

Overall, respondents did not like the first part of the two-step question. Two respondents in DC, four respondents in Portland, and two respondents in Fargo said they would not want to answer the first part or foresaw issues with others responding to it. One respondent in DC said that they would walk away from the survey when asked the first part of the two-step question, but another respondent said that sex assigned at birth is something with which respondents should come to terms.

“I don’t like it at all because why does it matter unless I’m going into the doctor and having a medical thing done what my sex was when I was born. Whether I have a penis or a vagina does not make any difference to anything that could be in [the survey], unless it’s medically related.”

“I’m never going to answer what my sex was recorded as. Like yes, it was recorded as male or female, whatever. There’s like two people who get in my pants and it’s my partner and my doctor and nobody else needs to know.”

“I also just don’t see how the first question could have any bearing on anything that’s happening toward going into the future.”

“Why do you need to know what my sex was at birth? Is that important information for you to know and if so explain to me why.”

“On the two step, ‘was your sex recorded as male or female at birth,’ I want to answer that question only under duress as male, but that’s only because I know it’s being recorded for some particular reason...There are a lot of other organizations, governmental, that ask that question that have no need to know the answers to that particular question.”

This may be due in part because respondents did not seem to understand how the two questions would be used together to identify transgender respondents through a mismatch in sex assigned at birth and gender identity (e.g., a respondent who answers “male” to the first question and “female” to the second). A few respondents in DC and Portland understood that the questions would be analyzed jointly, and tried to explain it to other respondents.

One respondent in Portland explicitly said they liked the two-step question as compared to the one-step, and another saw the value from a data perspective.

“To me it felt like I could both be authentic to myself and come out, and [on the one-step] I didn’t feel like I could do that, so I appreciated the [opportunity] to still count myself as trans here.”

“The other thing I was thinking of was if there was a good reason to try to capture the directions of transition separately. The differences in employment and suicide attempt rates and murder and all the stuff that happens to us. I think trying to capture that [has] some value.”

Despite the limited time allotted to the subject, respondents still had many criticisms and comments about the question wording. The most prominent were the inability to mark all that apply, lack of adequate response options, not using “cisgender,” and the language of the question stems.

There was a common sentiment across groups that many transgender people may identify as both transgender and something else, and therefore respondents should be able to mark all that apply. One gender-fluid respondent chose to mark all that apply anyway, and circled everything except for male at birth when answering the two-step question.

“I describe myself as both male and transgender so I don’t know which one to pick.”

“Someone can identify as male and transgender and that does not necessarily mean that they’re [not] a man.”

The lack of adequate response options was heavily discussed in the DC, Portland, and Fargo groups in particular. On the other hand, one respondent in DC said that “transgender” was not needed in the second question. The most frequent criticisms voiced were that there are not enough response options generally, and that some respondents do not describe themselves as transgender. A genderqueer respondent circled that they identified as a woman because they do not use the term transgender.

“I would say trans-‘modifier’, trans non-binary person. Because transgender is, I don’t know. I guess it’s some people’s gender identity but ‘trans man’ would be a gender identity that I hear people use. ‘Transgender’ is not something people use.”

Both respondents in Nashville said there is a distinction between using the term “transgender” to come up with a count of members of the community versus identifying those who specifically use the word “transgender” to describe themselves.

Two respondents in Portland wanted there to be more response options in order to provide data that are more granular.

“I will accept the term transgender but it’s too superficial as a label. It beats male or female as the only options but it doesn’t let me fill in enough information there.”

“There should be some sort of drop down or option or something that allows you to define under that umbrella.”

Some response options respondents specifically suggested including are:

• Intersex for both parts of the question (mentioned in Portland, Nashville, and Fargo)
• Agender (mentioned in DC)
• Genderqueer (mentioned in DC)
• Nongendered (mentioned in Fargo)
• Bigender (mentioned in Fargo)
• Gender-fluid (mentioned in Fargo)
• Non-binary (mentioned in Fargo)
• Other (mentioned in Fargo)
• Neither (instead of transgender) (mentioned in Fargo)
• Prefer not to answer (mentioned in Fargo)
• None of the above (mentioned in Fargo)

Many respondents also felt strongly that “cisgender” should be used. For example, they thought having cis male, cis female, trans male, and trans female as response options would be more appropriate. One respondent in Fargo said that “man” and “woman” should be used instead of “male” and “female” because “male and female are sexual terms, not gender-related terms.” There was a detailed discussion in the Portland group on this issue when the two-step variants were presented (see Section 4.4.3).

Respondents also commented on the language of the question stems. For the first question, one respondent in DC said they liked the language of “was your sex recorded as...” as opposed to something like “were you born as...” This got wide agreement from the group. Respondents in Portland also agreed with this.

An additional suggestion made during the concluding remarks of the Portland group was to ask, “‘Do you identify as a gender other than the one you were assigned at birth?’ Yes/No.” This suggested re-wording got some support from the group.

For the second question, one respondent said, “do you describe yourself as” was a strange wording. However, both respondents in Nashville liked this wording because it puts the emphasis on self-description rather than societal assumptions.

4.4.2.2 Differences across Groups

One respondent in the Nashville group did not seem to understand the intention of the first part of the question to capture what sex was recorded on the birth certificate. They seemed to conflate this with gender identity, reading it more as a “what you were born as” type of question.

“It depends upon who you ask. Technically, female. I was assigned male...We were assigned male or female but that doesn’t mean that’s what we are or ever were.”

One respondent in Fargo said that this question does not go far enough to provide information to inform differences in employment (i.e., the value of collecting gender identity on the CPS), and thought that there should be information collected on how others perceive their gender.

4.4.3 Two-step Question Variants

After discussing the one-step and basic two-step questions, respondents were presented with a card that had both the original two-step question as well as a few other options. Respondents were asked how they came up with their answers to these questions, which question they preferred, and what other answer choices they would like to see included. Because respondents had many comments about the questions, not all follow-up probes were asked in every focus group due to time constraints. The two-step question variants were worded as:

1. Was your sex recorded as male or female at birth?

• Male
• Female

2A. Do you describe yourself as male, female, or transgender?

• Male
• Female
• Transgender

2B. Do you describe yourself as male, female, or transgender or something else?

• Male
• Female
• Transgender
• Do not identify as male, female, or transgender

2C. How do you describe yourself? (Check one)

• Male
• Female
• Transgender
• Genderqueer/gender non-conforming/non-binary

2D. How do you describe yourself? (Check one)

• Male
• Female
• Trans male/ Trans man
• Trans female/ Trans woman
• Genderqueer/gender non-conforming/non-binary
• Different identity (please state): ___________________

4.4.3.1 Similarities across Groups

There was not a lot of direct comparison of the 2A-D alternatives against each other specifically; instead, respondents mostly spoke generally about weaknesses they saw with the questions more generally.

No respondents commented on 2A in this section; this is not surprising seeing as 2A was part of the initial two-step question introduced to respondents.

Only two respondents across all groups commented on 2B. One respondent said that they thought that 2A and 2B were the same question and did not like the response options. Another respondent commented that they thought the response options should match the question stem (i.e., use “something else” as a response option instead of “do not identify as male, female, or transgender”).

Three respondents in the DC focus group explicitly mentioned these question variants as being better than the prior questions, with one of these respondents saying they would be very happy with 2D if “cis” were added. Another respondent said they would prefer 2C if “cis” were added but would otherwise prefer 2B. The Fargo focus group generally agreed that 2C and 2D “seemed a little more well thought out” and one respondent in this group said they liked 2D. In Portland, a few respondents liked 2C, a few more liked 2D, and a few said they did not like any of them.

As with the basic two-step question, general themes in the discussion about all the questions included the inability to mark all that apply, lack of adequate response options, not using “cisgender,” and the language of the question stems.

While not discussed as frequently as it was with the previous one- and two-step questions, respondents in the DC and Fargo focus groups brought up again that respondents should be able to select all responses that apply. One DC respondent specifically commented on Questions 2C and 2D, both of which instructed respondents to check one response option. However, most of the discussion was about all of the question variants generally.

One respondent in DC said that they liked having “trans male/trans man” as a response option “because it kind of separates out of the transgender and it still says male.” However, this distinction was not useful for everyone.

“If you give me the options [of] male and transgender, I’m gonna go with male because that’s how I live my life and that’s how I see myself.”

This respondent also said when they first started transitioning they would have embraced the transgender label more. Another respondent said that they never use transgender as a label. The DC group generally agreed with the sentiment that a single option of transgender does not convey their full identities, as discussed by respondents in other groups in response to the two-step question earlier.

“I am transgender but I’m also still male and I don’t want that to be taken away from me.”

Respondents also made various suggestions regarding the response options that were very diverse and varied by group. Some of these suggestions were also voiced in response to the previous two-step question. Respondents suggested that we:

• Maintain current language:
• • Continue to use the word “transgender” (mentioned in DC)
  • Keep “at birth” language in the first step (mentioned in Fargo)
  • Keep genderqueer / gender nonconforming (mentioned in Nashville)
• Add new language:
• • More response options generally (mentioned in Portland and Fargo)
  • Trans masculine/trans feminine (mentioned in DC)
  • Gender-fluid (mentioned in DC)
  • Agender (mentioned in DC)
  • Non-binary³⁰ (mentioned in Portland)
  • A question in which all respondents, including those who are cis, must explain their answer (mentioned in Portland)
  • Intersex (mentioned in Nashville)
• Change existing language:
• • Put gender nonconforming into the non-binary category (mentioned in DC)
  • Use man and woman instead of male and female (mentioned in Portland)
  • Separate out identities instead of collapsing them into one line in 2C and 2D (mentioned in Portland)
  • Use non-binary instead of transgender (mentioned in Portland)
• Remove:
• • Gender non-conforming “because that is a gender presentation” (mentioned in Portland)

One respondent in Fargo was particularly frustrated by the lack of options:

“When they say transgender, what are they asking? I mean that is such a big [expletive] word that could include anything...I know that we’re the smallest [expletive] segment of the population, but we’re also the [most diverse] internally, and we all just kind of fit together, [but] give me more options.”

In Portland, there was a lot of discussion about the terms “male” and “female,” with three respondents saying that the terms referred exclusively to sex and genitalia. One respondent in Fargo brought this up earlier in response to the first two-step question shown (see Section 4.4.1).

“You know the term mixed metaphors? Male and female has to do with your anatomy.”

“For me, male and female are [what people use] when they ask you what gender you are but what they really want to know is what your sex assigned at birth was.”

“I personally feel like when people assign me male in public I honestly feel a little like, ‘oh, you think I have a penis.’”

One respondent in this group strongly disagreed:

“This is a pet peeve of mine in the trans community in general. We can get to these really academic definitions that I don’t see in the real world. I really don’t think non-trans people say male or female to refer to genitalia. I think that non-trans people use male in the exact same way as man and use female as woman...It’s not that I necessarily disagree that the academic definition of these things is [that male and female refer to genitalia], it’s that I think societal views are more important than the academic definitions.”

Five respondents in DC, one respondent in Portland, and two respondents in Fargo mentioned using “cis” as a prefix, with large group agreement in Fargo. Respondents generally thought this would be clearer for members of the transgender population and also “educate” those who are not. Two respondents said they thought this was important because the use of “transgender” as a separate response option or “trans” as a modifier implies that people who are transgender are not “normal.” This received some agreement in a couple of the groups.

Again, some of this discussion seemed to stem from the fact that respondents did not understand how two questions about gender identity would be used together to identify transgender respondents.

“Either clearly state that you mean cis male and cis female if you don’t want transgender people to check those, and if you don’t do that, then you need to provide more gender identifiers in the questions.”

On the other hand, one respondent in DC and three respondents in Portland argued that including “cis” would cause confusion for people who are cis but unfamiliar with the term.

“Is the purpose of the survey to educate non-trans people or is it to gather data?...I feel like I’m a little attuned to like, the fact that you need to collect the data and in order to do that the terms that you use have to be familiar to the people who you are asking the questions to.”

One respondent in Portland said the question should be, “Are you cis?” As another potential alternative, one respondent suggested using “non-trans” instead of “cis.”

While not discussed as much as with the previous gender identity questions, respondents in the DC and Fargo focus groups discussed the “how do you describe yourself?” question stem. One respondent said they were not sure if this was trying to get at sex or gender. Another respondent said it should be replaced by “what is your gender?”

“It’s not how I describe myself, it’s not how I choose to tell you who I am, it’s what I am.”

“I feel like saying how you describe yourself is how you want to be, whereas this is who I am.”

This sentiment was agreed on by a few other members of the DC focus group. On the other hand, one respondent in Fargo liked the language of this stem and also liked “current gender” used in the one-step question because “gender identity is not static.”

4.4.3.2 Differences across Groups

There were mixed opinions on question 2D. Both of the respondents in Nashville liked this question.

“[It] has the option to state a different identity, so that’s the most inclusive which is good.”

“You’ve included the main types that people would identify as.”

On the other hand, a few respondents in Fargo felt very negatively about the answer choices for 2D. One respondent said:

“It’s obvious that 2D was not created by a gender aware person...it just made me feel very uncomfortable to look at that.”

Four respondents in DC and one respondent in Nashville were unsure how to answer 2D, as they felt it forced them to choose between being counted as part of the transgender population and reflecting their gender identity accurately. One respondent in DC said that even with the transgender response option, they would be inclined to put male, as that is how they self-identify. These comments were similar to those voiced in response to the two-step question earlier.

“I would describe myself as female. That’s how I would prefer to describe myself but for what you’re looking for it would be transgender.”

One respondent in DC and two respondents in Portland were skeptical of the utility of an “other, specify” response option.

“Whenever you do a write in...you just get washed out. [I saw] ‘other’...and I was kind of like, that’s gonna be like one percent and no one is really going to look into that, so just wanting to be seen I put [selected] trans male.”

While most of the respondents across groups had criticisms of the gender identity questions, one respondent in Fargo had an immediate negative, visceral reaction to question shortcomings:

“I’m physically ill. I’m dead serious. I mean I don’t want to make progress like this. Here are the issues for me. Male and female are biological terms, they are not gender terms, so when you put those and then add transgender, there you are comparing apples to oranges number for one. But also if you look at 2D, you’ve got male, female, and then trans male, trans female, genderqueer, you’re still setting male and female out as distinct, as different...when I fill this out I feel more excluded and more isolated than I would if I were not able to have my identity recognized on here.”

4.5 Concluding Remarks

To conclude the focus group, we asked respondents how we should word or introduce gender identity questions to address any concerns or hesitation that respondents may have when answering. This included both responding for themselves as well as for responding for other members of the household. We also gave respondents one more opportunity to voice opinions, and asked respondents for additional suggestions on how to improve the collection of gender identity information.

4.5.1 Similarities across Groups

Most of the comments across groups were on data use, confidentiality, and interviewer training. In Portland and Nashville, respondents reiterated comments about the importance of being clear on how gender identity information will be used. Respondents in Nashville said that additional information about data use can help explain why respondents should “out [themselves] to the government if [they] haven’t already.” One respondent in Portland said that the blurb we provided in Section 4.2.6 to explain potential CPS data collection was reassuring.

“Are you just wanting to know the number of transgender people in the U.S., or do you want to know the amount for a specific reason? Do you need it to go for health care, or to Congress, or to other services that the government offers?”

“It should be apparent what that information might be used for because that might inform how someone chooses to answer.”

In DC and Nashville, there were a few more comments about confidentiality concerns.

“The minute you ask for my name, I’m shutting down. Or even my address.”

“You have transgender people out there that have made no transition at all that would identify but they’re too scared to; they’re definitely not going to identify on a census for it. You don’t give anything to the government you don’t have to.”

Respondents in the Portland and Fargo groups also emphasized the importance of proper interview protocol and interviewer training. They said it was important for everyone to be asked the same gender identity questions, and to be asked them explicitly instead of the interviewer assuming whether or not someone is transgender. One respondent in Portland added that whether or not they disclose they are transgender “comes down to a gut feeling [of], ‘is this person safe to disclose to?’”

“Make sure your [interviewers], the people actually asking these questions, have had some sort of training or done some sort of research or [have] a little bit of awareness, and so you’re not just having some regular cisgendered person who only understands transgender as a basic term, who has no understanding of anything else. Make sure they really know what they’re asking.”

“Make sure they have the dignity to not make weird faces.”

“A lot of times you’ll hear someone’s voice and then they’ll assume your gender immediately based on that, and then that’s a person gone. They’re not going to talk to you again.”

4.5.2 Differences across Groups

Some of the concluding probes were asked spontaneously in the DC group after the recorder was turned off, and not all of the discussion was captured in the notes. Structured probes were added to the guide for the remaining groups.

In Fargo, there was discussion on the usefulness of leveraging community organizations, providing a centralized location for safe survey-taking, and involving transgender individuals in this line of research.

Two respondents in the Portland focus group and one respondent in the Fargo focus group took the opportunity to emphasize here the importance of collecting data about the transgender population.

“I have problems with it, and I want it asked. And I think that’s something about this. There’s no perfect way to ask these questions and there’s definitely no way you can ask these questions that every trans or non-binary person is going to agree on...And so for me, I hope that whatever conclusions you’re able to draw based on this, the fact that you’re not going to be able to get it right doesn’t stop you from doing it.”

“It’s forward progress, and just because it’s not completely accurate...it’s a step, and I generally think it’s good to see the questions and it’s not just ‘male/female.’ That alone is a positive thing.”

5. CONCLUSIONS

5.1 Summary of Findings

5.1.1 How do transgender respondents define “transgender” and “gender identity,” and relate these terms to their self-identity?

Respondents generally agreed that gender identity involves how someone feels and sees themselves, with some respondents commenting that the term “gender identity” makes them think of “anyone who is not cisgender” or “cis.”³¹ Some respondents also thought that people’s gender identities involve an interaction between their sex and gender, while others disagreed.

Respondents were also generally able to agree that to be “transgender” means to be the opposite of “cisgender.” While some respondents said that they would not use the word “transgender” to describe themselves personally, most agreed that it could be used as an umbrella term to describe members of a diverse community.

Respondents in the focus groups self-identified in a variety of ways, with relatively few respondents electing to describe themselves as “male,” “female,” or “transgender.” For some respondents, defining their gender identity was a difficult process, while others said that it was easy. A common sentiment across the groups was that everyone has their own journey in deciding how to self-identify, and that their self-identification varies over the course of that journey. Some respondents said that the disclosure of their gender identity depends on the setting, with some hesitant to disclose it in work situations, for example.

5.1.2 How do transgender respondents feel about the collection of gender identity information by the Federal government, both generally and in the context of the CPS?

While nearly all respondents were familiar with surveys, very few reported experience with government surveys in particular. Most respondents thought that government surveys should include gender identity questions for reasons including getting a count of the transgender population, using this count to advocate and allocate funding, and understanding potential discrimination, amongst other things. However, there were a few comments in each group from respondents who were skeptical the data would be used at all or were generally distrustful of the government. The question of who in the government would be using the information was raised in two of the groups.

The majority of respondents’ comments were regarding Federal surveys in general. Some respondents were concerned about having their gender identity tied to their name or other identifying information, and some respondents said that they do not always disclose their gender identity in order to protect their safety, avoid inappropriate questions, or prevent harm to their career prospects. Concerns about general malicious use and confidentiality/anonymity were amplified in the three focus groups outside of Washington, DC that were conducted in March, April, and June 2017. In Portland, there were also concerns about interviewer misconduct resulting in personal information being shared in non-official ways.

A later explanation of the reasons for possible collection in the CPS was reassuring for respondents, and the idea of collecting gender identity information for the purposes of the CPS received wide support across the focus groups. A few respondents talked about it being useful to see how unemployment differs between people who are transgender and not transgender, but most had just general reactions to the idea as a whole. Nonetheless, most of the respondents did have some sort of concern about how this data would be collected by the Federal government and how the information would be used. Concerns voiced by respondents in all four groups included leak of information outside of the government, general malicious use, and confidentiality/anonymity.

Across all four groups, several of the respondents who shared concerns about the government collecting information about gender identity also explicitly stated that they support the idea of it, suggesting there may be internal tension between understanding the value of having statistics on the transgender population and hesitating to provide their own personal information. Respondents suggested being clear about data use and confidentiality, as well as providing adequate interviewer training. Some respondents suggested involving LGBT organizations in data collection.

5.1.3 How do transgender respondents feel about gender identity information being collected via proxy response?

The immediate reaction to the idea of proxy response was negative in three out of the four focus groups. Some respondents explicitly asked why proxy response had to be used. The two respondents in the Nashville focus group were okay with proxy response, though this may be a result of their two-person households, and there were a few respondents in each of the other groups who felt similarly.

Respondents were concerned about issues with accuracy and the inappropriateness of someone answering gender identity questions on someone else’s behalf. While not all of these concerns applied to respondents personally, they reported that they knew people for whom proxy response would be problematic.

A handful of respondents in each of the DC, Portland, and Fargo groups said household members would refuse to report gender identity on their behalf. Very few respondents thought that members of their household would answer accurately. They were concerned about the CPS respondent being unaware of household members’ correct gender identity or refusing to accept it. Parent-child and roommate relationships were cited as being particularly difficult.

Most respondents felt proxy response of gender identity was sensitive, citing concerns about general confidentiality, potential for sharing person information that the person would not want shared, and safety.

Respondents were split on whether it would matter who in the household was answering on their behalf. Most of those who said it did matter commented on the inability of family (particularly parents) to correctly answer questions about their gender identity; however, others felt the answer was likely to be inaccurate regardless of who in the household was giving the proxy response.

5.1.4 What feedback do transgender respondents have on wording of questions about gender identity?

Respondents in all four groups gave strong reactions to the presented survey question wording. Some respondents explicitly said that they thought it would be difficult for researchers to create questions with adequate response options, given the diversity of terms used and debate within the transgender community itself about terminology. Respondents saw shortcomings with all of the questions that were presented, but the question with the largest number of current gender identity response options (question 2D, on the two-step variant card) was seen as the most promising. Not all respondents were clear on how both of the questions in the two-step questions would be used together to identify transgender respondents.

Some respondents were uncomfortable selecting the transgender response option to describe themselves. There were also respondents who were unsure, unwilling, or uncomfortable identifying as transgender instead of male or female. General question criticisms and comments included the inability to mark all that apply, lack of adequate response options, not using “cisgender,” and the language of the question stems. None of the question versions tested would be recommended for use with the transgender community without further testing.

5.2 Study Limitations

Overall, results from these focus groups suggest that respondents see the importance of Federal collection of gender identity information, but the methods through which data collection occurs is critical. Respondents’ comments made it clear that there are weaknesses in proxy measurement and question wording that present obstacles to accurate collection of gender identity on the CPS for which there are no clear solutions. Survey context and Federal collection of the information is a lesser concern, but was raised across all four groups. It is very important to consider these obstacles carefully, as they have the potential to create serious classification errors.

These focus groups were just one part of a larger study on the feasibility of asking about gender identity on the CPS. A decision on overall feasibility of collecting SOGI information in the CPS should consider the findings of the focus groups as well as those of the cognitive test on gender identity questions (Ellis et al., 2017).

This qualitative study was carefully designed to evaluate the feasibility of asking gender identity questions in the context of an employment survey—specifically the CPS—which relies on proxy response. While the above described results are sound and provide valuable information in response to the specified research questions, there are some limitations that need to be kept in mind when considering the implications of the findings.

5.2.1 Qualitative Research

Focus groups are a qualitative research methodology that is not designed produce point estimates or standard errors, or to be representative of any given population. While we attempted to recruit respondents from a variety of backgrounds and conducted four groups in regionally diverse locations, it is necessary to remember that the results reported in this study may or may not reflect the reactions other respondents may have.

5.2.2 Respondent Characteristics

Although attempts were made to recruit respondents with a wide variety of demographic characteristics and backgrounds, the nature of the recruiting methods (e.g., Craigslist ads mentioning LGBT; use of a known LGBT contractor) may have brought in people who are more actively aware of and advocate for transgender issues. As a result, these respondents may react differently to gender identity questions.

Additionally, as typical with these type of studies that rely on volunteer samples, respondents to this study are by nature more cooperative and comfortable with the Federal government and/or research studies than actual respondents outside the lab setting.

5.2.3 Testing Locations

The four data collection locations (Washington, DC; Portland, OR; Nashville, TN; and Fargo, ND, and their outlying areas) were selected with the goal of collecting information from a variety of respondents with differing experiences and cultural backgrounds. While these cities were expected to represent a variety of cultural perspectives, they are not expected to be comprehensive. Additional testing in other locations will reveal the extent to which the opinions expressed by the respondents in this study are similar or different from those in other regions of the country.

5.2.4 Respondent Understanding of CPS Methodology

It is difficult to fully explain the Federal survey process overall, let alone the CPS in particular, and how gender identity would be collected in that context, without actually exposing respondents to the CPS interview. We erred on the side of giving respondents less information in order to avoid biasing them, but this sometimes resulted in misunderstandings we would not expect to see in actual data collection, such as the conflation of the CPS with a job application or the belief that survey data is not kept confidential. Lack of exposure to the CPS interview also meant we were unable to see how respondents would react to being asked to answer gender identity questions by proxy in context. The focus groups were more hypothetical in nature, and opinions may or may not be predictive of behaviors (Fazio, 1986; Horwitz & Finamore, 2017).

Additionally, while we tried to ask respondents about gender identity data collection in the context of current CPS methodology (e.g., by discussing survey mode and use of proxy response), we were not always successful in communicating that some aspects of data collection are nonnegotiable. This resulted in respondents discussing infeasible options, such as a public forum for data collection. In future work, it will be preferable to conclude with probes that force respondents to choose between imperfect measurement and no measurement at all.

5.2.5 English Language Only

These focus groups were conducted in English, and all respondents spoke English fluently. We anticipate cultural and language issues may arise when translating gender identity questions to other languages. The CPS is regularly administered in English and Spanish, and translators are called on when necessary for other languages. Thus, translation and accompanying cultural issues need to be explored before adding gender identity questions to the CPS.

5.3 Recommendations and Future Research

If it is deemed feasible to include gender identity questions in the CPS based on this study and its companion cognitive interview study, the next steps are to identify the outstanding potential issues that need to be addressed by future research, such as:

• Question wording, and wording of response categories
• Translation and cultural issues for non-English populations
• Impact of survey administration mode on respondent reactions
• Further examination of the sensitivity of questions, and whether this varies by demographics
• Optimal question placement within the CPS
• Appropriate age cutoff for questions, and procedures for obtaining consent
• Quality of estimates generated using the CPS, including whether the sample size would be sufficient to develop reliable labor force estimates for the LGBT population, or an analysis of the likely error bounds of such estimates
• Comparison of methodologies and estimates of gender identity questions included in other surveys

We emphasize that there remain serious concerns about classification error due to the small estimated size of the transgender population. Mistakenly classifying respondents who are not transgender as transgender, or vice versa, would likely increase the statistical error in population estimates, although the full extent and statistical consequences of these errors are beyond the scope of this research. We cannot yet make any conclusions about the quality of data these questions would collect on the CPS.

Given the dearth of research available on this topic, we encourage researchers working on other surveys to further explore perceptions and accuracy of proxy data collection of gender identity items. This could be done with additional cognitive interviews and focus groups, with both transgender and non-transgender respondents, as well as larger scale feasibility and field testing to understand item nonresponse, response distributions, impact on response rate, and attrition. In addition, we have specific recommendations related to proxy response, survey context, and question wording.

5.3.1 Proxy Response

Currently, the only option for collection of gender identity in the CPS is for one household member to report about all members of the household including themselves. With a few exceptions, respondents were concerned about the accuracy of gender identity information reported by proxy. They also thought it was inappropriate for someone to answer these questions on the behalf of someone else. Consequently, there may be reason to believe that these concerns could result in an undercount of the transgender population. This undercount could be amplified in households with parent-child or roommate relationships.

Respondents’ negative reaction to the use of proxy response indicates that collection of gender identity on the CPS may not be feasible.

5.3.2 Survey Context

Another area of study relevant to the collection of gender identity on the CPS is the potential for respondents to view questions as irrelevant on a survey that is not about health. If respondents view these questions as irrelevant to the subject matter of the survey, and therefore feel like their agreement to participate in the CPS is being taken advantage of, they may refuse to answer the questions, breakoff from the survey entirely, or refuse to participate in subsequent interviews.

A few respondents were initially hesitant about Federal collection of gender identity and wanted to know how the data would be used. However, the collection of gender identity on the CPS specifically was widely embraced by respondents. Thus, we found little evidence that the survey context of employment is of significant concern for the CPS among transgender individuals. It should be noted that respondents were provided with an explanation of statistics that may result from this collection of gender identity information.

If the CPS were to add questions about gender identity in the future, we would recommend testing a scripted help text for interviewer use if they encounter respondents who are skeptical of the relevance of the question. This could be similar to the text provided to interviewers in the NCVS.

5.3.3 Question Wording

Finally, while question wording was not a primary focus of this research, comments made by respondents suggest that, contrary to prior cognitive testing, current questions may be inadequate. Respondents’ feedback illustrated the difficulties inherent in designing questions that are acceptable to transgender respondents but do not introduce statistical error. Even with a “transgender” response option, some transgender respondents may select “male” or “female” instead. If the CPS were to add questions about gender identity in the future, we recommend considering revisions and conducting additional cognitive testing on revised wording.

We suggest additional research on the following aspects of question wording in particular:

1. Evaluate stems of questions about gender identity, as respondents differed in their opinions of phrases such as “current gender” and “do you describe yourself as...”
2. Test the two-step question with gender identity asked first. While we do not have direct evidence of this, we suspect that that asking gender identity prior to sex at birth may be clearer for respondents who do not understand how the two questions work together.
3. Consider adding broad categories such as “other,” “none of these,” or “something else.” While it is almost certain that responses in this category would be collapsed up for analysis, the existence of an alternative option seems to be reassuring for some transgender respondents. The NCVS uses “none of these.”
4. Continue to test “male/female” versus “man/woman” wording. Cognitive testing by Martinez et al. (2017) found that cisgender respondents preferred male/female wording, but understood both equally. One potential concern with man/woman wording is that younger respondents may not feel old enough to endorse these labels.
5. Explore the feasibility of allowing respondents to mark all that apply.
6. Conduct testing on the use of the gender-neutral pronouns “they/their/them” for transgender respondents in order to simplify question fill patterns. Currently, the CPS and other Federal surveys fill many questions with “he/she,” “his/her,” or “him/her” pronouns after sex is determined. However, this does not seem advisable for transgender respondents given that they may differ in their preferred pronouns. This is particularly problematic for health surveys with sex-specific questions (e.g., pregnancy), but because the CPS generally does not have this limitation (though some CPS supplements could be affected, such as the biannual supplement on fertility), “they/their/them” may be a suitable alternative for transgender respondents.

We encourage researchers interested in survey measurement of gender identity to test these question changes, as improvements to the wording will benefit all surveys currently collecting or considering collecting gender identity.

6. References

Adams, C. (2017). "The gender identity terms you need to know." CBS News. Retrieved from https://www.cbsnews.com/news/transgender-gender-identity-terms-glossary/.

Baker, K., & Hughes, M. (2016). "Sexual Orientation and Gender Identity Data Collection in the Behavioral Risk Factor Surveillance System." Center for American Progress. Retrieved from https://cdn.americanprogress.org/content/uploads/2016/03/05064109/BRFSSdatacollect-brief-04.05.17.pdf.

Bickart, B.A., Blair, J., Menon, G. & Sudman, S. (1990). "Cognitive Aspects of Proxy Reporting of Behavior." In M. Goldberg, G. Gorn, and R. Pollay (Eds.) Advances in Consumer Research, Vol. 17 (pp. 198-206). Provo, UT: Assoc. for Consumer Research.

Boehm, L.M. (1989). "Reliability of proxy response in the Current Population Survey." In Proceedings of the American Statistical Association, Survey Research Methods Section, pp. 486-489. Retrieved from https://www.researchgate.net/publication/258332621_The_Reliability_of_Proxy_Response_in_the_Current_Population_Survey.

Cahill, S., Singal, R., Grasso, C., King, D., Mayer, K., Baker, K. & Makadon, H. (2014). "Do Ask, Do Tell: High Levels of Acceptability by Patients of Routine Collection of Sexual Orientation and Gender Identity in Four Diverse American Community Health Centers." PLoS One, 9: e107104.

Ellis, R., Virgile, M., Holzberg, J., Nelson, D., Edgar, J., Phipps, P., & Kaplan, R. (2017). "Assessing the Feasibility of Asking about Sexual Orientation and Gender Identity in the Current Population Survey: Results from Cognitive Interviews." Technical Report, Center for Survey Measurement, U.S. Census Bureau; Office of Survey Methods Research, Bureau of Labor Statistics.

Fazio, R.H. (1986). "How Do Attitudes Guide Behavior?" In R.M. Sorrentino and E.T. Higgins (Eds.) Handbook of Motivation and Cognition: Foundations of Social Behavior. New York: Guilford Press.

Federal Interagency Working Group on Improving Measurement of Sexual Orientation and Gender Identity in Federal Surveys. (2016a). "Current Measures of Sexual Orientation and Gender Identity in Federal Surveys." Retrieved from https://s3.amazonaws.com/sitesusa/wp-content/uploads/sites/242/2014/04/WorkingGroupPaper1_CurrentMeasures_08-16.pdf.

Federal Interagency Working Group on Improving Measurement of Sexual Orientation and Gender Identity in Federal Surveys. (2016b). "Evaluations of Sexual Orientation and Gender Identity Survey Measures: What Have We Learned?" Retrieved from https://s3.amazonaws.com/sitesusa/wp-content/uploads/sites/242/2014/04/Evaluations_of_SOGI_Questions_20160923.pdf.

Federal Interagency Working Group on Improving Measurement of Sexual Orientation and Gender Identity in Federal Surveys. (2016c). "Toward a Research Agenda for Measuring Sexual Orientation and Gender Identity in Federal Surveys: Findings, Recommendations, and Next Steps." Retrieved from https://s3.amazonaws.com/sitesusa/wp-content/uploads/sites/242/2014/04/SOGI_Research_Agenda_Final_Report_20161020.pdf.

The GenIUSS Group. (2014). "Best Practices for Asking Questions to Identify Transgender and Other Gender Minority Respondents on Population-Based Surveys." J.L. Herman (Ed.). Los Angeles, CA: The Williams Institute.

GLAAD. (2017). "GLAAD Media Reference Guide—Transgender." Retrieved from https://www.glaad.org/reference/transgender.

Grieco, E.M. & Armstrong, D.M. (2014). "Assessing the 'Year of Naturalization' Data in the American Community Survey: Characteristics of Naturalized Foreign Born Who Report—and Don't Report—the Year They Obtained Citizenship." Paper presented at the Applied Demography Conference: San Antonio, TX.

Horwitz, R., & Finamore, J. (2017). "Do focus group participants mean what they say? A field test of mailing materials updated based on qualitative feedback." Presented at the 7th Conference of the European Survey Research Association: Lisbon, Portugal.

Hughes, I.A., Houk, C., Ahmed, S.F., Lee, P.A., & LWPES1/ESPE2 Consensus Group. (2006). "Consensus statement on management of intersex disorders." Archives of Disease in Childhood, 91(7): 554-563.

Institute of Medicine (IOM). (2011). "The Health of Lesbian, Gay, Bisexual, and Transgender People: Building a Foundation for Better Understanding." Washington, DC: The National Academies Press. Retrieved from http://www.nationalacademies.org/hmd/Reports/2011/The-Health-of-Lesbian-Gay-Bisexual-and-Transgender-People.aspx.

Joloza, T., Evans, J., O'Brien, R., & Potter-Collins, A. (2010). "Measuring Sexual Identity: An Evaluation Report." Office for National Statistics, United Kingdom. Retrieved from https://www.ons.gov.uk/ons/rel/ethnicity/measuring-sexual-identity---evaluation-report/2010/sexual-identity.pdf.

Kojetin, B.A. & Mullin, P. (1995). "The Quality of Proxy Reports on the Current Population Survey." American Statistical Association, Proceedings of the Survey Methods Research Section, pp. 1110–1115.

Krueger, R.A. (1998). Analyzing & Reporting Focus Group Results. Thousand Oaks, CA: Sage Publications.

Lombardi, E. & Banik, S. (2016). "The Utility of the Two-Step Gender Measure within Trans and Cis Populations." Sexuality Research and Social Policy, 13(3), 288–296.

Martinez, M., Henderson, A., Luck, J. & Davis, M.C. (2017). "Cognitive Pretesting of the National Crime Victimization Survey Supplemental Victimization Survey." Retrieved from https://www.census.gov/srd/papers/pdf/rsm2017-03.pdf.

Mellow, W. & Sider, H. (1983). "Accuracy of response in labor market surveys: Evidence and implications." Journal of Labor Economics, 1(4), 331–344.

Moore, J.C. (1988). "Miscellanea, self/proxy response status and survey response quality, a review of the literature." Journal of Official Statistics, 4(2), 155–172.

Organisation Intersex International in the United States of America (OII-USA). (2013). Submission to the National Institutes of Health Request for Information (RF I): Inviting Comments and Suggestions on the Health and Health Research Needs, Specific Health Issues and Concerns for Lesbian, Gay, Bisexual, Transgender, and Intersex (LGBTI) Populations. Retrieved from http://oii-usa.org/wp-content/uploads/2013/10/NIH-Submission.pdf.

Ortman, J.M., Bates, N., Brown, A., & Sawyer R.C. (2017). "Optimizing Self and Proxy Response to Survey Questions on Sexual Orientation and Gender Identity." Paper presented at the Annual Conference of the Population Association of America, Chicago, IL. https://paa.confex.com/paa/2017/mediafile/ExtendedAbstract/Paper13962/Ortman%20Bates%20Brown%20Sawyer%20PAA%202017.pdf.

Park, J.E. (2015). "Proxy Reporting of Sexual Orientation Information in Household Surveys." Internal Presentation, Office of Statistical and Science Policy, Office of Management and Budget.

Park, J.E. (2016). "Improving Measurement of Sexual Orientation and Gender Identity in the Federal Statistical System." Presentation at the Council of Professional Associations on Federal Statistics (COPAFS) Quarterly Meeting, September 2016.

Pascale, J. (2016). "Modernizing a Major Federal Government Survey: A Review of the Redesign of the Current Population Survey Health Insurance Questions." Journal of Official Statistics 32, 2, 461–486. http://dx.doi.org/10.1515/JOS-2016-0024.

Pierce, J.P., Farkas, A., Evans, N., Berry, C., Choi, W., Rosbrook, B., Johnson, M., & Bal, D.G. (1993). "Tobacco use in California, 1992. A focus on preventing uptake in adolescents." Sacramento, CA: California Department of Health Services.

Reisner, S.L., Conron, K.J., Tardiff, L.A., Jarvi, S., Gordon, A.R., & Austin, S.B. (2014). "Monitoring the Health of Transgender and Other Gender Minority Populations: Validity of Natal Sex and Gender Identity in a U.S. National Cohort of Young Adults." BMC Public Health, 14, 1224.

Sexual Minority Assessment Research Team (SMART). (2009). Best Practices for Asking Questions about Sexual Orientation on Surveys. Los Angeles, CA: The Williams Institute, UCLA School of Law. Retrieved from https://williamsinstitute.law.ucla.edu/wp-content/uploads/Best-Practices-SO-Surveys-Nov-2009.pdf.

Tamborini, C.R. & Kim, C. (2013). "Are proxy interviews associated with biased earnings reports? Marital status and gender effects of proxy." Social Science Research, 42(2), 499–512.

Tate, C.C., Ledbetter, J.N., & Youssef, C.P. (2012). "A two-question method for assessing gender categories in the social and medical sciences." Journal of Sex Research, 50(8), 767–776.

Truman, J., Morgan, R., Gilbert, T., & Vaghela, P. (2017). "Measuring Sexual Orientation and Gender Identity in the National Crime Victimization Survey." Paper presented at the Joint Statistical Meetings, Baltimore, MD.

University of California, Berkeley. (2017). "LGBTQ+ Resources: Definitions of Terms." Gender Equity Resource Center. Retrieved from https://cejce.berkeley.edu/geneq/resources/publications-media-faqs/resources-classrooms-and-groups.

University of California, Davis (2017). "Lesbian, Gay, Bisexual, Transgender, Queer, Intersex, Asexual (LGBTQIA) Resource Center Glossary." Retrieved from https://lgbtqia.ucdavis.edu/educated/glossary.html.

U.S. Census Bureau. (2017a). "Current Population Survey: Methodology." Retrieved from https://www.census.gov/programs-surveys/cps/technical-documentation/methodology.html.

U.S. Census Bureau. (2017b). "Basic CPS Items Booklet: Demographic Items." Retrieved from https://www2.census.gov/programs-surveys/cps/techdocs/questionnaires/Demographics.pdf.

World Health Organization (WHO). (2014). "What Do We Mean by 'Sex' And 'Gender'?" Retrieved from http://www.who.int/gender-equity-rights/understanding/gender-definition/en/.

7. Glossary

Last Modified Date: July 10, 2024