At a job-market event a few years ago, Brian presented results from our research using randomized controlled trials (RCTs) to shift public opinion to be more supportive of same-sex marriage. He self-identified as LGBTQ within the context of a response to a question during the Q&A. Toward the end of the session, an older white male scholar raised his hand to make a comment in the form of a question: “So…you’re not really a social scientist, you’re just some kind of gay activist?” Even after presenting quantitative results from empirical studies, Brian was still perceived as having an agenda in a way that he would not have if he were presenting his dissertation work on public opinion about the presidency.
Claims of “having an agenda” or “being an activist” are not new to scholars who research identity in political contexts including sexual orientation and gender identity but also racial and ethnic identities, as explored by other contributors to this symposium. These claims are particularly familiar to researchers like us who use RCTs to explore causal mechanisms that affect attitudes and behaviors. Often these accusations have another implication: that being activists comes at the expense of being true, unbiased social scientists. Because of the potential large-scale, real-world implications of RCTs, ingroup and outgroup positionality becomes a particularly controversial and important issue when research involves interventions. This article presents our counterarguments to those claims and to attempts to dismiss our and other scholars’ work as conducting so-called me-search. We also introduce a basic conceptualization for the study of underrepresented groups by researchers who do not identify with those groups, using LGBTQ studies as an example.
Claims of “having an agenda” or “being an activist” are not new to scholars who research identity in political contexts, including sexual orientation and gender identity but also racial and ethnic identities.
IDENTITY AND POSITIONALITY
Considerable attention has been given to the role of researcher identity and how it potentially legitimizes or delegitimizes the work (Dean et al. Reference Dean, Furness, Verrier, Lennon, Bennett and Spencer2018; Holmes Reference Holmes2020). As researchers, we should appraise and scrutinize our own research methods and engage in “methodological reflexibility” (Cassell et al. Reference Cassell, Buehring, Symon, Johnson and Bishop2005; Johnson and Duberley Reference Johnson and Duberley2003). As discussed in the introduction to this symposium, much of the scholarly discourse around positionality and researcher identity focuses on observational research. However, complex questions about their role in work involving RCTs remain, particularly around research with normative goals and partnering with advocacy groups (Davis and Michelitch Reference Davis and Michelitch2022). This article focuses on the following five points, with a particular focus on LGBTQ identity, highlighting each in turn:
-
(1) Some identity groups (almost) inevitably will be underrepresented among scholars and will continue to be severely understudied if researchers are required to be members of the groups that they study.
-
(2) Research on identity groups conducted by members of those groups should not be disqualified as me-search because doing so ignores legitimate expertise and insight that those researchers bring to the study of their own identity groups. It also risks further delegitimization and marginalization of research about marginalized groups.
-
(3) One researcher cannot represent or have experienced all of the perspectives of a given community; therefore, requiring researchers to be members of the groups that they study risks reifying those groups. Alternatively, a researcher should not be required to self-disclose any identity.
-
(4) Researchers who do not identify with the groups that they study should ensure that their research benefits the marginalized group and they should take precautions to not exploit, misunderstand, or devalue the marginalized groups during the research and in any subsequent publications.
-
(5) Acknowledging researcher positionality and the use of methods such as RCTs have the potential, in some cases, to address concerns about researcher identity without compromising points 1–4.
Underrepresented Groups and Researcher Representativeness
Political science traditionally has been overwhelmingly dominated by cisgender, heterosexual white men. Recent strides have been made: there is increased gender parity (i.e., between cisgender men and women) in political science. For example, the current editorial board of the discipline’s flagship journal, the American Political Science Review (APSR), is 100% women. Despite these high-profile advances, publication patterns continue to show that women scholars are disproportionately underrepresented in publications and that a significant gender gap in publication rates for men and women remains (Teele and Thelen Reference Teele and Thelen2017).
The discipline of political science also has been historically disinterested in LGBTQ politics; work on LGBTQ issues is more common in other disciplines, including sociology and psychology (Tadlock, Taylor, and Brettschneider Reference Tadlock, Taylor, Brettschneider, Brettschneider, Burgess and Keating2017). Until recently, there were few openly LGBTQ political scientists conducting research on their ingroup. There are exceptions, of course, including the groundbreaking work of Sherrill (Sherrill Reference Sherrill1996; Wolinsky and Sherrill Reference Wolinsky and Sherrill1993) but his example is the exception, not the norm. It is difficult to quantify the percentage of openly LGBTQ academics (and openly LGBTQ political scientists) in the United States but it is likely to be relatively small (<10%), reflecting the estimated size of the adult LGBTQ population. The percentage is likely even smaller for subgroups such as transgender people, who are estimated to be 0.06% of the population (Flores et al. Reference Flores, Herman, Gates and Taylor2016).
In terms of research on LGBTQ rights, much of the focus has been on issues that disproportionately affect LG people rather than the entire community. Although some notable political scientists study transgender politics (Currah Reference Currah2008; Currah, Juang, and Minter Reference Currah, Juang and Minter2006; Murib Reference Murib2020; Taylor Reference Taylor2007; Taylor, Lewis, and Haider-Markel Reference Taylor, Lewis and Haider-Markel2018; Wuest Reference Wuest2021), smaller segments of the LGBTQ community—including transgender, gender nonconforming, and nonbinary people—and research continue to be underrepresented. In the past 50 years, only seven articles on LGBTQ politics have been published in the APSR, with most published in recent years (Ayoub, Page, and Whitt Reference Ayoub, Page and Whitt2021; Cook Reference Cook1999; Girard Reference Girard2021; Kalla and Broockman Reference Kalla and Broockman2020; Lax and Phillips Reference Lax and Phillips2009; Magni and Reynolds Reference Magni and Reynolds2018; Reynolds Reference Reynolds2013). The publication records of other top journals in the field are similarly lacking.
Smaller identity groups almost inevitably struggle to find equal representation in the academy, in terms of both the number of people conducting research and the focus of the research. Outgroup members must study historically marginalized and stigmatized groups—particularly those that constitute small percentages of the population—to draw attention to the crucial political issues that affect those communities. In a discipline historically dominated by cisgender heterosexual white men, considering work in political science and/or using RCTs conducted only by ingroup members as valid perpetuates the deficiencies in research on smaller subgroups. LGBTQ studies cannot rely on openly LGBTQ scholars to drive the research forward.
Me-Search
Conducting research on and about an identity group with which we identify can be fraught with complicated dynamics. Critics argue that because we are researching ourselves—known colloquially as me-search—it is self-interested in a way that invalidates the study. We have encountered this phenomenon firsthand at conferences, at job-market events, and during book talks. Conversely, conducting research on and about an identity group with which we do not identify has the potential to introduce claims of appropriation and a lack of valid understanding.
Our argument is that researchers should be a mix of scholars who are ingroup and outgroup members because variations in positionality provide different insights and allow for the generation of different strains of knowledge. Although being open about our own positionality can be important, no one should be considered intrinsically more or less able to study a particular identity group based on our own identity. The strongest approach for the discipline and our communities is for both ingroup and outgroup members to study identity politics to contribute jointly to the literature.
…researchers should be a mix of people who are ingroup members and outgroup members because variations in positionality provide different insights and allow for the generation of different strains of knowledge.
Members of a historically marginalized group should be seen as valid and informed academics because they conduct high-quality, rigorous, and transparent research. A scholar’s preparation, methods, and transparency should be most important. RCTs are an important tool in this research; when used thoughtfully, they provide unique insights into political phenomena of theoretical and practical interest (Druckman and Green Reference Druckman and Green2021, 12). If members of a historically marginalized group conduct low-quality research, they should not be viewed as valid and informed academics simply because of their lived experiences. Conversely, outgroup members who conduct rigorous and thoughtful research should not be summarily dismissed simply because they do not belong to the member of a group that they are studying. Although researcher identity is certainly relevant, it is not determinative of quality or impartiality. Emerging norms among scholars using RCTs appropriately focus instead on the robust attention to ethical issues and the increased use of preregistration and pre-analysis plans (Druckman and Green Reference Druckman and Green2021).
Research with and on Outgroups
Outgroup members have an obligation when they are studying a different identity group: they must confirm the validity and benefit of the research itself. The work should benefit (or at least not harm) the marginalized group and should not exploit, misunderstand, misconstrue, or devalue the marginalized group. A simple way to accomplish this task is to consult with a member of the community being studied—perhaps a fellow scholar or a member of a local advocacy group—about the research plan, its feasibility, and its validity to that identity group. Often there are norms and expectations of a group that may not be known by outgroup members that easily can be identified in this simple conversation. Another way is to partner with other researchers who have relevant expertise or with political, interest, or advocacy groups studying similar topics. For example, in 2011, we conducted an experiment in cooperation with Equality Illinois, which was seeking to increase support for marriage equality among Black Americans. Using an RCT informed by colleagues with expertise in Black politics, we were able to avoid messaging that would be received poorly and to help the organization learn the importance of both message wording and recruiting Black volunteers in persuasion efforts (Harrison and Michelson Reference Harrison and Michelson2017). For our book on transgender politics, we often discussed our research ideas and progress with experts in transgender politics and members of transgender-serving organizations (Michelson and Harrison Reference Michelson and Harrison2020).
While being open about our own positionality as a researcher can be crucial, no one should be considered intrinsically more or less able to study an identity group based on our own identity. There should be no litmus test or obligation to be open about our own identity unless it potentially sacrifices the objectivity and validity of the study. Disclosures often are intrusive and potentially stigmatizing and may not augment a researcher’s credibility. However, depending on the research design, it may be appropriate or even advantageous for researchers to mention their identity as part of the discussion. Cassell et al. (Reference Cassell, Buehring, Symon, Johnson and Bishop2005, 5) found that identity has a role in credibility during data collection, particularly in qualitative research: “judgments of credibility were seen to depend on aspects of the nature and conduct of the research itself (e.g., methodical, conclusive, technically skilled, etc.) but also as influenced by symbolism and context.”
In our personal experience, the same can be true of quantitative research. The decision to disclose an identity is up to the individual researcher. However, much thoughtful, deep work is required to better understand the needs and wants of marginalized communities by people who are and are not members of those communities.
Our Positionality as Scholars of LGBTQ Politics
Our research includes three books: the first focuses on LGBTQ rights in general and on same-sex marriage in particular (Harrison and Michelson Reference Harrison and Michelson2017); the second focuses more exclusively on transgender rights (Michelson and Harrison Reference Michelson and Harrison2020). Our third book focuses more broadly on the entire LGBTQ community (Michelson and Harrison Reference Michelson and Harrison2022). As the focus of our research shifted over time, so also did our positionality. Brian is openly gay, making him an ingroup member for the first book project. As a cisgender man, he was an outgroup member for the second book project and then an ingroup member for the third. Melissa does not identify as a member of the LGBTQ community and was an outgroup member in all three book projects. We have experienced challenges to our expertise and positionality from multiple perspectives.
When we were promoting our first book at various public events, audience members—including political science faculty and students as well as the general public—asked multiple times about our positionality. As an ingroup member, Brian was able to come out to audiences to deflect concerns about our ability to appropriately understand the community that we were studying. We also emphasized that our research was informed by and often conducted in cooperation with members of LGBTQ advocacy organizations. The latter aspect of how we grounded ourselves in the community, we believe, is a more appropriate means of ensuring that the work we do is culturally competent. To assume that, as an openly gay man, Brian understands the entirety of the political experiences of all LGBTQ people—or even of all gay men—is to essentialize his identity and neglect the diversity of those communities. One gay man cannot possibly understand all of the different experiences and perspectives of this larger group. Instead, we spoke to representatives of multiple LGBTQ organizations about the needs and concerns of their members. Although Brian’s sexual-orientation identity may have opened some doors, we both were able to listen and to learn from those conversations. For our book about transgender rights, we were both positioned as outgroup members (Michelson and Harrison Reference Michelson and Harrison2020). In the few virtual presentations that we were able to make, we faced questions about our cisgender identities and whether it was valid for us to conduct research on transgender politics. Again, we emphasized that our work was informed by conversations with members of transgender-serving advocacy organizations and colleagues with expertise in transgender politics.
On the one hand, we understand why people asked these questions: they wanted to understand our positionality so they could consider it when evaluating our research. On the other hand, we were concerned that some people thought our identities as cisgender people meant that we were inappropriately studying members of an outgroup. Identity disclosure has the potential to cause real harm to a researcher, in terms of both safety and employment. However, if the risk of harm is low for a researcher and there is a conflict of interest between the identity and the research—or, conversely, there is a promise of a deeper applicability and connection of the research based on that identity—there are reasons to believe that disclosure might be a positive development.
Scholar Activists
Many political scientists, including ourselves, were drawn to the field to conduct research to challenge racism, inequality, sexism, homophobia, transphobia, xenophobia, and apathy. We are focused not only on contributing to theory and climbing the retention and promotion ladder; we also are driven by an interest in making a positive difference in the real world and a commitment to advancing justice, equality, equity, and inclusion. Scholars may disagree about what that positive change entails, but “[t]here is certainly no ethical requirement that engaged political science scholarship must be value-free” (Pepinsky Reference Pepinsky2018, 569). Different scholars can and do pursue activist scholarship that reflects their values and policy preferences.
A substantial body of recent political science scholarship, including our own, aims to increase support for LGBTQ rights domestically and internationally (Ayoub Reference Ayoub2016; Broockman and Kalla Reference Broockman and Kalla2016; Burack Reference Burack2018; Flores et al. Reference Flores, Haider-Marke, Lewis, Miller, Tadlock and Taylor2018; Harrison and Michelson Reference Harrison and Michelson2017; Lewis et al. Reference Lewis, Flores, Haider-Markel, Miller, Tadlock and Taylor2017; Lewis et al. Reference Lewis, Tadlock, Flores, Haider-Markel, Miller and Taylor2021; Magni and Reynolds Reference Magni and Reynolds2021; Michelson and Harrison Reference Michelson and Harrison2020; Tadlock et al. Reference Tadlock, Flores, Haider-Markel, Lewis, Miller and Taylor2017; Taylor, Lewis, and Haider-Markel Reference Taylor, Lewis and Haider-Markel2018). Much of this work uses RCTs to test theories about what is most likely to reduce homophobic and transphobic attitudes and behaviors. Because RCTs often involve large-scale interventions, they can be more explicit about altering the world as we know it than other types of research. The ultimate aim of this work is not to merely explore theories but rather to change public policies and to make the efforts of LGBTQ organizations and advocacy groups more effective at increasing public support for their rights. When we deliberately act as scholar activists, choices that we make concerning what to write about—regardless of methodology—can have real-world effects. The work can set the agenda and bring certain experiences to the forefront; it can shape political debates in living rooms, legislatures, and courtrooms. Whereas some political science scholarship does not try to generate change, other research does, including, we hope, our own.
CONCLUSION
Scholars often engage in research that will enact fundamental and needed change: virologists want to kill a virus, biologists want to cure disease, ethnomusicologists have affection for a culture and want to bring attention to its music. Our RCTs call attention to prejudice against LGBTQ people and help to develop best practices for how to reduce it. All researchers bring a motivation or bias to their work; however, that bias sometimes is seen as harmful or less than ideal. Rather than cast all such research as illegitimate, concrete steps—including researcher positionality—can begin to erode these outdated views. We should not have to omit information about our identities because we are concerned about whether it undermines credibility; conversely, we should not be forced to disclose our identities, particularly if they are sensitive or fraught. Although unlikely, a system of forced disclosure might incentivize people to lie about their identity (e.g., Rachel Dolezal and Jessica Krug). Demanding the expression of research identities while delegitimizing the expertise of those who study their own groups marginalizes the study of those groups. At the same time, the assumption that only ingroup members have valid perspectives can encourage that type of lack of disclosure.
Demanding the expression of research identities while delegitimizing the expertise of those who study their own groups marginalizes the study of those groups.
Members of a historically marginalized group should be seen as valid and informed academics when conducting research relevant to their own group because they have the lived experience of that group. However, nongroup members also have important perspectives based on their lived experiences, such as the surprise (and perhaps discomfort) they felt when learning that someone they knew identified as LGBTQ. Insider and outsider perspectives are both vital and valid approaches to scholarship relating to LGBTQ identity.
