1
Feminist Theory and Survey Research
“The idea that there is only ‘one road’ to the feminist revolution, and only one type of ‘truly feminist’ research, is as limiting and as offensive as male-biased accounts of research that have gone before.” ~Liz Stanley and Sue Wise, 1983, p. 26.

Introduction
Over the past three decades, feminist methodologists have hammered home one point with surprising regularity: Feminist research takes a variety of legitimate forms; there is no “distinctive feminist method of research” (Harding, 1987; see also Chafetz, 2004a, 2004b; Fonow & Cook, 2005; Hawkesworth, 2006; Hesse-Biber, 2007; Risman, Sprague, & Howard, 1993; and Sprague, 2005). And yet, to this day, the relationship between feminist theory and quantitative social science research remains uneasy. Among feminist scholars, quantitative research is often seen as suspect for its association with positivism and its pretense of objectivity (among other things). At the same time, among quantitative researchers, feminist-identified work is often dismissed as “biased,” “activist,” or “substantively marginal.” While a number of scholars have recently published works outlining a “feminist” approach to social science research, these books have generally steered clear of quantitative survey research. Some authors of feminist
1

2— Feminist Measures in Survey Research

methods texts limit their discussion of feminist survey research to a small section (e.g., Hesse-Biber, 2007; Reinharz, 1992; Sprague, 2005), while others overlook survey research entirely (e.g., Hesse-Biber, Gilmartin, & Lydenberg, 1999; Jaggar, 2008; Naples, 2003). Sociologist Joey Sprague (2005) aptly describes the situation:
Because feminists and other critical researchers have tended to assume that quantitative methodology cannot respond to their concerns, there are relatively few analyses of specific procedures that are problematic in mainstream quantitative methodology and there is even less written on feminist ways of implementing experiments or surveys. (pp. 81–82)

In this book, I hope to offer a new approach for viewing (and doing!) quantitative feminist research. Rather than asking, Can quantitative research really be feminist? (as many other feminist methodologists have already asked and answered), I ask, What do quantitative researchers risk by continuing to ignore feminist theories? My answer, which I happily reveal up front, is, A lot! Though a feminist approach will certainly add more to some branches of quantitative research than to others, a feminist perspective can inform virtually every aspect of the research process, from survey design to statistical modeling, to the theoretical frameworks used to interpret results. Throughout the book, I hope to show how feminist theory can measurably and significantly improve a wide range of quantitative social science research. In addition, I want to suggest that the relationship between quantitative research and feminist theory is especially fruitful when an interdisciplinary, multiracial feminist approach is used. Those who are relatively unfamiliar with both feminist theory and quantitative research and those who have already discovered for themselves the usefulness of integrating feminist theory and quantitative methods may see the aforementioned goals as relatively straightforward: I hope to show how a multiracial feminist approach can improve quantitative social science research in a variety of areas. Readers with a background in the humanities, feminist philosophies of science, postmodern feminist theories, or queer theories, however, are likely to see these goals as something else: complex, perhaps even misguided or naive. As psychologist Carolyn Wood Sherif (1979/1987, p. 51) wrote some thirty years ago, “If the issues of [gender] bias in psychological research were as simple as turning the methods and instruments prized by psychology into the service of defeating bias, many battles would have been won long ago.” Readers who approach this book with a background in social science are, perhaps, just as likely to view my aims as suspect. Science infused explicitly with ideology and activist agendas is

Chapter 1. Feminist Theory and Survey Research——3

no longer science, one might argue. As Janet Saltzman Chafetz (2004a), sociologist and self-described feminist, asserted, although there is such a thing as feminist theory—even if I do not think of it as social scientific—I find the very idea of feminist methodology in the social and behavioral sciences fundamentally untenable. . . . The research design and tools of data collection and analysis one selects ought to be chosen on the basis that they are the most appropriate to answering a given research question (pp. 971–972)

—not on the basis of political or ideological commitments. My goal in this book, then, is to address both of these concerns headon. I argue that feminist theory and survey research can be used together. In fact, much existing research already points to the advantages of feminist social science research. At the same time, however, elements of Sherif’s and Chafetz’s comments ring true. A feminist approach to social science research does require something other than redeploying the same old instruments and methods (recall feminist theorist Audre Lorde’s similar assertion that “the master’s tools will never dismantle the master’s house” [1984, pp. 110–113]). And while I certainly disagree with Chafetz on the tenability of feminist methodology, I wholeheartedly agree with her second point. The research design and tools of data collection and analysis should be chosen on the basis that they are the most appropriate to answering a given research question. As I hope to show in this book, however, a multiracial feminist approach is oftentimes the most appropriate choice. In the remainder of this chapter, I provide a brief introduction to feminist theory and research. I begin with the question, What makes research “feminist”? and then examine the historically uneasy relationship between quantitative social science research and feminist research. After exploring how other scholars have navigated this relationship, I then focus on one particular branch of feminist theory—multiracial feminism—which has been largely ignored in quantitative social science research. In considering the relationship between multiracial feminist theory and quantitative social science research, I introduce three themes that together form the backbone of this book. First, multiracial feminist theories offer numerous substantive insights into the social world that have been underused by social science researchers. Second, multiracial feminist theorizing offers survey researchers a number of analytic interventions that can bring increased complexity and nuance to their research. And third, by highlighting difference, inequality, relationality, and the context of discovery, a multiracial feminist perspective can help survey researchers increase the quality of social science research. I describe the tenets of multiracial feminism and conclude this chapter with an overview of what is to come.

4— Feminist Measures in Survey Research

What is Feminist Research?
More than 30 years ago, historian Gerda Lerner called for a feminist transformation in the field of history. She wrote,
[H]istory as traditionally recorded and interpreted by historians has been, in fact, the history of the activities of men ordered by male values—one might properly call it “men’s History.” Women have barely figured in it; the few who were noticed at all were members of families or relatives of important men and, very occasionally and exceptionally, women who performed roles generally reserved for men. In the face of such monumental neglect, the effort to reconstruct a female past has been called “Women’s History.” The term must be understood not as being descriptive of a past reality, but as both a conceptual model and a strategy by which to focus on and isolate that which traditional history has obscured. (1979, pp. 168–169, italics added for emphasis)

In her approach to women’s history, Lerner called for something more than simply “finding women” in the historical record. She argued that the development of women’s history would require “challenging traditional sources,” challenging the “traditional periodization of history,” redefining categories and values, in short, a complete paradigm shift. “Women’s history,” she writes, “demands a fundamental re-evaluation of the assumptions and methodology of traditional history and traditional thought” (1979, p. 180). Over the course of the next decade, similar arguments were made throughout the humanities and social sciences. Sociologists Judith Stacey and Barrie Thorne, for example, called for a feminist revolution that would “transform the basic conceptual frameworks” of sociology (1985, p. 301). In addition to “correcting sexist biases” in research, and “creating new topics” that reflect women’s experiences, the feminist revolution they called for would produce a “‘gendered’ understanding of all aspects of human culture and relationships” and would “as equally attend to race, class, and sexuality as to gender” (1985, p. 311). In the same year, psychologist Michelle Fine published an article assessing the development of feminist psychology. She concluded that while some advances had been made, future feminist scholarship would be strengthened by situating individuals within social and historical contexts, increasing cross-disciplinary collaboration among researchers, and by “documenting the diversity of women’s experiences” (1985, p. 179). For purposes of this book, what is most interesting about these accounts is that these scholars see a feminist approach to scholarship as something more than “research about women.” For Lerner, Fine, Stacey and Thorne, and others, a feminist perspective challenges some of the most taken-for-granted conceptual and methodological assumptions in a given field.

Chapter 1. Feminist Theory and Survey Research——5

Feminist research requires a different approach to scholarship, but what does that approach entail? The answer depends on whom one asks. For example, feminist philosopher Sandra Harding (1987, p. 1) begins her classic book, Feminism and Methodology, by explicitly rejecting the idea that there is a “distinctive feminist method of research.” She brings a historical approach to feminist social science, asking, “What are the characteristics that distinguish the most illuminating examples of feminist research?” (p. 6). She identifies three. First, feminist research “generates its problematics from the perspective of women’s experiences.” Second, it is scholarship done for women—it seeks to provide women with explanations that they “want and need.” And third, feminist research emphasizes the “importance of studying ourselves and ‘studying up,’ instead of ‘studying down.’” In other words, feminist inquiry “locates the researcher in the same critical plane as the overt subject matter” (p. 8). In their book, Beyond Methodology, Mary Margaret Fonow and Judith A. Cook (1991, p. 2) identify four themes in feminist research, “reflexivity; an action orientation; attention to the affective components of the research; and use of the situation-at-hand.” Feminist approaches to social research, they explain, are “often characterized by an emphasis on creativity, spontaneity, and improvisation in the selection of both topic and method” (1991, p. 11). In her Handbook of Feminist Research, sociologist Sharlene Hesse-Biber (2007, pp. 16–17) describes feminists’ research in a similar way: It asks new questions, “going beyond correcting gender bias in dominant research studies”; centralizes issues of power, authority, ethics, and reflexivity in the practice of research; and is typically conducted at the margins of traditional disciplines. Common to each of these approaches is the idea that feminist research involves something more than adding women and stirring or simply controlling for gender by means of a single variable. Feminist research requires a shift in how we approach research, but it does not require a focus on women, per se. Feminist research requires a feminist perspective, but feminist research might not focus primarily on gender. And certainly, feminist research needn’t be produced by women. As Harding writes, “obviously, neither the ability nor the willingness to contribute to feminist understanding are sex-linked traits!” (1987, p. 11). Feminist research, then, is not necessarily distinguished by the topic of research, nor by the sex, gender, or political affiliation of the researchers involved. Rather, feminist research is distinguished by how the research is done and, to some extent, by what is done with the research. What theoretical perspective(s) does the researcher bring to the research, and how does this inform her or his approach to doing the actual research—formulating questions, planning research design, interpreting results, disseminating

6— Feminist Measures in Survey Research

information? Because there are multiple approaches to feminism and multiple varieties of feminism, there are also multiple approaches to feminist research.i

Feminist Theory and Survey Research
Despite feminist methodologists’ broad understanding of what constitutes feminist research, many contemporary scholars—both feminists and otherwise—continue to see quantitative survey research as being at odds with feminist theory. This is true for a number of reasons. Within the social sciences, feminist scholars have rightly critiqued survey research for reducing gender to “sex”—a dichotomous variableii that obscures the relationship between gender (which social scientists and feminist scholars typically consider to be socially based) and sex (which is typically understood as something more physiological).iii Too often in survey research, gender (which becomes synonymous with sex) then appears to be a stable property of individuals (“She is female.”)— rather than a “principle of social organization” (Stacey & Thorne, 1985, p. 307). Focusing on gender at the level of the individual, we lose sight of the processes through which gender is socially constructed and maintained. We also risk losing sight of how gender operates as a social institution— how gender “establishes the patterns of expectations for individuals” and “orders the social processes of everyday life” (Lorber, 1994, p. 1). A similar reductive process occurs with measures of race and ethnicity. As sociologist Tukufu Zuberi (in Zuberi & Bonilla-Silva, 2008, Introduction, p. 6) writes, “when we discuss the ‘effect of race,’ [in statistical models] we are less mindful of the larger social world in which the path to success or failure is influenced.” Analyzing racial difference and inequality by means of a single dichotomous variable, we risk losing sight of the institutional dimensions of racial inequality (Pager & Shepherd, 2008; Zuberi, 2001). We also risk losing sight of the dynamic social processes that create racial groups and
Naples (2003, pp. 3–4) makes a similar claim: “Since there are diverse feminist perspectives, it follows that there are different ways feminist researchers identify, analyze, and report ‘data.’” ii Dichotomous, or “dummy,” variables are those that have two and only two options, such as “male” and “female,” or, in attitudinal research, “agree” or “disagree.” In statistical analyses, these are typically coded 0 and 1, though what numbers are assigned to what category makes no difference, so long as they are interpreted properly. iii Increasingly, feminist scholars see sex itself as socially constructed. See, for example, Butler (1990), Kessler (1998), and Fausto-Sterling (2000). i Chapter 1. Feminist Theory and Survey Research——7

maintain differences among them. In other words, we risk essentializing race and racial differences (Zuberi & Bonilla-Silva, 2008). By understanding gender and race as social institutions, rather than as stable properties of individual people, we can see how culturally and historically specific ideas about gender are “built into the [other] major social organizations of society, such, as the economy, ideology, the family, and politics” (Lorber, 1994, p. 1). We can also see how gender is connected to the other major systems of inequality such, as race, nation, sexuality, and class. Feminist scholars in the humanities have been largely critical of quantitative survey research, although their criticisms are often different from those described above. While feminist social scientists have critiqued quantitative survey research primarily on the basis of method (i.e., “gender as a variable”), critiques from the humanities focus more broadly on methodology and epistemology (the study of knowledge—what can be known? And who can know?). In this critique, the very foundations of social science are called into question. As Harding (1987, p. 182) points out, “scientific knowledgeseeking is supposed to be value-neutral, objective, dispassionate, disinterested, and so forth. It is supposed to be protected from political interests, goals, and desires (such as feminist ones) by the norms of science.” And yet feminist research, by definition, has interests and values, for example, social justice and human rights. While Harding herself argues that these differences are not irreconcilable, others strongly disagree (e.g., Chafetz, 2004a). In addition to these critiques, feminist standpoint theorists from both the humanities and social sciences have argued that quantitative research is limited in so far as it rarely takes into consideration the social and historical contexts in which it is produced. Quantitative research is often presented as value-free, objective, and disinterested. Rarely do quantitative scholars cast a critical eye on the processes through which research is produced and how the research production process may reflect (and even reproduce) social inequalities. Feminist standpoint theorists, in contrast, have argued that knowledge about the social world is often structured by social inequalities (e.g., Hartsock, 1983/2003; Hill Collins, 1990/2000; Smith, 1974; Sprague, 2005; Valadez, 2001). Individuals who share particular social statuses or social locations often share meaningful experiences, which in turn can generate shared knowledge about the social world. If, however, in our scientific research, the voices and experiences of privileged groups are consistently represented but those of underprivileged groups are marginalized or excluded entirely, then the resulting knowledge claims are necessarily limited. Standpoint theorists emphasize the value in understanding all knowledge claims (whether they be made by privileged or underprivileged groups) as partial perspectives. For standpoint theorists, social science research is

8— Feminist Measures in Survey Research

never value-free, objective, or disinterested—it is always produced from a particular perspective and within a particular context—nor should it aspire to be so. Rather, standpoint theorists embrace the idea that knowledge is socially situated and seek to produce and value knowledge grounded in subordinate, social positions.iv Although feminist theory offers a number of important critiques of quantitative research techniques (and social science more generally), this does not necessarily mean that the two are fundamentally irreconcilable. In fact, quantitative research has been an important tool for understanding, documenting, and challenging gender inequalities and social inequalities more generally. Consider, for example, how quantitative research has helped to document feminist gains—and lingering inequalities—in higher education. Survey research shows us that in 1970, women represented 40% of college students enrolled in degree-granting institutions in the United States. By 2007, this percentage had increased to more than half (55%).v Fifty-seven percent of bachelor’s degrees conferred in the 2000–2001 school year were awarded to women, up from 43% in 1969–1970.vi But despite these gains, women still earn only 20% of the bachelor’s degrees awarded in the field of engineering and less than a third (28%) of the degrees conferred in computer and information sciences.vii
See also Hill Collins (2000), Haraway (1988, 1990), Hartsock (1983). Source: U.S. Department of Education, National Center for Education Statistics, Higher Education General Information Survey (HEGIS), Fall Enrollment in Colleges and Universities surveys, 1970 and 1980; 1990 through 2006 Integrated Postsecondary Education Data System, Fall Enrollment Survey (IPEDS-EF:90–99), and Spring 2001 through Spring 2007; and Projections of Education Statistics to 2017. U.S. Department of Commerce, Census Bureau, Current Population Survey (CPS), October, selected years, 1970 through 2007. (This table was prepared August 2008.) Retrieved March 3, 2010, from http://nces.ed.gov/programs/digest/ d08/tables/dt08_190.asp vi Source: U.S. Department of Education, National Center for Education Statistics, Higher Education General Information Survey (HEGIS), “Degrees and Other Formal Awards Conferred Survey;” and Integrated Postsecondary Education Data System, “Completions Survey” (IPEDS-C:01), 2000–01. Retrieved March 3, 2010, from http://nces.ed.gov/pubs2005/equity/figures .asp?PopUp=true&FigureNumber=K vii Source: U.S. Department of Education, National Center for Education Statistics, Higher Education General Information Survey (HEGIS), Degrees and Other Formal Awards Conferred Survey; and Integrated Postsecondary Education Data System, Completions Survey (IPEDS-C:01), 2000–01. Retrieved March 3, 2010, from http://nces.ed.gov/pubs2005/equity/figures .asp?PopUp=true&FigureNumber=K v iv

Chapter 1. Feminist Theory and Survey Research——9

Quantitative analyses of the U.S. Census Bureau’s Current Population Survey (CPS) have similarly helped feminists keep track of gendered economic inequalities. In March of 1964, the CPS revealed that the weekly wages of full-time, year-round women workers, aged 25 to 64, were 58% of what full-time, year-round men workers of the same age group earned.viii More than 4 decades later, the U.S. Census Bureau reported women’s earnings had improved relative to men’s, but, they noted, a significant wage gap remains. In 2008, women in the United States who worked full-time, year-round earned only 77% of what full-time, year-round men workers earned.ix Further, a recent report from the U.S. Bureau of Labor Statistics shows that the gender gap in earnings remains at every level of educational attainment: In the fourth quarter of 2009, the median weekly earning for men who were working full time but who had earned less than a highschool education was $686.x For women, the corresponding figure was $477, roughly 70% that of men’s earnings. For men and women with bachelor’s degree or higher, the weekly earnings for full-time workers were $1,896 and $1,384, respectively—a gender gap of 73%. In addition to documenting material inequalities, quantitative survey research has been a valuable tool for documenting cultural beliefs about gender and how these beliefs have changed over time. Over the past several decades, the U.S.-based General Social Survey (GSS), for example, has regularly asked respondents whether they believed that “most men are better suited emotionally for politics than are most women.” In 1974, nearly half of men agreed with this statement (47.6%). In 2010, 23.3% of men surveyed agreed. Attitudes have clearly changed since the 1970s, but with nearly one in four men still clinging to the belief that women are ill-suited for politics (and notably, nearly 1 in 5 women are also clinging to this belief!), women politicians and those aspiring to become politicians still face a tremendous obstacle. In another example, in 2010, the GSS also asked respondents about their views about balancing work and family. Strikingly, one out of three women surveyed (33.7%) and more than a third of men (39.1%) indicated that they agreed or strongly agreed with the statement that “it is much better for everyone involved if the man is

viii

Explaining Trends in the Gender Wage Gap. June 1998. A Report by The Council of Economic Advisers. Retrieved February 21, 2010, from http://clinton4.nara .gov/textonly/WH/EOP/CEA/html/gendergap.html#2 ix Retrieved February 20, 2010, from http://www.census.gov/Press-Release/www/ releases/archives/income_wealth/014227.html x This report was based on data collected in the fourth quarter of 2009. Source: Table 4. Retrieved from http://www.bls.gov/cps/earnings.htm#education

10— Feminist Measures in Survey Research

the achiever outside the home and the woman takes care of the home and family.” Though stories in popular culture tell of women’s advances leaving men in the dust, survey research presents an alternative, sobering view: Gender inequality persists. Survey research has clearly played an important role in the fight for gender equality in education, work, and families. But quantitative analyses of survey research have been important tools for understanding other manifestations of sexism as well—including those beyond the realm of what we might consider “liberal” articulations of feminism. For example, quantitative survey research has been important for understanding and challenging a culture of violence against women. In particular, survey research has helped reframe debates about sexual assault so that stranger rape no longer occupies the forefront in discussions of violence against women. Survey research has shown us that women in the United States are more likely to be killed in their homes than in any other setting. The National Coalition Against Domestic Violence (NCADV) reports that “almost one-third of female homicide victims that are reported in police records are killed by an intimate partner.”xi In addition, survey research has shown us that the vast majority (85%–90%) of victims of sexual assault on American college campuses know their assailants—sadly, about half of such incidents occur during a date.xii Survey research and quantitative data analysis more generally have also helped document changes in gender ideology and, relatedly, in cultural representations of men and women. Despite the much discussed “death of feminism” in the 1980s and 1990s, survey data analyzed by Bolzendahl and Myers (2004) and Huddy, Neely, and LaFay (2000) document increased support for feminist ideals over the past several decades.xiii Analyzing research from dozens of surveys across more than three decades, Huddy et al. (2000, pp. 316–317) conclude that support for the U.S. women’s movement “shows no sign of diminishing in the 1990s” and that “[y]oung people remain staunch movement supporters” (see also Harnois, 2008; Peltola, Milkie, & Presser, 2004).

The NCADV fact sheet cites: Federal Bureau of Investigation, Uniform Crime Reports, Crime in the United States, 2000. (2001). NCADV fact sheet. Retrieved March 3, 2010, from http://www.ncadv.org/ xii Retrieved March 3, 2010, from http://www.ojp.usdoj.gov/nij/topics/crime/rapesexual-violence/campus/know-attacker.htm xiii Bolzendahl and Myers (2004, p. 760) conclude that feminist attitudes among women and men, “have continued to liberalize . . . with the exception of abortion attitudes, which have remained stable.”

xi

Chapter 1. Feminist Theory and Survey Research——11

In brief, quantitative analyses of survey research have played an important role in helping to understand and challenge systems of inequality in many of its varied forms. As sociologist Christine Williams writes, quantitative analysis is necessary if feminists are to intervene in important political debates. . . . Sometimes we need numbers to present a compelling argument, to inspire activism, and to get things changed. . . . [W]e cannot and should not give up on the quantitative study of gender. (2006, p. 456)

Risman, Sprague, and Howard (1993, p. 608) sum it up nicely: “Some feminist questions demand quantitative answers.” While feminist critiques of quantitative research are numerous, feminist scholars have offered important critiques of (almost?) every kind of research in the social sciences as well as in the humanities and biological sciences. For example, feminist scholars have critiqued ethnography, participant observation, oral history, content analysis, literary criticism, experimental research, and medical trials, in addition to quantitative survey research. But rather than abandoning these approaches, many feminist scholars have sought to improve these techniques—and in many cases, to use them to different ends. For example, ethnographic research may, at one time, have been a tool of imperialism, but many anthropologists and sociologists today use ethnographic research to subvert neocolonialism and other systems of inequality, working with disempowered groups around the world to help achieve their goals (e.g., Booth, 2004; Hewamanne, 2008; Naples & Desai, 2002). Radical methodological critiques—whether they be feminist, postmodern, antiracist, and/or postcolonial—have not always advocated throwing the proverbial baby out with the bathwater; rather, they have often worked to transform and reappropriate these techniques. They have used these transformed techniques in combination with other approaches and have drawn post-positivistic conclusions about the social world.xiv As Risman (2001, p. 610) writes, “feminist [social science] scholarship expresses a commitment to science with and from a value position. This is a rejection of the belief in the possibility of value-free singular context-less scientific ‘Truth,’ but it is neither a rejection of all science nor an acceptance of relativism.” From a post-positivist perspective, feminist scholars seek to identify the “cultural elements” that shape scientific inquiry, and to “figure out which of these cultural elements are at this particular historical moment advancing and which blocking the growth of knowledge” (Harding 1998, p. 145; see also Risman, 1993; Sprague & Zimmerman, 1989, 1993). xiv Positivism emphasizes the promise and possibility of objectivity in science. As Sprague (2005, p. 32) writes, “positivism assumes that truth comes from eliminating the role of subjective judgments and interpretations.”

12— Feminist Measures in Survey Research

My approach to feminist survey research builds on the feminist transformations and reappropriations described above. While many varieties of feminism can (and have) informed survey research, in this book, I highlight the implications of multiracial feminist theory for social science survey research. Though it is seldom employed in conjunction with survey research, multiracial feminist theory offers the grounds for the transformation, reappropriation, and post-positivist interpretation of survey research.

What is Multiracial Feminist Theory?
In their article, “Theorizing Difference From Multiracial Feminism,” Maxine Baca Zinn and Bonnie Thornton Dill (1996) describe multiracial feminism as a broad-based theoretical perspective in which race, gender, class, sexuality, and nation are understood as intersecting systems of inequality. This idea of systems of inequality as intersecting with one another—intersectionality—is meant to suggest something beyond additive models of oppression that came before.xv While multiracial feminists acknowledge that many individuals are simultaneously disadvantaged by multiple systems of inequality (for example, racial minority women may face racism and sexism), they argue that additive models of inequality are insufficient for understanding the complexity of the social world. By examining systems of inequality as separate and distinct systems, additive models fail to address ways in which systems of inequality work “with and through each other” and influence the lives of all people, privileged and underprivileged alike (Baca Zinn & Thornton Dill, 1996, p. 326). In contrast to those feminists who seek to understand gender in isolation from other systems of inequality, multiracial feminists explicitly locate the

xv

As discussed in the preface, I use the language of “multiracial feminist theory” because it draws attention to the importance of race and feminism in the intellectual genealogy of contemporary “intersectional” scholarship. In addition, the term multiracial feminist theory draws attention to the importance of theorizing difference, as opposed to simply highlighting or “discovering” difference. As intersectionality becomes more mainstream, there is considerable risk of its becoming a “buzzword” (Davis, 2008) and in the process, risk of losing both its theoretical complexity and radical potential. The phrase “multiracial feminist theory” reminds us that differences must be theorized and that historically the most important intellectual work in this area has been done by women of color, that is, by multiracial feminists.

Chapter 1. Feminist Theory and Survey Research——13

social construction of gender (and other systems of stratification) within a broader context of intersecting social hierarchies. These intersections take place at the level of the individual, where “people experience race, class, gender, and sexuality differently depending upon their social location in the structures of race, class, gender, and sexuality” (Baca Zinn & Thornton Dill, 1996, pp. 326–327). They also intersect at the institutional level, where, for example, systems of race, gender, class, and sexuality reinforce one another and are each built into our political, economic, and cultural institutions.xvi Multiracial feminist theorists acknowledge that, in particular situations, any given social status or system of inequality may be more or less salient (e.g., Jordan, 1982/2003). But they refuse to designate one system of inequality as universally more significant than others, as the intersections of systems of inequality are both dynamic as well as “organized through diverse local realities” (Hill Collins, 1990/2000, p. 228; but see also Combahee River Collective, 1981; Jordan, 1982/2003; Weber, 2001). Multiracial feminists’ refusal to privilege universally one system of inequality over others has resulted historically in intersectional politics—political movements and global and community activism—that similarly refuses a single-oppression framework (e.g., Berger, 2004; Combahee River Collective, 1981; Roth, 2004). In addition to emphasizing the intersections of systems of inequality, multiracial feminists have emphasized the “relational nature of dominance and subordination” as well as women’s agency. As Baca Zinn and Thornton Dill explain, “intersecting forms of domination produce both oppression and opportunity. At the same time that structures of race, class, and gender create disadvantages for women of color, they provide unacknowledged benefits for those who are at the top of these hierarchies—whites, members of the upper classes, and males” (Baca Zinn & Thornton Dill, 1996, p. 327; see also Baca Zinn, Hondagneu-Sotelo, & Messner, 2000/2007; Barkley Brown, 1992). Multiracial feminist theory focuses not only on difference and particularity but also on the relationships, inequalities, and social processes that help create and maintain these differences. This focus on the relationships that structure difference and inequality stands in stark contrast to “patchwork quilt” (Baca Zinn et al., 2000/2007) and “mosaic” models (May, 2010) of difference. In these latter approaches, difference and particularity are highlighted, but the social structures in which these differences are embedded remain unexplored.

xvi

See Weber (2001) for an excellent discussion of how power relations intersect and are expressed simultaneously at the macro and micro levels.

14— Feminist Measures in Survey Research

In emphasizing relationality, multiracial feminism highlights the process through which differences are created and maintained. For Baca Zinn and Thornton Dill, the final “distinguishing features” of multiracial feminism concern issues of methodology and accountability. They explain, “multiracial feminism encompasses wide-ranging methodological approaches, and like other branches of feminist thought, relies on varied theoretical tools as well” (1996, p. 328).xvii These theoretical and methodological approaches come from across the humanities and social sciences. However, since many of the central works of multiracial feminism were neither produced by “traditional” academics nor produced in traditional academic spaces, multiracial feminism underscores the need to engage with intellectual work outside of academia as well. Historically, structured inequalities of race, class, gender, and nation have worked to limit educational and career opportunities for women of color. These intersecting inequalities have limited women’s ability to acquire prestigious positions within academia, and have also limited their ability to produce and publish scholarship in those outlets with the most academic legitimacy. As a result, many women of color intellectuals turned to nontraditional spaces to create and disseminate their work (and oftentimes, as in the case of Kitchen Table Press, they created these spaces in the process). As with feminist standpoint theorists, multiracial feminist theorists have drawn attention to how social locations help to shape knowledge; they have argued that “lived experiences . . . create alternative ways of understanding the social world and the experience of different groups . . . within it” (Baca Zinn & Thornton Dill, 1996, p. 328; see also Hill Collins, 1990/2000). Multiracial feminist scholarship embraces these alternative understandings, which have been marginalized within traditional scholarship and in some situations ignored completely. Finally, in emphasizing the inequalities built into the knowledge production process, multiracial feminism raises the issue of accountability. In her essay “La Güera,” Chicana feminist Cherríe Moraga writes, “so often the [white] women seem to feel no loss, no lack, no absence when women of color are not involved; therefore, there is little desire to change the situation” (1981/1983, p. 33). In her speech “The Master’s Tools Will Never Dismantle the Master’s House,” Audre Lorde (1979/1984) makes a similar point. She calls attention to the dearth of “women of Color” represented

xvii

See also Hancock (2007), who argues that an intersectional approach must be both empirical and theoretical and must draw from multiple methods.

Chapter 1. Feminist Theory and Survey Research——15

at academic feminist conferences, and she challenges her audience to think critically about that situation:
Why weren’t other women of Color found to participate in this conference? Why were two phone calls to me considered a consultation? Am I the only possible source of names of Black feminists? In academic feminist circles, the answer to these questions is often, “We did not know who to ask.” But that is the same evasion of responsibility, the same cop-out, that keeps Black women’s art out of women’s exhibitions, Black women’s work out of feminist publications except for the occasional “Special Third World Women’s Issue,” and Black women’s texts off your reading list. But as Adrienne Rich pointed out . . . white feminists have educated themselves about such an enormous amount over the past ten years, how come you haven’t also educated yourselves about Black women and the differences between us—white and Black—when it is key to our survival as a movement? (1984, p. 113)

In her essay “Age, Race, Class, and Sex,” Lorde writes of a similar phenomenon at work within the classroom:
The literature of women of Color is seldom included in women’s literature courses and almost never in other literature courses, nor in women’s studies as a whole. All too often, the excuse given is that the literatures of women of Color can only be taught by Colored women, or that they are too difficult to understand. . . . I have heard this argument presented by white women of otherwise quite clear intelligence, women who seem to have no trouble at all teaching and reviewing work that comes out of the vastly different experiences of Shakespeare, Molière, Dostoyefsky, and Aristophanes. Surely there must be some other explanation. (1984, p. 117)

Lorde and Moraga highlight the need for privileged groups to educate themselves about issues of difference and inequality and about groups who are different from themselves. They push women to think critically about what has become routine, normative, and taken for granted. They push women to take responsibility for the role they play in maintaining inequality and to hold themselves accountable to something beyond what is expected.

Multiracial Feminist Theory and Survey Research
The question of accountability—not, To whom are we accountable? but rather, To whom do we choose to hold ourselves accountable?—is an

16— Feminist Measures in Survey Research

important one for feminist research, as well as teaching and activism.xviii In “traditional” social science research, scholars are typically held accountable to discipline-specific expectations: What questions are the most central or important? Which theorists are important and thus worth reading and citing? What methods are typically used? How long is a typical article? Is it appropriate to use the first person? and so forth. Multiracial feminist theories push scholars to think critically about these norms. The goal is not to denigrate research grounded in traditional academic disciplines but rather to understand how disciplinary norms structure the knowledge production process and the resulting knowledge claims. Only then can we develop alternative research strategies that bring into focus what previous research has obscured. As we will see in greater detail in the next chapter, disciplinary norms structure the social science research process, rendering some questions, theories, methods, and interpretations more legitimate and others less so. In so doing, these norms can perpetuate inequalities already built into the system. As sociologist Barrie Thorne (2006, p. 477) explains, disciplines “discipline, in the positive sense of providing training, honing methodological skills, and sustaining communities of practice. But, they also enforce conventions, sustain hierarchies and mechanisms of exclusion, and police boundaries (as in the cursing phrase ‘that’s not sociology!’).” If convention dictates that class inequality is more “central” than race or gender, as has historically been the case in the field of sociology, then critical race and feminist theories are relegated to the margins and with them the experiences of racial minorities and women. If convention dictates that samples composed mostly of middle-class, white American college students are sufficient for making general claims about the social world, then the experiences of lower- and working-class, racial minority, and international students—not to mention older, nonstudent populations—will similarly remain hidden. In

xviii

In her recent article “Be longing: Toward a feminist politics of relation,” Aimee Carrillo Rowe (2005), asks feminists to reconsider the politics of location, as a way of thinking. She asks, “What gets left out when the conditions and effects of belonging to a ‘location’ are assumed as a starting point for our theorizing?” (2005, p. 15). She urges the reader to reframe the question, “To whom are we accountable?”—a question that takes the link between social location and accountability as a given (e.g., I am a woman, and so I am accountable to women; I am a sociologist, and so I am accountable to sociologists). Instead, she suggests, we might ask, “To whom do we choose to be accountable?” Accountability need not stem solely from our particular social locations. We can, she suggests, choose to hold ourselves accountable to a broader political or intellectual community, and doing so often involves building relationships across diverse communities.

Chapter 1. Feminist Theory and Survey Research——17

emphasizing the value of diverse methodological approaches and theoretical perspectives, multiracial feminism advocates expanding the intellectual and political communities to which we choose to hold ourselves accountable. Doing things as they are typically done is simply not enough. As Berger and Guidroz write of the “intersectional approach,” multiracial feminism emphasizes “border-crossing,” and “challenges traditional ways of framing research inquiries, questions, and methods” (2009, p. 7). What I hope to show here, though, is that multiracial feminist theory offers more than a challenge to—and more than a critique of—survey research. Multiracial feminism offers an alternate approach for doing survey research. It emphasizes difference and inequality, relationality, and the circumstances under which the research itself is produced—as Harding (1987, p. 183) calls it, the “context of discovery.” As I show in the following chapters, multiracial feminism offers the grounds for transformation and reappropriation, interdisciplinary border crossing, and post-positivist interpretations of survey research.

Organization of the Book
In what follows, I hope to show the promise of a multiracial feminist approach to survey research. Three themes in particular stand out. First, multiracial feminist scholarship offers substantive insights into the social world that have been underused by survey researchers. Second, multiracial feminist theorizing offers survey researchers a number of analytic interventions that can bring greater complexity and nuance to social science research. And third, by highlighting difference, inequality, relationality, and the context of discovery, a multiracial feminist perspective can help survey researchers increase the quality of social science research. The next chapter provides an overview of contemporary feminist survey research from across the social sciences and within women’s and gender studies. I analyze a sample of more than 50 quantitative articles published in five feminist journals in the past two decades and investigate the extent of disciplinary boundaries in scholars’ theoretical perspectives, as well as in their survey tools and their analytic techniques. While interdisciplinarity is a key theme in multiracial feminist theory and feminist theory more generally, feminist survey research remains largely structured by disciplinary boundaries. And while scholars in each discipline have engaged with some aspects of multiracial feminist theory, the majority of quantitative survey research does not. I conclude by considering how disciplinary boundaries might work to constrain the development of a multiracial feminist approach, and return to this idea in each of the following chapters.

18— Feminist Measures in Survey Research

What would it mean, in practical terms, to bring a multiracial feminist, or intersectional framework to survey research? In Chapters 3 to 5, I demonstrate several approaches for bringing a multiracial feminist framework to social science survey research. I focus on issues of meaning, measurement and modeling, and seek to show how multiracial feminist theorizing can inform each aspect of survey research. Chapter 3 focuses on sexism, Chapter 4 on racism, and Chapter 5 on feminism. I begin each chapter by discussing the contributions of prior survey research in these areas. My goal is not to denigrate prior research but rather to highlight its importance for social justice, social change, and social theory. I then guide the reader through some of the limitations of this research. My focus is on the hidden assumptions of survey instruments and multivariate models. For example, what kinds of questions do we ask when we want to gauge women’s experiences with sexism? Do our measures make sense for women of different ages, racial groups, and socioeconomic classes? Imagine that we were to design a new survey, focusing just on young women’s experiences. What survey questions would be most relevant? Would any of the measures be inappropriate? What additional questions would we ask? What literature would we consult to help us answer these questions? How are differences represented in our statistical models? And what assumptions of sameness are challenged when we begin from a multiracial feminist perspective? Finally, what analytic strategies are available for multiracial feminist analyses of survey research? Throughout these chapters, I explore the feminist theories that I have found most valuable for answering these questions. As we think through these and other questions, we begin to see how our measures are constructed with an eye toward the experiences of particular groups. Often, our measures work best for groups that are more privileged. Often, our measures and our models help obscure the experiences of those who are already marginalized. In the concluding chapter, I bring together the methodological findings from the previous chapters and outline six considerations for thinking about survey research from a multiracial feminist perspective. A multiracial feminist approach is useful not only for survey research on racism, sexism, and feminism but also for many kinds of survey research and for social science more generally.