Fertility

The research we describe in this section is on the topic of racial and class differences in women's fertility and other family formation patterns. As in most scholarship on inequality, studies of fertility usually examine the effects and causes associated with gender, class, and race/ethnicity as distinct from one another. Each dimension is studied as an independent or interactive effect and evaluated relative to others in determining how relevant it is in shaping individual experiences. Extensive analyses of this kind have produced a consensus that class differences in fertility are more substantial than racial differences, and that racial differences narrow significantly or are eliminated once socioeconomic status is controlled. Thus, many scholars in this field emphasize the importance of economic motivations in constraining women's fertility decisions, even though they are well attuned to differences across racial and ethnic groups. After briefly describing research in this area in greater detail, we argue that a more fully intersectional approach could significantly augment and alter its conclusions.

Recent empirical research on women's fertility patterns has led to the development of theories of women's family formation practices that are rooted in rational economic behavior. This contrasts with what many previously had thought of as irrational behavior. For instance, contrary to some frames associated with a culture of poverty thesis,Footnote ⁴ alternative explanations of women who choose to have children earlier in life indicate that they may do so because their educational and labor market opportunities are limited and their aspirations thwarted (Geronimus Reference Geronimus1986; Kaplan Reference Kaplan1997). Conversely, professional women postpone and thereby limit childbearing, often to undesirable levels, to devote their early employment years to careers (Blossfeld and Huinink, Reference Blossfeld and Huinink1991; Hewlett Reference Hewlett2002). Taking this research on poor and privileged women together, it is widely agreed that early childbearing, while rationally motivated and personally fulfilling in the short term, compromises the economic viability of mothers and their families in the long term (Hoffman et al., Reference Hoffman, Foster and Furstenberg1993).

Averil Clarke's (Reference Clarke2011) intersectional research on fertility began in this intellectual context, in which rational choice understandings of women's marital and childbearing decisions prevailed and where class-based explanations were central. Clarke's study relies on similar data and analytical methods as other studies of fertility, but it differs in the explanatory frame that guides the research design, analyses, and conclusions. Clarke's frame allows for the possibility that the dynamics of racial and class inequality are not independent of one another but interact in a variety of ways. After a brief description of key aspects of Clarke's research, we explain how it can refine the conclusions of previous research that has broadly similar findings yet is less sensitive to the simultaneous operation of multiple social dynamics that are often hidden in the details.

Clarke's project uses qualitative data from college-educated Black women and survey data on all groups from a nationally representative sample. It 1) explores connections between elite Black women's educational, occupational, romantic, and reproductive decision-making and their family formation outcomes, and 2) compares their processes of family formation to those of less educated Black women and degreed White and Hispanic women. Clarke identifies the same class differences in fertility levels as those found in the studies described above. But she argues that although college-educated women's fertility is lower than that of non-college-educated women, the road taken to arrive at this low fertility destination varies substantially by race and ethnicity. These alternative paths reflect different social explanations of college-educated women's relatively lower fertility.

Different combinations of marriage and childbearing are conceptualized in Clarke's research as routes to women's different fertility outcomes, within education groups as well as between them. She finds, for instance, that Black women at all education levels spend less than 25% of their reproductive years in marriage while White women spend about 40%. She also finds that less than 70% of college-educated Black women's births occur within marriage, while the corresponding figures for whites and Hispanics are over 95% and 85%, respectively. Thus the opportunity for having children within marriage, the preference of most women, is lower for college-educated Black women than for similarly educated White and Hispanic women, and it is more akin to that of less educated Black women.

Clarke uncovers the same limited opportunities in sexual behavior. Black women have the lowest average ages of sexual initiation at every education level, and college-educated women have the highest average ages of sexual initiation regardless of race (Warren et al., Reference Warren, Santelli, Everett, Kann, Collins, Cassell, Morris and Kolbe1998). Putting these two patterns together, Clarke shows that college-educated Black women have the longest average distance between virginity loss and the first family formation event of all race-education groups under study.Footnote ⁵ Clarke also finds that college-educated Black women have the highest likelihoods of experiencing periods of celibacy. Finally, she finds that among college-educated women, Black women terminate a higher fraction of pregnancies than White and Hispanic women mainly because they are more likely to be single. Single women are more likely to terminate pregnancies than married women, but among single women, Black women are actually less likely to do so.Footnote ⁶ The implication is that with greater marital experience (or less singleness), Black women would use abortion less and thus have higher fertility levels.

Clarke rounds up all of these findings to conclude that when it comes to the achievement of low fertility, a race-based deprivation in romance differentiates the experiences of Black women with college degrees from similarly educated White and Hispanic women. The advantages of class in desired family formation practices are thus distinctly racialized. We believe this conclusion, buttressed by a detailed analysis of group differences, augments and modifies the conclusions of studies that elevate the role of class-based explanations. Several such studies—large and small, qualitative and quantitative, with racially diverse and single racial or ethnic group samples—have led to the development of more nuanced understandings of the ways in which class and fertility are associated; nevertheless, racial or joint racial and class dynamics are often obscured. When multiple inequalities are studied and the relevance of each determined by its statistical significance in models that are not fully interactive, the risk is that researchers will make note of, but ultimately bypass, the dynamics of individual systems of inequality that appear secondary in the results. We describe three examples along these lines, from new research on fertility, that illustrate both the promise and pitfalls for uncovering intersectional dynamics.

One example is Kathryn Edin and Maria Kefalas' (Reference Edin and Kefalas2005) qualitative study of marriage avoidance among low-income mothers. Consistent with past literature, these authors explain young women's decisions to have children early in life without marrying as the consequence of economic motivations. At the same time, they notice but underplay some of the most novel findings in their study. These have to do with unexpected racial differences in attitudes about marriage: low income African American mothers are more likely to desire marriage than their counterparts in other racial/ethnic groups and to have ongoing relationships with the fathers of their children. The critical but unanswered question is why Black women are nevertheless less likely to marry, despite expressing greater interest in marriage and having more fertile conditions for marriage?Footnote ⁷ Using an intersectional frame, the study could have gone the extra mile to explore whether this racial difference was rooted in class-based processes or, more likely, some combination of racial and class dynamics.

We believe that some of the most recent research in family demography is moving more promisingly, but still too tentatively, in the direction of a genuinely multidimensional understanding of family formation patterns among both the poor and the privileged. These studies document racial differences but without necessarily illuminating their connection to class differences. One such study (like Clarke's) describes fertility differences across education separately for Blacks and Whites. It found that while education reduces fertility for both Blacks and Whites, the effect is much greater for Black women: they have fewer children than White women among the college educated but more among the non-college educated (Musick et al., Reference Musick, England, Edgington and Kangas2009). These differences, the authors argue, “reflect the large racial differences in union status across the education distribution; Black women are much less likely to be married, and married women are much more likely to intend their births” (Musick et al., Reference Musick, England, Edgington and Kangas2009, p. 557). But because the study's foremost objective is to examine educational differences, substantial racial differences within education groups are noted but not explained. They are not linked to the kinds of racial processes that inhibit desired childbearing among college-educated women described by Clarke.

New research using propensity scores offers what may be a more promising approach to answering this question, though it too has limitations. This technique consists of ranking individuals according to their propensity to attend college, for example, based on factors such as family income, parents' education, race, and academic achievement. It categorizes individuals into groups along a continuum from those with multiple disadvantages and a low propensity to complete college to those with multiple advantages and a high propensity. Sorting the population like this incorporates ideas of multiple jeopardy or interlocking systems of oppression, in which groups are irreducible to a single dimension of subordination (Collins Reference Collins1990; King Reference King1988).

Within these categories of likely and unlikely college-goers, we can observe the fertility and marriage outcomes of those who actually did and did not attend college. Jennie Brand and Dwight Davis (Reference Brand and Davis2011) and Brand and her colleagues (Musick et al., Reference Musick, Brand and Davis2012) find that the top strata of likely college-goers had more children and were more likely to marry if they went to college than if they did not, whereas the reverse was true for the lowest strata of unlikely college-goers, whose fertility and marriage prospects decline if they attend college, as Clarke found among Black college-educated women. Importantly, this interaction effect means that the main effect of college, without looking at how its effect differs for advantaged and disadvantaged groups, is statistically insignificant. Moreover, the interaction that stands out the most is the extremely low fertility rate of college-educated women from disadvantaged backgrounds. As the authors rightly note, this should not be considered a “positive” effect of a college education.

These scholars are going beyond conventional explanations of delayed childbearing and marriage rooted in economic incentives. However, the new methods deploy a multidimensional framework that obscures the dynamics of specific intersections. Emphasized instead is the generic role of deficits in social and cultural capital:

While in broad strokes this is probably true, it is unclear what explains almost diametrically opposed experiences among the educated. Clarke's study delves more deeply into the racial dynamics underlying these “class” differences and provides compelling detail of their centrality in explaining restrictions in romantic opportunity, marriage, and fertility among Black college-educated women, many of whom come from multiply disadvantaged backgrounds and encounter racial restrictions in relationship markets. This alternative rendering is only in part a function of the space afforded in a book length study exploring detailed interactions of this kind. The rest must be attributed to the explanatory framework of intersectionality that Clarke employed from the start of her research to understand outcomes for Black college-educated women that challenge rather than ratify theories of economic rationality.

Homogamy

Next we describe the development of two lines of inquiry that emerged gradually out of our desire to reconcile conflicting normative interpretations of martial homogamy (i.e., the marriage of individuals with similar characteristics) with the perspectives of gender and class/income inequality.Footnote ⁸ Taken in the form of a thought experiment, our first line of inquiry considers the conflicting implications of homogamy for achieving greater gender equality on the one hand (which it does), and less income inequality on the other (which it does not). This leads to an intersectional view that attempts to reconcile these conflicting views by developing a pattern of partnership that we call “gender neutral heterogamy” (marriage between unequals) that could lead simultaneously to greater income and gender equality. The second example then examines the class and gender processes that obstruct the achievement of heterogamy and reinforce homogamy.

The point of departure in the first example is two opposing perspectives on marital homogamy. The first perspective is grounded in the desire to achieve greater gender equality within households and the second from the perspective of desiring greater equality of income across households. With respect to the former, there is an enormous body of feminist research on the exploitative aspects (more typically for women) and economic benefits (more typically for men) of economic dependence between romantic partners (Sorensen and McLanahan, Reference Sorensen and McLanahan1987). Research has repeatedly shown that women's disproportionate responsibility for unpaid reproductive labor within households constrains their labor market opportunities, weakens their bargaining power with male partners, and makes them more dissatisfied with familial relationships (England and Farkas, Reference England and Farkas1986; Rogers and DeBoer, Reference Rogers and DeBoer2001). Consequently, much research focuses on the degree to which women and men in heterosexual partnerships increasingly contribute equally to household income, thus fostering more equitable gender relations (e.g., Raley et al., Reference Raley, Mattingly and Bianchi2006). Although rarely put in exactly these terms, the consensus is that increasing “income homogamy” is good for gender equality.Footnote ⁹

But increasing economic equality within households also has implications for income inequality across households. Rising income homogamy means that the rich are increasingly marrying the rich and the poor marrying the poor, leading to a greater concentration of income at the top and poverty at the bottom (Schwartz Reference Schwartz2010). This is one of the factors that many believe contributed to the spectacular growth of income inequality in the United States over the past several decades (though we recognize that there are deeper causes, such as housing and other forms of segregation that have increased alongside income inequality). Focusing on this one factor, then, from the perspective of income inequality, one can argue that it was actually more equalizing in the past to have the wives of highly paid men out of the labor force and the wives of working-class men in the labor force, than it is today when wives are more likely to be employed, regardless of their husband's income (Cancian and Reed, Reference Cancian and Reed1999; Juhn and Murphy, Reference Juhn and Murphy1997). In short, something that appears to be a prerequisite of gender equality—income homogamy in heterosexual relationships—leads directly to increases in another form of inequality, income inequality. Yet neither of the literatures on gender or income inequality addresses this conflict because the literatures are unidimensional. The unintended consequences of choosing one site to advance greater equality, while excluding others, are therefore obscured.

If one is interested in advancing gender equality and income equality simultaneously, how does one go about resolving the contradictory implications of homogamy for achieving greater equality? We raise an option that involves rethinking the typical definition of gender equality as parity of earnings between partners within each household (McCall Reference McCall, Lareau and Conley2008). The arrangement we propose is to maximize dependency within couples in a gender-neutral way, so that men are as likely as women to be economically dependent within relationships. This minimizes income homogamy and hence income inequality among families. Note that couples could continue to be homogamous in terms of education but would eschew income homogamy to the extent possible by relying on a primary breadwinner. What matters in this scenario is the achievement of a gender-neutral distribution of dependency in the population as a whole rather than an equal distribution of earnings between members of each household.

Here we demonstrate, then, that putting different kinds of social dynamics into conversation with one another may challenge the normative interpretations derived from any single social dynamic understood in isolation. This process uncovers alternative normative interpretations of the same fact (i.e., income homogamy) and lays the groundwork for a more inclusive normative perspective. That this normative solution is more feasible among the affluent, who can better afford to forego the earnings of one partner, than it is among most other families, does not necessarily negate the utility of this intersectional thought experiment. Rather, it points further down the road to the economic (e.g., living wages) and social (e.g., subsidized care) conditions that would have to be in place for its realization.

In the second example, we begin to open the black box of marital homogamy to ask how it comes about in practice in order to assess the prospects for fostering greater “gender-neutral heterogamy.” We focus on the lives of Black college-educated women, although this time we have no data for other groups against which to compare our findings since this is a research area in its infancy. The data are qualitative and explore whether and how women come to increase commitment to romantic partners who share and/or exceed their socioeconomic status as well as to reject partners of lower status. The conclusions we draw once again bridge the arguments from two separate literatures on economic and gender theories of marital decision making.

The first literature uses data on the traits of those already married to test economic theories of marital decision making (Bennett et al., Reference Bennett, Bloom and Craig1989; Lichter et al., Reference Lichter, LeClere and McLaughlin1991; Oppenheimer et al., Reference Oppenheimer, Kalmijn and Lim1997). It adjudicates among theories claiming that individuals try to marry those with the highest income possible or, alternatively, with equivalent traits so that they can produce children like themselves. Others emphasize that individuals are constrained by their own and potential partners' economic viability (Oppenheimer et al., Reference Oppenheimer, Kalmijn and Lim1997; Wilson and Neckerman, Reference Wilson and Neckerman1987). The second literature focuses on explanations having to do with changes in gender relations and heterosexual dating practices and norms (Bailey Reference Bailey1988; Bogle Reference Bogle2008; Coontz Reference Coontz2005). While neither of these literatures makes specific claims about how women's dating behavior relates to their eventual marital choices, or directly observes the mate selection process, both provide insight into the constraints and opportunities college-educated Black women face and the decisions they make in the achievement of homogamy.

Using open-ended survey data on premarital romantic relationships from college-educated Black women, we found that factors having to do with men's relative economic status did arrest the development of romantic relationships. But the majority of the qualitative data did not evince this reasoning in the straightforward manner suggested by economic theory. For example, a woman might reject a partner of higher status if the partner believed that his superior status entitled him to special treatment that she was unwilling to accommodate. Or a woman might report that her partner of a lower status initiated separation or resisted marriage because he did not feel he was “contributing enough.” Still other women were criticized by parents or friends who interpreted their partners' artistic or entrepreneurial goals as “irresponsible,” reflecting a willingness to take advantage of the women's higher economic status once married. In a few cases, pregnancy and/or childbearing led to a breakup when partners disagreed about whether and how educational or economic activities should adapt in response to parenting (e.g., should one or the other leave school/employment or give up trying to be an artist?). These data make fairly consistent reference to the fulfilled or disappointed expectations of women, their partners, and their friends and family members. Because these expectations tied privilege and responsibility in romantic relationships to partners' economic and non-economic (e.g., age, gender, pregnancy, or family status) characteristics, the literature on gender and its relation to dating practices and norms became a critical resource in explaining how heterogamous relationships were thwarted.

The literature connecting changes in women's premarital romantic behavior to an evolving gender order suggests a more cultural explanation for college-educated Black women's constraints and choices in premarital romance. Making sense of women's evaluations of their partners' socioeconomic fitness and of the appropriateness of particular occupational pursuits for male and female partners became easier after considering the possibility that individuals made decisions about socioeconomic fitness while struggling with various ideologies of heterosexual romantic partnership. Individuals used ideologies of both the past (e.g., mother's primary responsibilities are in the home, men who marry assume responsibility for breadwinning) and present (e.g., both partners should have careers established before marrying or childbearing) in order to evaluate the long-term prospects of a relationship. Market-based choices are made, but they are not made outside of the context of present and historical ideological frameworks that define the still gendered nature of marriage and romantic partnerships. Thus neither the economic nor the gender-based explanations of homogamous relationships can account on their own for how such relationships are often formed in practice.

In considering both of our examples, we recognize that selecting a partner is a deeply intimate and individual activity, and therefore one that is going to be resistant to formulaic models. However, partner selection is also a social practice with complex and significant ramifications for social inequality. Thus, scholars working in the fields of economics, demography, and gender, race, and sexuality studies have all become interested of late in the issue of homogamy. Often, however, scholars approach the subject from different and even opposing perspectives. In the above examples, attempts to place competing perspectives in conversation with each other produced intersectional accounts that identified new connections and conflicts among multiple dynamics (e.g., rising income inequality across households and falling gender inequality within households); and constructed new research questions (e.g., how does income homogamy actually happen in practice?), social explanations (e.g., the joint gender and economic determinants of partnership formation), and normative solutions that extend beyond the confines of unidimensional accounts (e.g., gender-neutral heterogamy).

Liberalism

Feminist and critical race theorists of all stripes have long decried the universalist pretensions of liberal and democratic political theory in societies with deep racial and gender inequalities. Over the past few decades, however, several authors have brought these strands together in new theoretical frameworks incorporating multiple forms of social inequality as central divisions in liberal and democratic societies. Although two of the three authors we discuss emphasize racial over gender divisions and do not claim the intersectionality label, they each build on a foundation of feminist and antiracist scholarship. In these examples, the mix of feminist and antiracist criticism shapes the development of a new interpretation of familiar histories that retains valuable elements of prior interpretations. Thus a new synthesis was enabled after the accumulation of separate areas of inquiry over time. We emphasize the temporal evolution of intersectional frameworks in this section, rather than the validity per se of each author's arguments.Footnote ¹⁰

We look first at Rogers Smith's “multiple traditions” framework. In an article dubbed “an instant classic among scholars of American political thought and history” (Hochschild Reference Hochschild1998, p. 321) and a subsequent book, Smith (Reference Smith1993, Reference Smith1997) exposed the contradiction between ideals of liberalism, democracy, and republicanism on the one hand, and the exclusion of women and minorities from full citizenship, what he terms “ascriptive hierarchy,” on the other. Smith's framework likely held wide appeal because it did not position one tradition as secondary to another, as much mainstream scholarship had previously done; nor did it treat one tradition as logically dependent on or irredeemably tarnished by the other, as more critical scholarship had done. Rather, much like earlier strains of feminist scholarship in the dual systems tradition, Smith (Reference Smith1993) saw each dynamic as relatively autonomous with “a complex pattern of apparently inconsistent combinations, accompanied by recurring conflicts” (p. 558).

Smith suggested a careful reinterpretation of historical and social contexts, in which particular configurations of race and liberalism, for example, crystalized at some points and other configurations at different points. The challenge for subsequent scholars was to define the parameters of those relationships at various junctures of history and to formulate a more nuanced and critical story of liberalism's shifting boundaries and racist legacy. Such a story would partially supersede the “multiple traditions” framework by comingling liberalism and ascriptivism to a greater degree than did Smith.Footnote ¹¹ Still, one could say that Smith succeeded in bridging intellectual movements that were hitherto severed from one another, galvanizing dialogue across scholarly boundaries defined by critical studies of race, gender, class, and American liberalism. Indeed one of the few scholars cited in the original APSR article as informing and embodying the spirit of a multiple traditions framework was Kimberlé Crenshaw and other writers associated with intersectionality and Critical Race Studies.

The reweighting of issues of race and gender that occurred in the wake of Smith's multiple traditions intervention in political science has its counterpart in political philosophy in the scholarship of Charles Mills's The Racial Contract (Reference Mills1997) and, more importantly for our purposes, Contract and Domination (Reference Pateman and Mills2007), coauthored with feminist political theorist Carole Pateman. Mills and Pateman analyze social contract theory, which is the basis upon which equal individuals emerge from a state of nature to form social contracts resulting in a just liberal society. These authors intervened in debates over the different types of contract implied by theorists such as John Locke, Jean-Jacques Rousseau, Immanuel Kant, and John Rawls, calling attention to what they saw as a more fundamental obfuscation: social inequalities among subjects.

As above, here we note the evolution of theorizing from more or less isolated feminist and anti-racist camps (i.e., the Sexual Contract by Pateman and the subsequent Racial Contract by Mills) to their more flexible and encompassing notion of the “domination contract” that can both bridge and decipher among multiple axes of inequality. In an article published a year after the book with Pateman, Mills (Reference Mills2008) discusses the domination contract and distinguishes “ideal” from “nonideal” theory, not only as a matter of what normatively ought to occur versus what occurs in existing society, but as a matter of what “justice demands in a perfectly just society [versus …] what justice demands in a society with a history of injustice” (p. 1384). For Mills, an ideal theory of a just social contract in a hypothetical society of equals is all but irrelevant in an unjust society. Ultimately a raceless and genderless liberalism may materialize, but only if the racial and gendered dimensions of liberal theory and society are first made transparent, through an explicit confrontation with the “facts” of racial and gender subordination. Here again, few would dispute these facts; the challenge is to reinterpret and utilize them to construct new theories of a fully just society.

Evelyn Nakano Glenn (Reference Glenn2002) undertakes this challenge in her explicitly intersectional analysis. (Neither Smith nor Mills and Pateman identify their work as based in themes of intersectionality, yet we suggest that each fosters opportunities to engage in intersectional thinking.) Glenn's subject is the nineteenth-century development of U.S. citizenship and labor. She maintains from the outset that her approach is intersectional because it examines the simultaneous and interconnected evolution of racial and gender inequality in politics (citizenship) and the economy (labor). Like the two prior examples, Glenn begins with the central contradiction between the theoretical presupposition of American liberalism and the realities of race and gender exclusion. But she also makes her task one of showing how the liberal tradition developed in practice through the “efforts of dominant groups seeking to enforce their own definitions of citizenship and its boundaries, and by the efforts of subordinated groups to contest these definitions and boundaries” (Reference Glenn2002, p. 55). The dialectical unfolding of this contest between elites and subordinates, on the ground, rather than the “discursive conflicts among competing members of the elite” (p. 52), is what she says distinguishes her study from Smith's.

This definition of Glenn's subject matter places everyday struggles over the meaning of belonging and freedom among social groups in deeply unequal relations to one another at the center of analysis. Glenn demonstrates this in three historical case studies, which articulate the way that race and gender inequalities are presupposed in White men's definition of and access to citizenship rights. In addition to discussing the more well-known history of conflicts over citizenship rights between whites and African Americans, her work exposes the regional specificity of this pattern to the southeast, and how it deviates from constructions of racial boundaries between Anglos and Mexicans in the southwest, and Haolis, Japanese-Americans, and native Hawaiians in Hawaii. Thus Glenn reinterprets the same facts three times over: she takes what we have tended to see as a single broad construction of White male citizenship and transforms it into three particular versions of the making and remaking of a racialized and gendered republic for all citizens, White and non-White alike.

Debates over the meaning of American liberalism are notorious and enduring. What we have endeavored to show is the significant theoretical progress made in defining how multiple systems of inequality are connected to the canonical traditions of American history and ideology. While covering similar territory on one level, however, there are real conceptual differences among the approaches outlined here: autonomous but equally important traditions versus fully imbricated ones; generically named but encompassing concepts such as “ascriptive hierarchy” and the “dominance contract” versus regionally and historically situated configurations of citizenship; and discursive debates among elites and philosophers versus the politics of everyday struggle and social change. Many will consider the second side of each opposition as the proper side for a study of intersectionality; and if forced to take sides, we might not disagree. But our point is that there are advances being made on the first side of these oppositions as well, ones that help to refine the practice of bridging and borrowing across the borders of subfields and improving our explanations of how multiple social dynamics interact with one another.