Chapter 1
Data and Discipline: Discovering the Feminization of Migration
cholars who discussed the feminization of migration at the end of the twentieth century suggested that it was a recent development even as they presented findings pointing toward much earlier change. In this book, we present considerable evidence of earlier periods of feminization and offer an explanation for a shift toward gender balance that stretched across the entire twentieth century. Before we present that historical evidence, however, it seems important to answer the question implicit in Hania Zlotnik's acknowledgment that most feminization occurred before 1960. If migrations had begun to feminize already in the 1910s and 1920s (as we show they did), why did discussions of feminization begin only seventy to eighty years later? And why was the discussion not led by scholars in the new and rapidly expanding scholarly field of gender and women's studies?
Answering those questions becomes even more important when we realize that scholarly interest in the sex of migrants is not new. On the contrary foundational scholarship in the multidisciplinary field of migration studies addressed the topic beginning in the nineteenth century. Furthermore, women scholars interested in sex and gender have been studying migration for more than a century. Explaining why the feminization of migration was labeled only in the final years of the twentieth century thus reveals much about the origins, development, and practices of the scholarly field—interdisciplinary migration studies—to which this book contributes.
More than a literature review of the most recent research findings on sex, gender, and migration, this chapter (and the next) offers what German-speakers might call a Begriffsgeschichte1 or what English-speakers describe as an historical sociology of knowledge2 of the three analytical
20     Gender and International Migration
categories—gender, sex, and migration—that define the intersection of gender and migration studies. Begriffsgeschichte and the sociology of knowledge share a profound respect for the ways in which knowledge \ production—a term more commonly used in gender studies—intersects with and depends on social structures, relations, and processes of schol- 4 arly inquiry. In this chapter, we argue that migration studies scholars' i understanding of sex, gender, and migration always reflect the influence % of gender ideology and gender relations on scholarship itself.3 Changing S societal assumptions and prescriptions about sex difference (sometimes:i called gender ideology) structure gender relations society-wide, and in scholarly institutions, training, research, and professional employment. For a very long time, gender ideology and relations, not biological sex, closed off scholarly knowledge production to women, along with many. other activities, limiting them, to familial, domestic, and reproductive life. Gender relations and ideology also subsequently made knowledge production in some scholarly disciplines seem more appropriate for female or male scholars. As a result, disciplines and even methods could, be gendered male or female. Tracing how gender affects disciplines, data, and | research methods over time, this chapter suggests why few scholars in i;i migration studies looked for or recognized the feminization of migration as it began in the early twentieth century; and why it was demographers I and statisticians—rather than scholars in gender studies—who first labeled J feminization in the 1980s.
The State Sciences and the Study of Population, Sex, and Migration
Scholarly study of migration began with states' efforts to govern people | residing in their territories and to differentiate resident insiders from potentially ungovernable outsiders or foreigners.4 Almost all contempo- ;g rary scholars in gender studies trace such insights about the governmental-ity of humans by states to Michel Foucault's critique of the modem state.5 i But state efforts to describe and register people in order to govern them | are, in fact, very old. The so-called state sciences (Staatswissenschaften)—- :? the scholarship and disciplines mat produce knowledge useful to states"— and the development of demography and population studies of mobile J populations—are more recent developments.7 From its origins in the sev- | enteenth century, the discipline now called demography made biological f sex a fundamental category of analysis that gave states important information about the strength, decline, or increase of populations through repro-duction, birth and aging. The earliest scholarly studies described migrant % populations, too, as structured by sex difference. Almost all these early | studies of populations—whether sedentary or mobile—were the work of § male scholars and researchers. M
Data and Discipline 21
Both world historians and social scientists have recently described how a sedentary bias has profoundly affected scholarly knowledge of human lifes Since the invention of the written word, humans have differentiated sedentary from mobile people. The first agrarian states, which emerged ivith the spread of agriculture and the Neolithic revolutions in east, west, and south Asia, typically distinguished themselves from foragers, hunters, and nomadic herders—groups whose lives and cultures remained defined by constant patterned movement.9 The differentiation was not neutral:
"whether in China or the classical Mediterranean, imperial states associated civilization with sedentarism and mobility with their enemies, the "barbarians." Although the civilized (who alone left written documents for historians to study) had every reason to exaggerate barbarian mobility,10 world historians have instead pointed to the many forms of mobility that made civilization possible, including military mobilizations, imperial expansion, labor migrations to build Egypt's pyramids or China's Great Wall, and long-distance trade.11
Despite ancient cultural suspicions of mobile people, the scholarly
: study of migration began only with the development of the state sciences within emerging, expanding, and often imperially motivated nation-
: states. In tine seventeenth century, rulers of Europe's states first began to see the persons they governed as constituting a population. In England, Trancis Bacon suggested that rulers must understand and provide for their populations if they expected to govern them.12 Counting a population and evaluating its size and attributes soon became key empirical foundations for enlightened or scientific governance. Thus the first English population researcher, the hatter John Graunt, along with his better-educated collaborator, William Petty, called their recordings and enu-
. merations of births and deaths political arithmetic. Graunt was the first of many demographers to note that male babies outnumbered girl babies at birth, setting into motion population scientists' long-term interest in tracking age and sex in human populations. By the 1800s, population statisticians even called themselves statists.13 Across the nineteenth century, statists demonstrated the interaction of age and mortality through what they called population sex ratios. Because mortality differed by sex across the life course, male predominance among newborns gradually shifted
.... toward female majorities among the elderly. The early modern scientific revolution that Bacon pioneered had challenged theocratic explanations for state power by making sex and biological population reproduction, not God's grace, key measures of a state's strength.
!3y the eighteenth century, Thomas Malthus's (1766-1S34) more pessimistic predictions, including that a growing population inevitably outstripped the agricultural production and livestock required to feed it,14 deeply disturbed a British society in the midst of rapid industrialization, urbanization, and population growth and ultimately motivated greater
22     Gender and International Migration
attention by statists to human mobility.13 It is striking that the earliest mobility statistics in France and England focused less on foreigners' arrivals than on internal movements as mobile citizens and subjects relocated -from the countryside (where they raised food) to cities and industrial jobs (where they consumed food). Scholars today debate whether any transition from lower to higher rates of mobility accompanied these changes; the issue remains unresolved.16 The political elites at the time certainly believed both that mobility was increasing and that it was potentially threatening. Statisticians' long-standing interest in the relationship of sex, reproduction, and population growth assured that the earliest studies '" of mobility, too, paid some attention to sex and thus to male and female ™ migration patterns.17
The oft-cited work of British statistician and geographer E. G. Ravenstein . (1834—1913) and his laws of migration—which appeared in publications -of 1876, 1885, and 1SS9—illustrate how one founder of modern migra- jt, tion studies understood the relationship between sex and migration. Born into a family of German cartographers, Ravenstein migrated as a young "4
man to England and became a naturalized citizen.18 He developed his .....
ideas about sex and migration as he analyzed the 1871 and 1881 population censuses of Great Britain. In addition to listing individuals' current places of residence, sex, and age, census-takers had asked about both the ^ place of birth and sex of all natives of the British Isles (including colonial Ireland). They had distinguished natives from aliens but had not noted aliens' places of birth or nationality. On this basis, Ravenstein claimed to -!~ have discovered what he called laws of migration in 1885. In a second article, Ravenstein claimed universality for his laws by providing evidence , from other European countries—Germany, France, Austria, Hungary— and from Canada and the United States.1" .««;
In one of his most influential laws, Ravenstein argued that sex deter- "r mined the distance migrants journeyed, asserting that "Woman is a greater migrant than man." In a clear nod toward Victorian ideology of sep- ~t arate (public and domestic) spheres for male and female activity,20 he ; acknowledged that "This may surprise those who associate women with ' domestic life, but the figures of the census clearly prove it."21 However,
Ravenstein mainly documented that women were more migratory than ___
men within the kingdom (for example, England, Scotland, or Wales) of their birth. In his typology of local, short- and long-joumey stage migration and temporary (short-term) migration (a schema that did not distin- ■-■»-guish migrations within the British Isles from those with foreign origins), short-journey migrations were by far the most common, making women i the greater migrant. Men, by contrast, predominated among migrants 7 venturing beyond the kingdom of their birth, whether internally (witiiin Great Britain) or internationally (into Great Britain from foreign states). Ravenstein also noted and commented on an important exception to this ~$
Data and Discipline 23
law linking masculinity and long-distance migration: he found more women than men among migrants bom in Ireland but living in England,
Scotland, and Wales.
Ravenstein's conclusions about sex and migration were empirically ^rounded but also mirrored late-Victorian understandings of proper sex roles. In explaining migrations, for example, Ravenstein largely ignored familial or marital motivations for moving about. Thus, he did not consider tine possibility22 that marriage customs (especially women's moves to their husbands' village or farm) explained women's greater propensity to move short distances.23 When Ravenstein did note the influence of private or domestic concerns, furthermore, he linked them exclusively to female migrants and female morality. He suggested, for example, that counties with spa towns had attracted grass widows, who were women -whose husbands had deserted them, forcing them to work as prostitutes, or were away in the military, thus releasing them from household responsibilities and allowing them to enjoy indolent lives as hotel boarders.
Ravenstein—and much subsequent scholarship—focused on migra-tions motivated by labor demand in receiving or host societies. Ravenstein especially emphasized how sex-specific demand for waged workers in urban labor markets drove migration, such that towns with large military garrisons drew predominantly male migrants. He also described male migrants outnumbering female migrants in mining and metal-working towns, but acknowledged that women, too, sought work as migrants. He described women as migrating "quite as frequently into certain manufacturing districts" and concluded that "the workshop is a formidable rival of the kitchen and scullery" (in urban middle-class households) in demanding female labor.24 Presumably, the manufacturing districts described as attracting female migrants included England's many urban . centers of textile production.25 Thus, while Victorian gender ideology prescribed domestic roles for women, Ravenstein acknowledged that both sexes responded to labor market demand as they traveled to different destinations.
:: Despite his argument that "woman was tire greater migrant," Ravenstein also became the first in a long line of researchers who imagined men as the primary migration decision-makers. As Ravenstein sought to explain
...the. Irish anomaly of female majorities among long-distance movers, he also noted that "Whilst emigrants from England or Scotland depart in
.most instances without 'incumbrances' it appears to be a common practice Tor entire families to leave Ireland in search of new homes."26 Ravenstein
..wrote several decades after the end of the Irish potato famine—an event
: that might have provided him with an explanation for Irish distinctiveness and the international migration of more women than men. Instead, he ™agined male decision-makers traveling with or without family dependents (a group that presumably included both women and children),
Od
Gender and International Migration
but he never explained why male migrations encumbered with families would result in female majorities. Nor did he consider the possibility that more Irish women than men migrated because demand for female labor in textiles and domestic service may have surpassed demand for male labor in mining and industry.
Over time, the study of internal or domestic migrations and the study of international movements developed as separate scholarly fields of study. The male majorities that characterized long-distance movers became a persistent focus in research on international migration. Some critics have even claimed that the unencumbered or individual labor migrant who traveled in pursuit of economic advantage became the archetypical figure in migration research.27 This also made the long-distance male migrant most deserving of study; female migrants seemed interesting only when they outnumbered males or when they—like males—sought wage-earning : work as unencumbered persons without families. A century later, feminist; scholars continued to complain that scholarship treated female and chil--. dren migrants as uninteresting dependents of men.28
Historians have described how the ideas of Ravenstein and others cir- '■ culated widely, influencing governance and record-keeping around the Atlantic.251 In the Americas, however, it was international migration— what Americans increasingly called immigration30—and not urbanization that garnered the greatest attention from the founders of migration ; studies, hi America, where Christiane Harzig and Dirk Hoerder locate a second birthplace of modern migration studies, sex and migration remained objects of inquiry but were researched within a scholarly world that included more female scholars.31 Complex scholarly gender relations ■ help draw disciplinary boundaries around the production of knowledge .: about sex and migration in important ways.
Sex, Gender, and American Immigration Studies
Like Ravenstein, male American statists were interested in sex and rnigra- | tion. For example, they repeatedly demonstrated that years of massive : immigration had raised the proportion of men in the U.S. poptilation \: beyond normal levels.32 Over time, however, statists' conclusions about sex | and migration became marginalized in U.S. immigration studies, in part, :T this reflected declining levels of international migration after 1920 and a v. shift in the immigration research disciplines that bifurcated scholarship into | separate groups focused on internal and international moves. U.S. scholars ••' have long celebrated the Chicago School of Sociology and the immigrant- f origin Harvard historian most influenced by the Chicago sociologists' :ii work, Oscar Handlin, as founders of U.S. immigration studies.33 By contrast, scholars in statistics and demography and the state sciences—rather
Data and Discipline 25
than in history, anthropology, and sociology—produced the most research on sex, gender, and migration, whether internal or international.
Although some sociologists now question whether the early sociologists at the University of Chicago constituted an intellectually coherent school,34 few in the United States question the importance of Chicago sociologists such as Robert Park (1864-1944), VV.I. Thomas (1863-1947), Ernest Burgess (Canada, 1886-1966), and Louis Wirth (Germany, 1897-1952). These men were intensely interested in foreign-born immigrants, but more as social than as sexual or reproducing mobile people.35 Rather than study migrants' mobility, the Chicago sociologists analyzed immigrants' social, cultural, and socio-psychological adjustments to modern American society. The Chicago School was not disinterested in mobility : but treated it mainly as the initiator of the social and cultural changes that Chicago scholars found most interesting.36 The Chicago sociologists used an eclectic but decidedly qualitative mix of methods that emphasized careful, direct, ethnographic, or biographical observations of urban life. From these.micro-studies emerged ecological models of urban development as well as theories of immigrant disorganization, alienation, and marginality and of assimilation and a race relations cycle.37 Oscar Handlin later brought these themes to a broader audience with his influential history of immigrant alienation and assimilation.33
Historians have recently identified other intellectual fonts of U.S. immigration studies, all with significant input from women researchers, few of whom could obtain positions in American universities. Dirk Hoerder labels Jane Addams (1860-1935) and the Progressive social reformers who created the American settlement house movement as founders of a women's Chicago School.39 As reformers, political activists, and researchers, they documented and analyzed immigrant women's and men's lives at work, in neighborhoods, and at home. Mary Jo Deegan argues that departments of sociology excluded the Chicago women researchers as politically biased naive reformers or mere social workers.40 A multiracial, multiethnic, and interdisciplinary network of both male and female students trained at Columbia under immigrant anthropologist Franz Boas (Germany, 1858-1942), also produced new research on immigrants. Donna Gabaccia and Jon Gjerde identify sons and daughters of immigrants working at Midwestern state universities as the earliest historians of immigration.41 However, Kate Asaphine Everest Levi—who wrote the first history dissertation on immigration at the University of .Wisconsin—found work as director of a social settlement house, not in a university history department.42 Only the students of Boas found significant academic employment.
: Still invisible in these newer genealogies were the women researchers— many with interests in sex and migration—who became population scientists in the 1920s and 1930s. Even as specialized academic training,
26
Gender and International Migration
research and teaching in other fields became progressively more masculine professional arenas after 1920, women trained as statisticians found work and even exercised leadership inside universities and as state scientists.
This association between women researchers and quantitative methodologies and statistical population research was quite powerful in the twentieth-century United States. The social settlement movement (an integrated but female-predominant arena of scholarship) and the Women's Chicago School had together helped pioneer the development of social ■: surveys. Designed explicitly to influence public policy, surveys were col- : laborative projects to collect and then analyze massive amounts of data; about urban and industrial life. Early surveys facilitated intimate, detailed,: but often statistical analyses of immigrants' jobs, families, budgets, homes,: and neighborhoods. Whether in Chicago,43 as part of the 1907 Pittsburgh ■ Survey team,44 or with funding from the initially woman-initiated and i -controlled Russell Sage Foundation,45 women researchers published ; important surveys of immigrant life; some even wrote explicitly about the lives, jobs, and consumer habits of immigrant women.4''
Separate from the Chicago School of Sociology wras the School of Civics and Philanthropy (later renamed the University of Chicago's School of i Social Administration). Many graduates were women and found employ- % ment in municipal and state governments as researchers and administra- :\ tors of social welfare and public health agencies or in the federal Women's l Bureau. Edith Abbott (1876-1957), an early dean of the Chicago School of Civics and Philanthropy, was an expert on immigration and one of the drafters of the American Social Security Act. Known today mainly as a % founder of social work, she authored so many quantitative analyses of immigrants, female employment, and criminality that she was initially J known as "the passionate statistician."47 I
Passionate statisticians of both sexes ensured that migration and mobility remained central themes in demography and increasingly in economics. In the 1920s, statisticians' senior academic representative, f Cornell University Professor Walter F Willcox, supported several projects | about international migration from his position at the National Bureau | of Economic Research (NBER), including a collaboration with Hungarian j Imre Ferenczi (1884-1945) at the International Labour Organization (ILO). § Their work confirmed—on a far more international scale than earlier studies—the overall male predominance among nineteenth-century trans- 1 continental migrants and the anomalous female majorities among Irish migrants in many places, not just Great Britain.45 |
University-trained U.S. women demographers with expertise on migration quickly assumed leadership in tine discipline of demography and in ~}< the Population Association of America (PAA). Because international migrations declined after 1920, however, most of these women studied internal |
mi:
Data and Discipline 27 io-ration, using methods and sources that Ravenstein would have recog-
nized.49 During the 1950s, Irene B. Taeuber (1906-1974), Dorothy Swaine Thomas (1899-1977), and Margaret Jarman Hagood (1907-1963) served as back-to-back presidents of the PAA. Thomas also served as first female president (in 1952) of the American Sociological Society (later Association). Her election reflected the degree to which quantitative methodologies had—by that date—moved closer to the mainstream of U.S. sociology and, by the time of her death, had come to define it generally50—although not American immigration studies. Thomas and other female demographers distinguished themselves and their empirical research from earlier sociologists, but they also differentiated themselves from the Chicago women and survey methods that Harriett Bartlett (herself a Chicago-trained social worker) characterized in 1928 as a form of fact-finding aimed at social control and thus different from real research that "deals with genera] data divorced from time and place" and "seeks to testa general hypothesis."31 These demographers did not, in other words, embrace research as a form of political arithmetic; they considered themselves social scientists, not state scientists.
E.G. Ravenstein's foundational laws of migration echoed through the work of Dorothy Thomas, Irene Taeuber, and a third woman demographer, Hope Tisdale Eldridge (1904-1991). Thomas pioneered in applying quantitative methodologies to the study of internal migrations in both Sweden and the United States.-Although she began as astudent of internal migration,33 Eldridge later edited the UN Demographic Yearbook, a publication that continued Willcox's and Ferenczi's documentation of international migrations after World War II. Eldridge also worked with Thomas on the third volume of the monumental Population Redistribution and Economic Growth, United States, 1870-1950, a work that sought to understand the economic causes and consequences of internal migrations.54 Conrad and Irene Taeuber too initially researched internal migrations in the United States, but Irene Taeuber then turned to international work on development, demography, and migration issues in Europe, Japan, and China.55
Attentive to how sex structured both migration and fertility, Irene Taeuber and Hope Eldridge pondered the demographic foundations of what feminist scholars in the 1970s would call sex or gender roles.56 Eldridge's work went well beyond Ravenstein in noting how shifts in sex composition were understood by nonspecialists.57 She argued that popular alarm over the new (and very small) female majorities emerging in the U.S. population (largely as a consequence of immigration restriction and declining fertility) was not warranted. Although less attentive to sex in her co-authored publications, Taeuber repeatedly called attention to how the usefulness of population records diminished when record keepers failed to differentiate individuals by sex.58
28      Gender and International Migration
Although it would clearly be an exaggeration to claim that population statistics was understood as scholarly work especially appropriate for women, it had become a field of significant achievement and leadership by women. Taeuber was remembered after her death not only for making international studies more central to the field of U.S. demography but also as a "feminist and a humanist" who strove at the same time for scientific objectivity.59 Taeuber's desire to combine feminist humanism with scientific objectivity is worm emphasizing because within a decade of her death a newer generation of female scholars, seeking to shift the analytical focus of migration studies from sex to gender, sometimes offered harsh critiques of both science and scientific objectivity. As the new women's . studies of the 1970s gave way in the mid-1980s to gender analysis, many;-: completely rejected the epistemological foundation of the women's stat-:; isticians' research on sex and migration, making it as impossible for them to see the feminization of migration as it had been for the Chicago School:; sociologists fifty years earlier.
The Gender Challenge
The challenge of gender analysis to the long trajectory of studies of sex1 and migration within the state sciences and population studies developed slowly but ultimately produced a dramatic reversal of the positive association of quantitative methodologies with female scholars. Initially, a new feminist movement in the 1960s, accompanied by growing numbers ; of women gaining specialized training in history, anthropology, demography, and sociology, promised a bright future of new insights about the .' relationship of sex and migration. Feminist scholars first studied what: had been ignored in the past, notably the lives and experiences of women migrants. New research on migrant women and families—much of it : drawing on population and survey data created by the Women's Chicago ■. School—flew off the presses of American universities and journals in the \ 1970s and 1980s.60 Even in work focused exclusively on migrant women, : however, a new generation of women studies researchers increasingly.. distinguished gender from sex and sought to explain sex differences : through attention to gender relations and to the exercise of power in pri- 4 vate and public spheres. After 1985, the rise of postmodernist philosophy ;:; in the humanities and (to a lesser extent) the social sciences changed the % meaning of gender analysis, initiating debates among feminist scholars and further complicating communication between migration scholars in ■ humanities and the social sciences.
Etymologically, the concept of gender originates in classificatory ■:>. grammars typical of Indo-European languages, which distinguish among ? masculine, feminine, and neuter nouns and pronouns. The word moon, ;> for example, is masculine in some languages and feminine or neuter in
Data and Discipline 29
others. Because linguistic gender is completely unrelated to biology, it provided a powerful example of how human behaviors completely unrelated to biological sex were wrongly understood as determined by sex difference.
Already in the 1960s, scholars across many disciplines began to substitute the term gender for sex to bring more sharply into focus and to question and criticize this type of unexamined attribution of malleable and diverse social and cultural practices to biological sex differences. Thus, the analysis of gender accompanied a general, contemporary intellectual : shift away from biological, sexual, and racial modes of explanation and toward social and cultural theories of causation. The popularity of social constructionist theory contributed greatly to this early adoption of gender as an analytical category in the social sciences and in history."1 For scholars in feminist and women's studies, the difference of sex and gen-der. was not merely semantic. Such scholars wanted to know how and why humans transformed socially constructed notions of gender into certainties, about biological difference because they wanted to change such assumptions. Anthropologists' studies showed that gender ideologies differed cross-culturally, but that many cultures understood human life to be organized around separate masculine or public realms and feminine or private realms. Many furthermore associated human culture with the public realm and the private sphere as ruled by nature and biological reproduction.62 Their work influenced new research on immigrant women, as did other studies about how boys and girls were socialized into appropriate gender roles and employment.63
Such social constructionist understandings of gender did not deny the biological reality of sex difference but did reject explanations of social and cultural behavior as originating in biology. Focused mainly on internal movements, demographers of the previous generation had done little to disrupt Ravenstein's laws because the laws were applied to long-distance : migrants only/'4 One result was that social scientists studying female international migrants in the 1970s confronted persistent questions about the value of their work, given the normalcy of male majorities and male decision-makers among international migrants. Anthony Leeds offered : a particularly dismissive critique of new studies of female migrants as reductionist and unnecessary.65 Forced to document the presence and .significance of women migrants, feminist scholars continued to turn to ..statistical evidence. Still, their growing skepticism of the most abstract ...claims of scientific objectivity that accompanied quantitative methodologies was already palpable.66 Feminist scholars proved particularly skeptical of what Harriet Bartlett in 1928 called "data divorced from time and place."67 They generally preferred grounded theory and data, which often meant analysis of data either created by women or in which women's .voices and agency could be heard or recognized/'s Many feminist scholars
30      Gender and International Migration
of the 1970s also hoped to produce knowledge that would be more useful to the powerless than to the powerful.65'
Even those women studies scholars who rejected the intellectual traditions of the state sciences did not always jettison quantitative analysis of . statistical data on sex. The special issue "Women in Migration," edited in 1984 by Mirjana Morokasic for International Migration Review {IMR), the flagship journal of migration studies, included the first report of the labor statisticians on female predominance in U.S. immigration and devoted an entire section to "Census-Based Quantitative Analyses of Female Immigrants and their Labor Market Characteristics: An International Comparison." A subsequent series of interdisciplinary monographs and essay collections on immigrant women also included or referred to quantitative research.70
Although not focused exclusively on migrant populations, the posthumously published 1983 book Too Many Women ? The Sex Ratio Question, by social psychologists Marcia Guttentag (1932-1977) and Paul Secord illustrated how social constructionist notions of gender changed analysis of sex composition while rejecting the expectations of policy relevance typical of the state sciences. The two authors may have been unaware of.-. the work of Ravenstein and the female demographers studying internal migration, but they had absorbed the conventions of demography, with its interest in tracking and counting the relative numbers of men and ■■ women in human populations. At the same time, they made tlie feminist movement's concern with personal power central to their analysis and clearly wanted to assist feminists in understanding the origins of femi-nist mobilization and the demographic challenges inherent in overcoming patriarchy,71
Guttentag and Secord explored the impact of changing shares of male and female on the interaction of what they called dyadic or personal relationship power—which routinely favors the underrepresented sex, . especially in the negotiation of marriage—with structural or societal power that is, or was, almost universally monopolized by men. When men seriously outnumber women, Guttentag arid Secord argued, women use their increased dyadic power to gain marital advantages and men . use structural power to control and restrict female sexuality. The result is highly traditional gender relations, which includes male-headed households, high rates of female nuptiality, high rates of marital fertility, and low rates of female labor force participation. When women outnumber men, by contrast, a sexually more permissive society, with higher rates of female employment, emerges but men so undervalue women as sexual . and marital partners that their misogyny sparks female resentment and encourages feminist mobilization.
Guttentag and Secord illustrated these ideas with a series of historical and sociological case studies, some more plausible than others.72 Several :
Data and Discipline 31
case studies posited migration as a cause of unbalanced sex composition. For example, they treated Chinese immigrants in the United States as an example of a heavily male population in which almost all women were prostitutes/3 and contrasted the migrations of medieval France and Spain74 to explain the development of courtly love. Their study of the nucrations of European colonizers and settlers of North America75 initiated a lively discussion of the politics of marriage formation in early America.76 The authors also identified sharp differences in the composition of African American and Jewish American populations (both female predominant)77 and Hispanics (male predominant, largely as a result of mio-ration). They encouraged scholars to explore the social, demographic, and cultural consequences of such variations, including changing "sexual behaviors and sexual mores . . . patterns of marriage and divorce, chil-drearing conditions and practices, family stability, and certain structural aspects of society itself."78 In addition, they focused special attention on how the fall in U.S. fertility in the 1930s and its swift rise again after the end of World War II created a marriage squeeze for baby-boom females and they traced to this "marriage squeeze" the rebirth of American feminism in the 1960s. Guttentag and Secord, in other words, suggested that changing sex ratios had consequences as well as causes.
Guttentag and Secord's work did not gain broad recognition or acceptance from feminist scholars, perhaps because the shift away from women's toward gender studies had already begun by the time it was published.71* By ■■■ the mid-1980s, feminist scholars' uncomfortable awareness about how Women-centered inquiry had contributed to marginalizarion within : the academy preceded but was strongly reinforced by the turn toward studies of culture and discourse, based on the foundation of postmodernist philosophy and radically relativist scholarly epistemologies.50 Even those who did not accept postmodernists' assumptions came to appreciate that gender was not so much a measureable thing but a relationship between culturally constructed notions of masculine and feminine such that any change in ideas about masculinity required changes in ideas about femininity. Gender, unlike sex, was relational but also changeable and fluid. Pushing further, queer theorists soon also chal-: lenged the biological foundations of sex difference, too, denying that sex was fixed, invariable, dichotomous or even measureable.51 Indeed, many gender scholars challenged all binaries, or dichotomies, seeing in them . the origins of unequal power, hierarchy, and patriarchy. In this view, . only a challenge to the binary of male and female sex can challenge male power itself.
Scholars in migration studies did not ignore these developments,82but they had little impact on scholars working with quantitative method-: ologies. Without binary, dichotomous, or bivariate categories of data— notably of male and female sex—quantitative and statistical study of
r-
32     Gender and international Migration
migration seemed impossible. Indeed, feminist scholars since Taeuber generally insisted that the creation of data on sex was necessary if scholars are to study the lives of women as well as men. In a review of scholarship on gender and migration, Katharine Donato and her colleagues also see data distinguishing male from female as the foundation for gender analysis using quantitative methods.83
Exacerbating this epistemological divergence were other closely related methodological concerns. Beginning in the 1970s and 1980s, increasing numbers of feminist scholars began to reject quantitative methods as hierarchical, female-unfriendly, or even malestream tools of analysis, approvingly citing African American poet Audre Lorde's trenchant observation that "The master's tools will never dismantle the master's house."S4 In this formulation, quantitative methods and the state sciences are the master's tools; the master's house is patriarchy. This kind of thinking suggests but rarely states outright that quantitative work—so long an ■ arena of female accomplishment—is instead a masculine methodology. By the 1990s, feminist scholars in gender studies increasingly privileged qualitative methodologies while, for their part, some quantitative social. scientists characterized gender analysis as tainted by tlie radical relativity of postmodernist philosophers.65 As a result, according to Sara Curran. and her colleagues, gender is a key constituting factor in only a minority of recent sociological migration studies; other scholars agree.*''
Despite the so-called linguistic turn, the methods of postmodernist literary scholars have never completely dominated newer studies of gender and migration. While they certainly devote greater attention to subjectivity, to gender relations, and to masculinity and femininity, and certainly employ a wider range and mixture of methodologies, newer studies rarely question the material reality of a human body—that of the migrants—that move from one place to another. In very few studies of gender do scholars treat the migrant—as was sometimes tire case in the humanities—as a disembodied "text" to be interpreted independent of ■ social or historical context. Attention to the fluidity of gender certainly empower Patricia Pessar and Sherri Grasmuck to challenge scholarship organized around the male migrant and to suggest that complex gender dynamics within families, kin groups, and households are as important in shaping migration decisions and trajectories as the more frequently studied search of male employment in global labor markets.37 Divisions of labor within families determine which persons migrated without jeopardizing the functions and needs of the household. Both in Europe and in the United States, new studies seek to demonstrate how gender mattered in migration theory, decisions to move, work, or reproduce, and in citizenship, welfare, and politics.88 Depending on social and economic resources, young men may have more or fewer incentives to migrate than young women; access to and rates of naturalization may also differ by
Data and Discipline 33
render, as well as the forms of political activity open to or undertaken by male or female migrants.
For their part, demographers, population scientists, and economists interested in migration sometimes wrestle with and sometimes try to ignore the linguistic turn. Russell Menard concludes simply that "the notion of a postmodern demography is an oxymoron."m Despite such pessimism, feminist demographers sometimes seek to bring into their social science discipline some insights from postmodernist understandings of gender more common in the humanities, when studying the relationship of cultural understandings to fertility.-'0 Some social science journals publish occasional quantitative studies of migrant women,1" but quantitative studies in history have almost completely disappeared.92
As gender theorists mounted their challenge to studies of dichotomous sex, feminist demographers working outside universities in governmental '■: and international organizations continued to follow the paths of their own ■'; discipline and to build on the example of earlier feminist demographers. At the United Nations, feminist demographers called for better recording of migrant sex and more analysis of women migrants. As they increas-; ingly sought—and found—new data, they also created the foundation for the discovery of the feminization of migration by scholars who still used . quantitative methods and statistical data, even as fewer and fewer scholars in gender studies did so.
Naming the Feminization of Migration
Among the newer generation of feminist demographers, none did as ..much as Hania Zlotnik to facilitate the discussion of the feminization of ,: migration. By the time she retired as the first woman director of the UN ■Population Division in 2012, the feminization of migration had become a v buzz word in scholarly discussions. The United Nations, along with many
■ other international and nongovernmental organizations (NGOs), were
■ also being asked to solve problems that seemed to snap into focus after .scholars had identified the feminization of migration and acknowledged— statistically and demographically—the presence of women. Still, as Eldridge noticed in 1947, that awareness could also spark popular alarm. In this case/ concern focused on the victimization of women migrants as exploited domestic workers and as trafficked victims of a global sex industry.
Zlotnik studied mathematics in Mexico before completing a 1977 Ph.D. in demography at Princeton. In a recent interview,-3 she expresses her early frustrations that the ILO (the international agency that had initiated the collection of data on migration in the 1920s) collected data in which women migrants remained largely invisible. At the time, the ILO was concerned mainly with the problems of male labor migrants in Europe. Many of Zlotnik' s generation shared that perception: Mirjana Morokvasic
34      Gender and International Migration
expressed her own frustration in titling her important early publication, "Birds of Passage [that is, labor migrants] are also Women."94 A few years later, with more knowledge acquired as a UN insider, Zlotnik acknowledged that both the ILO and the UN had been working hard to correct problems in the collection of migration data. Still, she summed up her reading of decades of such work with the conclusion that "the crusade to achieve greater homogeneity in the concepts underlying flow statistics on international migration was started at least 55 years ago, but despite some encouraging developments, it is unlikely that homogeneity will be achieved during the rest of this century."95
As an employee of the UN Population Division after 19S2, Zlotnik and like-minded demographers spearheaded a movement to analyze sex and migration based not on ILO data but on stock or census data that the UN had begun to compile in the 1950s. By using new data, the demographers hoped, the world could gain a clearer understanding of sex and migration because almost all censuses differentiated by sex and noted place of birth, allowing researchers to identify "persons living outside their country of birth" as migrants, hi rapid succession, the UN Statistical Office issued new recommendations on how national censuses should collect data on the sex of persons born abroad, and specialists in both Europe and Asia tackled the data collection problems specific to migrations in their regions.96
By 1986, the UN Department of International Economic and Social Affairs (through its Population Division) had surveyed the availability of national data sources (again, mainly census reports but also population registration systems) and proposed methods to estimate international migration. The UN Population Division was itself busily engaged in analysis of this data.97 At the start of the next decade, Zlotnik wrote that "women constituted 48 percent of all persons enumerated outside their country of birth at some point during 1970-1987," but added (undoubtedly after reading Houstoun) mat "in terms of flows, until the early 1980s women had predominated over men among permanent immigrants admitted by the United States."" In 1990, the UN Secretariat sponsored an Expert Group meeting on International Policies and the Status of Female Migrants in which Zlotnik was a prominent participant, well positioned to spread the word about the importance of women among international migrants counted in census data.9-
Then, in 1993, as indicated in the introduction to this volume, Stephen Castles and Mark Miller published the first edition of their influential and widely read Age of Migration. Well aware of the UN initiatives, they made the feminization of migration, along with the ubiquity and rising rates of migration in every region of the world, a signal characteristic of the new global age they proclaim. Their book introduced the feminization of migration to a readership well beyond university graduate seminars, ■ sparking: discussions that have continued to the present. The book also
Data and Discipline 35
inevitably suggests that feminization is a recent development, closely linked to new'forms of global integration.
What is most striking about the spreading attention to feminization is that it occurred only after the UN demographers chose to estimate, model, and begin to publish migrant population characteristics using census data. Only then could social scientists, accustomed to working with such data, see and name the feminization of migration. Yet as we discovered in the course of our research, and as we show in the chapters that follow, other data had long existed and pointed toward the rising presence of women among international migrants. For example, a 1953 compilation of ILO data that had thoroughly documented feminization had been almost totally ignored, even after being advertised in sociology, statistics, and demography journals in 1954 and 1955. Social scientists did not use it, and its first scholarly citation did not appear for forty years, in a study of modern slavery written by an historian.™' That statisticians had been using census data to study sex and migration since the days of Ravenstein, and that most studies of internal migration by feminist demographers in the interwar years were also based on census data, were powerful influences on Zlotnik and colleagues. The use of new data made feminization itself seem a new development.
Discussion of the feminization of migration has had mixed consequences. On the positive side, it has helped make gender dynamics of migration, understood as shaped by gender relations and ideology rather : than by sex, a somewhat more central concern of migration scholars work-■;;;ing with quantitative methods. At the same time, it meets feminist theorists' call for extending gender analysis beyond the study of local, family, or household movements to all spatial scales, including the global, an approach that Pessar and Mahler call gendered geographies of power.101 In ■small ways, it also encourages scholars of migration to rethink the causes of male and female migration and the composition of different types of migrations.
The discovery of the feminization of migration also encourages greater s attention to women as autonomous labor migrants, especially as workers ■doingpaid, reproductive work as nannies, caregivers, and domestic servants.102 Such research has begun to call into question the still-powerful .association of individual labor migration with masculinity.103 Scholars :now also emphasize the high proportions of women and children in refu-gee migrations,1114
Positively, too, scholars have begun to explain the causes of feminization by examining both emigration and immigration. In 1984, U.S. labor .statisticians suggested that feminization was a consequence of U.S. immx-: gration restriction, reduced volumes of immigration, and economic crisis.105 Using the same evidence, Gabaccia finds that early U.S. restrictions— ■imposed, for example, on the Chinese and Japanese—resulted in significantly higher proportions female among Japanese than Chinese migrants
36     Gender and International Migration
Working with Immigration and
(discussed further in chapter 4). Naturalization Service (INS) data, Donato finds that heavily female recent migrations to the United States are from countries with major U.S. military bases, suggesting that family formation and family unification— once privileged by U.S. rules for allotting scarce visas—offer a powerful explanation for feminization.107 Comparing Zambia and the Philippines, Maria Floro and Kendall Schaefer follow Sassen-Koob in attributing the feminization of migration to increasing demand for care and service workers, as does Kofman.10s In addition, Erin Hofmann and Cynthia Buckley suggest that the rising prevalence of divorce, lack of local economic opportunities, and the importance of human capital were the initial motivators for women's outmigration from postsocialist Georgia.109
Recent scholarship on female migrants focuses considerably more attention on the most exploited and vulnerable of female migrants rather than on the majority of women who are well educated, high- and low-skilled workers, or traveling to unify families. This focus is evident in the 2006 UN report; its analysis of migrant gender compositions worldwide provided a springboard to two chapters titled "Trafficking in Women and the Exploitation of Domestic Servants" and "Refugee Women and Asylum Seekers." In addition, although domestic violence, HIV/AIDS; genital cutting, honor crimes, and violence against women are not limited to migrant populations, these themes also receive ample coverage. References to the feminization of migration often frame studies of women and girl trafficking.110 Indeed, although no one has ever suggested that more than a small minority of migrant women are trafficked, about one-third of scholarly articles appearing after 1983 and addressing the feminization of migration focus exclusively on the sexual trafficking of migrants or on women working in the sex industry. In addition, studies. of labor exploitation of largely female domestic servants and health-care workers often reference the feminization of migration.111 Rarely do authors present data on how common or uncommon such labor exploitation is. Intentionally or not, and without recourse to data, such studies suggest that the feminization of migration has been largely driven by or resulted ■ in the exploitation of women migrants. One important exception is Rhacel, Parrefias, who challenges this perspective by describing how Filipina' hostesses in. Japan are working women who migrate by choice and are;;: not coerced into prostitution, though they remain vulnerable due to the.;: imposition of regulation by nation-states.112
Scholars have now begun to describe how more popular and journal-v istic writings about female migration also encourage a discourse of popu-.: lar alarm over rising levels of female migration.113 Focusing largely on. Europe, Marlou Schrover concludes that in popular discourse, "migrant-men are seen as causing problems and migrant women as having them."!M Such negative attention to women's problems during periods of massive,,
Data and Discipline 37
fnigtation are by no means new; neither is the tendency to understand migration as threatening women morally, in the nineteenth century, awareness of male majorities among migrants raised questions about women's moral behavior when they were left behind in sending societies,115 and : evidence of even relatively small shares of women among migrants fueled concerns about a white slave trade operated by exploitative migrant male procurers.116
Advocates for migrant women and scholars in gender studies sometimes also build their arguments for policy interventions around the vulnerability of contemporary female migrants. Thus Glenda Labadie-;Johnson notes, "Until recently governments and international fora have done very little to address . . . the issues raised" by the feminization of migration.117 In the course of our own research, we repeatedly encountered evidence that awareness of the feminization of migration was "provokingpopular alarm. In March 2010, we received an e-mail invita-i: tion to participate in a virtual briefing (by telephone) on "Gender, Race, and Migration," organized by representatives of advocacy groups such as Women Watch Africa, Domestic Workers United, Global Fund for "Women, and Priority Africa Network. The invitation began by citing the 2006 UN report but then asserted, "What is particularly alarming is the increase in the number of women who are now leaving their homes in highernumbers than ever before. Traditionally/' the invitation continued, "men had left home and sent remittances back home to women, children : and the elderly who stayed put on the land carrying on traditions, farm-: ing lands, and caring for communities. Women are now almost equal : to men in leaving their homes seeking employment in largely service :industries." It concluded, "the sex industry and trafficking of women are -also contributing to increased migration"—an argument never made in the 2006 UN report.118 Most recently, a professor of European immigration law responded to a scholarly presentation of our finding (that gender balance and not feminization was the most important trend in the present moment) with a blunt request for an assessment of whether feminization .of migration was good or bad for women. He was quite clear in enunciat-; ing his fear that exploitation of trafficked women and of domestic and sex workers had produced what we instead describe as gender balance and a kind of demographic normalcy in migrant populations. One goal of this book-is to extricate the study of gender and migration from such popular discourses of alarm.
Conclusion
By focusing on how migrations became relatively more female and relatively less male, new scholarship on the feminization of migration has ^eSUn to reverse the sharp divergence of quantitative studies of migration
38      Gender and International Migration
k
and gender studies mat emerged in the 1980s. Numbers, too, can reveal and help scholars understand the fluidity of gender ideology and gender relations, and even those in history and the humanities seem more willing to learn the language of statistics. In 1994, a high-profile historian proclaimed the arrival in migration studies of a post-structuralist structuralism that incorporates the insights of both structuralism and its critics.11" The past decade has seen both rising enthusiasm for mixed methodologies and a growing awareness of how qualitative and quantitative research produce perspectival but useful and sometimes complementary knowledge about gender.120 Our book appears then as the methodological chasm separating gender studies and quantitative social science has narrowed somewhat.
Using an eclectic array of methods from diverse disciplines, we seek to demonstrate that empirical evidence and systematic quantitative analysis j of bivariate data on sex reveals the fluidity and relationality of gender as it is constructed in both sending and receiving societies, assisting scholars to move bevond the idea that the feminization of migration is new or unprecedented. In the next chapter, we focus specifically on key concepts such as the difference between sex ratios and gender composition, and : the origins, creation, and uses of flow and stock data. We use these data to present a four-century survey of how gender ideologies and gender relations created differing kinds of migrations with significantly diifer-ent balances of male and female migrants. For readers in both the social sciences and humanistic gender studies, a thorough understanding of : how evidence is produced is an important first step toward analysis and interpretation.
This approach, we believe, can both assist scholars in the humanities » in understanding the language of statistics and social scientists in understanding the causes and consequences of global and regional shifts in migrant gender composition that have produced gender balance. Thus we draw insights from the humanities by calling attention to the constructed nature of both statistical data and the categories used to measure the relative numbers of male and female migrants, and from the social sciences by remaining resolutely and unapologetically empirical in our quantitative analysis of statistical data on binary sex. As we show in sub-. sequent chapters, an analysis of variations in the gender composition of migrant populations reveals the importance of understanding both their causes and consequences.
= Chapter 2 =
Analyzing Migrant Gender Composition with Statistical Data on Sex
The sex ratio is the most widely used analytical category for the study of the relative numbers of males and females in populations.1 Yet recent studies of the feminization of migration—including the 1984 and 2006 statistical studies previously discussed—do not measure the relative numbers of male and female migrants this way. Instead, they compare the percentage female among migrants over time and space. In this book, we too focus on the percentage female among migrants, and we analyze migrant gender composition rather man sex composition to explore how gender ideology and gender relations influenced migration over four centuries of human history.
One of the most important contribu tions of social constructivis t gender studies for scholars working with quantitative methods has been its attention to how and by whom data are created, for what purposes, and how key concepts have been operationalized. As chapter 1 reveals, new data figured prominently in the discovery of the feminization of migration. This chapter examines the creation of data and analytical categories that have been central to quantitative analyses of sex, gender, and migration. It also provides an explanation for the choices—of data, categories, and measures—that structure our analysis in the chapters that follow. Acknowledging the historical construction of data, categories, and measures opens them to interpretation, critique, and revision. Such interpretation, critique, and revision are essential if gender analysis is to be combined effectively with quantitative analysis of migration.
This chapter begins with an examination of the history of data created about mobile people and their (binary) sex. It examines the use and opera-tionalization of the sex ratio for the study of variations in the numbers of males and females in human populations, and explains why we measure migrant gender composition rather than the migrant sex ratio. It also
1Q