Progress in Human Geography 22,1 (1998) pp.

3953

Hierarchical households and gendered migration in Latin America: feminist extensions to migration research
Victoria A. Lawson
Department of Geography, Box 353550, University of Seattle, Seattle 98195, Washington, USA

Abstract: In this review essay, I argue that migration theory can be advanced by analysing gender differences in migration processes. I bring together feminist empirical work from diverse settings within Latin America in order to illustrate and discuss theoretical extensions to migration research. In particular, the discussion focuses on the centrality of intrahousehold power relations and dynamics for understanding who migrates, and with what consequences. I further argue that these theoretical understandings emerge from the culturally and historically specific operation of processes in particular places within Latin America.
One of the crucial tasks for future theory (and policy) . . . will be to take on board not only investigation of the causes, but also the consequences of gender-differentiated population movement (Chant, 1992: 200).

Introduction

In this review essay I argue that migration research can be advanced by researching gender differences in migration processes in their cultural context. In particular, I bring together feminist empirical research from diverse settings within Latin America in order to argue for extensions to existing migration theory. I argue that in order to understand the consequences of migration it is crucial to look at feminist research on intrahousehold dynamics and power relations. Feminist scholarship has revealed how origin and destination household social relations help explain who migrates (i.e., migrant's position within household divisions of labour) and with what consequences (i.e., urban destination labour market experiences). I further argue that understanding gendered migration
c Arnold 1998 *

03091325(98)PH181RA

40 Hierarchical households and gendered migration in Latin America

requires culturally and contextually specific research, and so I draw here on empirical research from Latin America to illustrate the diversity of processes operating within this region. In so doing, I invoke some generalities in process across the region, while also illustrating the ways in which gendered migration is produced in specific ways in particular local contexts that are constituted by historically specific social, economic and political relations. Despite two decades of attention to `women in development' in both academic and policy circles, women's `place' in Latin America is `. . . still the private sphere of home and family, and men's, the public realm of work and politics' (NACLA, 1993: 16). For many women, their participation in paid work, politics and popular culture is circumscribed by their domestic roles and responsibilities. A growing volume of research is finding that women's migration in Latin America is also profoundly shaped by gender divisions of productive and reproductive work in households. Specifically, we know from research on gender and migration that men are more mobile than women, that men migrate to a greater variety of destinations and have a greater range of employment opportunities in destination areas. However, while much migration research has assumed that women migrate as part of a household unit rather than independently (Bilsborrow, 1993), recent studies show that women equal or outnumber men in urbandestined migration streams in Latin America and that the majority of low-income women are economically active in their urban destinations (Chant, 1992). Further, recent research on labour market segmentation in Latin America shows that far larger proportions of women than men work in informal jobs where incomes are low and labour protections are loosely enforced (Lawson, 1995). These job characteristics have crucial welfare implications for women and men in-migrants and their subsequent life trajectories in urban destinations. Given these marked gender differences in migration processes, I argue for increased attention in migration research to feminist theory on the role of intrahousehold dynamics in both migration decisions and subsequent destination labour force participation.1 Migration researchers have paid little attention to the ways in which household divisions of labour (gender ideologies, roles and responsibilities) and significant life events shape migration and access to employment for new arrivals in Latin American cities (Radcliffe, 1991, and authors in Chant, 1992, are notable exceptions). Feminist researchers have emphasised the importance of intrahousehold power dynamics in shaping divisions of labour by assigning domestic work and income-earning responsibilities to various household members (Folbre, 1988; Bruce, 1989; Hart, 1992; Wolf, 1992). These household divisions of labour are fluid and complex, emerging from material inequalities among household members and from the operation of socially constructed gender ideologies. Individuals' positions within household divisions of labour are typically gender ascriptive but their assignment can vary according to household type (extended, nuclear, women headed), stage in household life course, rural or urban setting and cultural context. I argue here that the inner workings of migrant households (in both origins and destinations) are key to re-examining the classic migration questions who migrates and why, what are the material consequences of migration, and how does mobility rework gender divisions of labour and gender ideologies and identities? This work is timely for as Gilbertson and Gurak (1992: 23) note `. . . few studies have empirically explored the role of families and households in the migration and settlement processes as a way to better understand adaptation patterns of immigrant groups'.

Victoria A. Lawson 41

II

Mechanisms and consequences of migration: theoretical intersections

Migration research is supported by a rich and voluminous literature encompassing several theoretical perspectives and scales of analysis, each of which contributes to our understanding of gendered migration processes.2 Neoclassical economics has generated two influential approaches to migration modelling macro and micro models. Macro models analyse net migration flows, highlighting the importance of labour supply and demand conditions across regions in motivating population movements (Lowry, 1966; Greenwood, 1975). By contrast, microeconomic models focus on the behaviour of individuals rather than the determinants of overall flows. These approaches conceptualize population movement as migrants maximizing returns on their individual labour by comparing potential lifetime earnings available via migration with those in their place of origin (Todaro, 1969; Corden and Findlay, 1975; also see Brown and Lawson, 1985, for a review of both approaches). These neoclassical models highlight the volumes, patterns and contrasts in men's and women's migration and so have provided a foundation for study. However, these approaches elide gender differences by assuming that economic motivations (higher wages in destinations) are equally important for both genders. The restrictive definition of work adopted in these models (public sphere, paid productive work) has effectively marginalized some women's motivations for migration since many of their activities in the household are undervalued and/or rendered invisible (Radcliffe, 1991; Chant, 1992; Bilsborrow, 1993). Another shortcoming of these models is that they pay little attention to intrahousehold dynamics which shape, or even directly determine, individual decisions. Political-economy approaches to migration present a complementary perspective, emphasizing the organization and transformation of macroscale political and economic systems of relations, and resultant impacts upon the sites and conditions of women's and men's work (Roberts, 1978; Shrethsa, 1988). Recent research has focused attention on changing gender divisions of labour in the international economy and the expanding role of women in industrial production (Nash and Fernandez-Kelly, 1983). Free trade zones and export-led growth strategies in many settings have expanded employment opportunities for women and have influenced their mobility patterns (Mies, 1986; Massey, 1994). At a more microscale, political-economy approaches conceptualize migration in terms of the class position of migrants rather than in terms of individually held attributes (i.e., education, age, gender). From this perspective, migration decisions and their consequences involve distinct class biases. Members of dominant classes control productive resources and appropriation of economic surpluses and may make strategic choices to migrate, whereas members of subordinate classes are more likely to migrate as a survival response to changes in a production system over which they have no control (Shrethsa, 1988). This literature has contributed in important ways to our understanding of why migration occurs and how it impacts migrants from distinct class positions. Here again, however, economic factors and production side shifts are assumed to drive migration, with little attention paid to household dynamics, reproduction activities and gender relations as determinants of migration. The household strategies approach to migration extends the above perspectives by emphasizing the importance of reproductive work in shaping the mobility of various household members (Bilsborrow, 1993). In these approaches, the household is viewed as a key socioeconomic unit that organizes resources, including labour, in order to accomplish a combination of productive and reproductive tasks (Mincer, 1978; Katz and

42 Hierarchical households and gendered migration in Latin America

Stark, 1986; Radcliffe, 1991: 135). The neoclassical bargaining perspective has opened up the household, arguing that household members do not have identical utility functions (Lauby and Stark, 1988). The bargaining approach conceptualizes a type of co-operative conflict bargaining process in which each household member has a `breakdown point' (Manser and Brown, 1980; Sen, 1990; reviewed in Folbre, 1986). Under this model, an agreement will be reached if it provides a better outcome for participants than their breakdown position. Ultimately, however, a co-operative solution is reached and the analytical focus is upon conceptualizing migration as a strategy of risk minimization for the household as a whole (Lucas and Stark, 1988). These approaches raise, but do not resolve, important questions about what bargaining power is, and why certain members of the household have greater bargaining power than others. Recent feminist research responds to these questions, arguing that intrahousehold hierarchies of gender and generation can help to explain the mobility of women and men (Young, 1980; Collins, 1986; Chant and Radcliffe, 1992; Hondagneu-Sotelo, 1994). Research in a variety of settings has examined cultural constructions of women's and men's roles finding that women's positions within household divisions of labour produce very different reasons for, and patterns of, mobility. While still focused on household divisions of labour, this work differs from the above in its critique of household `strategy' as uncritically implying harmony and co-operation. Rather, feminist analyses focus on intrahousehold hierarchies of power and examine struggles over decisions, rules, obligations and gender ideologies (Westwood and Bhachu, 1988; Bruce, 1989; Chant, 1992; Hart, 1992; Gonzalez de la Rocha, 1994). This research emphasizes the ways in which patriarchal authority constructs household divisions of labour, including both power relations operating in the home and in the wider sociocultural context (Chant, 1992; Hart, 1992). Examining these power relations helps us understand the construction of particular household divisions of labour, the presence or absence of income-pooling behaviours (see case studies in Chant, 1992), the impacts of changing workforce participation on the proportion of women's and men's incomes going to household budgets (Beneria and Roldan, 1987; Beneria and Feldman, 1992; Chant 1992), and the impacts of changing workforce participation for children's welfare (Bruce, 1989). Feminist research also goes further, examining not only the empirical outcomes of intrahousehold inequality but also the mechanisms producing those unequal outcomes. As Hart (1992: 121) argues, we need to understand how domestic divisions of labour are produced and maintained.
In other words, to come to grips with intra-household bargaining and contestation, we have to engage directly with questions of ideology and the construction of meaning, and recognize how struggles over resources and labor are simultaneously struggles over meaning.

Research by Carney and Watts (1990), Radcliffe (1991) and Hart (1992) demonstrates clearly that the paid and unpaid contributions of male and female household members are valued differently within the domestic sphere. A hierarchy of importance (valuation) of the contributions of household members and of their access to subsistence is defined in ideological terms invoking gender roles and responsibilities. Notions of `custom' which are dominant representations of gender, family, domestic harmony and sexuality are frequently invoked as the basis for domestic control. This approach moves us away from an exclusive focus on the household, incorporating constructions of gender into communities, workplaces, and other sites of interaction. As will be revealed in the case studies below, intrahousehold bargaining and contestation over these

Victoria A. Lawson 43
meanings are used to maintain or redefine forms of control over the labour of various household members and a key mechanism of this control is gender. Another key difference between neoclassical household strategy models and feminist approaches is that the latter re-examine the category household. Feminist research has moved away from the household as a fixed, static category (of male-headed household) and has directed attention to variations in household composition (for example, women headed, male headed, extended, nuclear; Bruce, 1989) and how these variations themselves may be crucial to understanding both reasons for, and consequences of, migration. Recognizing the fluidity of household forms is also important for understanding the various combinations of reproduction, production and consumption that household members will be involved in. For example, in Latin America we find levels of female household headship at around 20% of all households, and in the Caribbean the proportion is about 26%, and this is higher than in many other non-Western settings (Folbre, 1991; Momsen, 1992; Safa and Antrobus, 1992). In both colonial and contemporary periods, male labour migration has often led to the formation of de facto womenheaded households (Folbre, 1991). As land tenure arrangements and the profitability of peasant agriculture have been transformed, many households have experienced male out-migration (either permanent or as circulation) and this has produced female headship on both a temporary and permanent basis (Gisbert et al., 1994). Migration across boundaries has also driven female headship formation in Mexico and the Caribbean, as economic opportunities in the USA have led households to adopt spatially extensive survival strategies (Momsen, 1992). Women's household headship, whether through nonmarriage, separation or widowhood, must often entail wage-earning activity in the home, but often for different reasons than occur under patriarchal controls in many maleheaded households. Considering the diverse household power relations under which paid work and labour market participation occur, and understanding that much of that work occurs beyond the public sphere of the labour market, can substantially extend our understandings of the consequences of migration within the contexts of particular household forms (Allen and Wolkowitz, 1987; Beneria and Roldan, 1987; Fortuna and Prates, 1989). In addition to opening up the household category, feminist researchers have argued for the need to re-examine critically the connections between geographic scale and the concept of household (Hart, 1992; Momsen, 1992; Staeheli and Lawson, 1995). Prevailing conceptualizations of households assume that they are local in scale and form. This assumption leads to theorizations of the household as derivative of migration that is, households are discretely located in space and are uniquely left, brought or formed in origin or destination areas. While this conceptualization represents one set of relationships between household structure and migration, feminist research suggests an expanded conceptualization. Households may be structured supralocally, incorporating multiple members in diverse places who remain part of the income-pooling unit directly, or who continue to exercise influence over household dynamics (through direct remittances, ownership of assets, visits and/or gift giving, and input/control over key decisions). This phenomenon is common in Latin America and has been documented by Hondagneu-Sotelo (1994) and Mountz and Wright (1996) for MexicoUSA migration, Momsen (1992) for Caribbean migration, and Arguello (1981) and Bilsborrow and Peek (1992) in studies of internal migration. I have argued here that migration theory will be significantly enriched through attention to feminist research on the inner workings of migrant households. In the next

44 Hierarchical households and gendered migration in Latin America

sections of this article I review empirical evidence from Latin America which demonstrates the links between household divisions of labour and the dynamics of gendered migration. This discussion draws on a series of case studies and life histories of migrants from various Latin American countries. The focus on case studies rather than aggregate statistical analyses is necessary because the majority of multivariate studies treat migrants as individuals (due in part to data constraints). Aggregate analyses in that genre lose the connection between individual experiences of migration and the household and community structures within which migrants are embedded. By contrast, case-study evidence can reveal the dynamic and context-specific power relations operating in households and the ways in which these vary according to the cultural construction of gender relations, political-economic macroforces operating and the like. The case studies discussed below shed light on the ways in which origin household gender divisions of labour shape mobility and on the ways in which destination household divisions of labour shape women's entry into destination labour markets. As such, my discussion is suggestive of ways to widen the focus in migration research I pose questions for further research rather than seek to generalize about broad and predictable patterns.

III

Origin household divisions of labour shape the dynamics of migration

Each of the following case studies illustrates the ways in which household divisions of labour in origin areas are culturally constructed and reworked in the context of broader political-economic dynamics. In this section, I discuss three cases to illustrate how material and discursive processes operate to shape gender roles and ideologies within households, how these influence who migrates and why that migration occurs. Radcliffe's (1991) study of peasant households in the Peruvian province of Calca (in the southern Andes) reveals the way in which position within household division of labour shapes out-migration. Within peasant communities, household divisions of labour are continually in flux in response to broader political-economic forces and intrahousehold struggles. Historically, the social construction of gender was based on complementarity between men and women, a system in which both partners held equal status through their contributions to subsistence. Children were trained in gendered skills and were taught about their gender roles and rights in households, with kin, and in the community. As this agricultural economy has increasingly commoditized, peasant households have become multiclass units in which the class position of household members is differentiated in terms of their insertion into various labour markets (circulatory agricultural work, urban domestic work). This has led to the breakdown of the complementarity system and patriarchal relations within households, and communities have redefined women's and men's access to the peasant economy and wage labour spheres. According to Radcliffe, patriarchal sociocultural relations in Calca communities exclude women from key agricultural tasks (ploughing, harvesting staples such as maize and potatoes) such that while women may own land, they cannot accomplish the entire agricultural cycle without male labour (Radcliffe, 1991). Indeed, Radcliffe found a strong taboo against women engaging in `male' tasks such as ploughing and that unmarried women could not participate in reciprocal labour exchanges between households. The breakdown of complementarity and the exit of both men and women into external labour markets have disrupted and transformed

Victoria A. Lawson 45
long-standing relations of exchange and reciprocity and in the process have transformed the relative valuation of men's and women's contributions to peasant households. Thus women's labour is actively being constructed as marginal to agricultural subsistence work so that those women not needed for domestic work are `surplus' labour. Accordingly, the reworking of peasant community divisions of labour is shaping women's and men's propensity to migrate. During fieldwork in the mid-1980s, Radcliffe found that mothers are typically nonmigratory, being responsible for childcare, household maintenance and daily needs of male agricultural workers, whereas daughters are less involved in subsistence production and so are more likely to be expelled into external labour markets where they are no longer a drain on family resources. The rigidity of gender divisions of labour in the Calca case is not repeated in all Peruvian peasant communities, but in this example it is the reworking of household gender relations in the context of particular, locally constructed household divisions of labour that is driving the mobility of daughters. Radcliffe found that one third of eldest daughters had out-migrated as compared to only 14% of sons. Since women command lower wages than men in destination labour markets the household is not maximizing potential economic returns by sending women. It is locally constructed, gendered social relations, designating women as reproducers, as marginal to agricultural work, and as secondary, supplemental wage earners that shapes the mobility of both daughters and mothers. Research by Gisbert et al. (1994) in Bolivia in the Campero and Mizque provinces also found that young, unmarried women in rural peasant households are frequently migratory. In these peasant communities, it is important for women to bring a dowry to their marriage. The contemporary political-economy of Bolivia includes a rapid commercialization of agriculture in recent decades, parcelization of farms and continuing highly unequal land distribution patterns, and an inability for peasant households to maintain their subsistence livelihoods as they had previously. These complex processes are interacting to produce high levels of rural poverty and landlessness. In this context of land scarcity, and under patriarchal cultural systems, inheritance patterns have favoured passing scarce land to male offspring. Under these conditions daughters, who also experience strong pressures to bring resources to their marriages, often generate a dowry through wage labour migration to cities. In this case study, nearly half of the women interviewed had migrated as young, unmarried women between the ages of 16 and 20 years. Here again, gender relations play an important role in determining who migrates. Local cultural constructions of marriage require that both men and women bring land to their marriages. Under conditions of increasing poverty, the greater ideological and material importance of sons over daughters is reproduced by passing the scarce land available only to sons. Gendered expectations of women also having resources in order to be marriageable promote daughters' (temporary) out-migration from rural households and external (urban) labour markets which then further reinforces their marginal positions within households and the community. Further, it is a particular group of rural women who migrate and are so marginalised, defined in terms of gendered position within household unmarried daughters. The above cases illustrate how gender prompted the migration of women from their origin households. A recent book by Hondagneu-Sotelo (1994; Gendered transitions) examines how the migrations of single women and single men from Mexico to the USA were situated within very different household power struggles. These differences have their roots in patriarchal cultural relations that have historically maintained tight controls

46 Hierarchical households and gendered migration in Latin America

over the work and mobility of Mexican women. Hondagneu-Sotelo's work is interesting, however, in that it stresses both the persistence of these gender differences and also the fluidity of gender roles and relations under changing political-economic conditions. The dramatic income gradient between Mexico and the USA has historically fomented male labour migration (which has periodically been embraced and demonized by the USA). However, by focusing on MexicoUSA migrations during the 1970s and 1980s a time of economic crisis, increasing female industrial and home-based employment, and cutbacks and layoffs in those sectors that typically employed men Hondagneu-Sotelo illustrates changes in gender relations and so mobility. Specifically, she demonstrates how the long-term out-migration of fathers and brothers, women seeking out new types of paid work, and profound changes in the role of the state in daily life under crisis have combined to foment changes in patriarchal gender relations. Through this study, gender is seen as a complex and contradictory system. A key finding of Hondagneu-Sotelo's study is that the independent migration of young men is frequently viewed as a rite of passage, a way to assert their masculinity and autonomy from their fathers. Migration to the USA is seen as part of the maturation process for these young men. As such the migration was often assisted by siblings, uncles, mothers or cuates (male friends). By contrast, single women's migration typically occurs from weakly bounded families which lack a strong patriarchal authority figure. Hondagneu-Sotelo found that many women who migrated as young and single came from family situations in which the father was absent. A hierarchy emerged from her research in which fathers had a particular authority over daughters and an authority which was not to be challenged by young women in the same ways that it was tolerated and even sanctioned for young men. As such, even though the numbers of women migrating have been expanding, many more of the women who migrate have a husband whom they join, or migrate as part of a family unit, rather than as single and unmarried. Hondagneu-Sotelo argues that for this sending community, patriarchal authority plays a clear and distinctive role in the mobility of women as compared to men. This authority is exercised over women through arguments about morality, sexuality and the need for male protection of women migrants. This authority, however, is not unambiguous or unchanging. Hondagneu-Sotelo's research also demonstrates that gender relations are contradictory and complex by citing examples of how husbands' departures rearranged gender relations in origin households. With the absence of menfolk from a household, women's work and responsibilities changed and in some cases this added to these women's repertoires and sense of power within their households. In both Hondagneu-Sotelo's and also other origin area studies in Mexico (Gonzalez de la Rocha, 1994), women began to identify themselves as mujer fuerte (strong woman) and this was symbolic of them becoming much more independent of patriarchal control. In another case the malleability of strong patriarchal control is demonstrated (Hondagneu-Sotelo, 1994: 91). A young woman who wanted to migrate to the USA, but whose father would not allow it, threatened to marry a man whom her father despised if she was not allowed to migrate. In this case she successfully used this threat to marry with all its connotations of changing her life around that decision as leverage for migration. In doing so, she manipulated her father's strong patriarchal control over her life to facilitate her migration. These case studies raise intriguing questions for a feminist migration research agenda. Each of these examples provides some insights on how case-study approaches expose the relentlessly fluid and complex nature of household divisions of labour, gender roles and

Victoria A. Lawson 47
ideologies. The above cases illustrate most vividly the fluidities produced through broader political economic forces which reverberate in communities and households in particular places. These case studies also suggest that struggles over access to both material and discursive power rework household divisions of labour and gendered mobility. These two themes in combination must be part of the future research agenda. A crucial lesson from these case studies is that the static categories of `women' and `men' do not reveal the causes of diverse mobility patterns, but rather, that we must critically examine people's positions within their household divisions of labour to understand fully the power relations fomenting mobility and its potential consequences for migrants themselves, and for their origin households. Further, these processes are context specific and will also vary according to household form (female headed, male headed, extended, nuclear, etc.). The case study by Hondagneu-Sotelo suggests that household dynamics differ for male versus female headed households and future research needs to take seriously the nature of household composition in seeking to understand the dynamics and consequences of gendered migration.

IV

Destination household division of labour and the consequences of migration

The dynamics of gender and the inner workings of migrant households also affect the consequences of migration. I now turn to two case studies that reveal some of the ways in which mobility reworks (or reinforces) gender divisions of labour, gender identities and access to material and discursive power for various household members. In particular, these case studies illustrate powerfully how women migrants' different positions within households, and at different moments in their life course, shape their access to paid work in their destinations. I focus on the ways in which cultural constructions of femininity reflected here in women's work shape women migrants' access to material resources and to the conditions under which they labour. Through this inquiry, I seek to broaden the focus of migration research to reveal the operation of gendered power relations in social institutions such as the labour market and to integrate these insights with those that examine household dynamics in destinations. Domestic work, perhaps more than any other destination job, illuminates the complex interactions between household divisions of labour, kinship arrangements and experiences of migration. Eaton (1992) notes that in Brazil, a large majority of in-migrant women are first employed as domestics. Similarly studies by Jelin (1977), Orlansky and Dubrovsky (1978) and Gill (1988) found domestic work to be the most common form of employment for newly arrived women in-migrants a job that has complex consequences for women. On one hand, host families may assist domestic workers to attend school and train for other opportunities, yet on the other, domestic work has consequences for migrants themselves, being low paid and often involving sexual abuse (Radcliffe, 1990). Looking at the prevalence of domestic work (which is in many ways the quintessential example of the feminization of women's paid work in destinations) poses questions about the material consequences of migration and about the gender segmentation of destination labour markets since the most readily available jobs for in-migrant women are clear extensions of domestic, caretaking and mothering roles. This segmentation of paid work has been extensively documented in a vast literature (see Nash and Safa, 1986; Massey, 1994; Hanson and Pratt, 1995, for summaries) and I want to focus here on how domestic work in particular resituates migrant women within destination

48 Hierarchical households and gendered migration in Latin America

households in contradictory ways or, put another way, about how mobility reworks gender ideologies and identities. Radcliffe's (1990) study of Peruvian domestics brilliantly draws out the intersections of household and labour market dynamics that result for in-migrants, indicating that domestic workers live and work in situations shaped by class and ethnic differences between themselves and their employers. Approximately half of all Peruvian domestics in Lima are peasant women from rural areas and Indian communities who become partially assimilated into mestizo culture in cities in ways which devalue peasant culture and identity. This process of assimilation is partial and contradictory, with peasant women dropping mannerisms and dress which mark them as peasant while, at the same time, employer families maintain linguistic and status barriers between themselves and their domestics (Radcliffe, 1990: 38486). Quasi-kinship relations also develop in which domestics are often included in the family, but are treated as children with two effects. On one hand, it denies the nature of their relationship to the family as employee with certain rights and responsibilities; on the other, these quasi-kinship relations involve strong patriarchal control over the `daughter's' movements and personal freedoms. Radcliffe's (1990) study demonstrates the complex and often ambiguous impacts of migration for women. Migrants are sometimes able to break out of domestic work, depending on the nature of their relation with destination families who may even assist in a migrant gaining education or a better job, or facilitate a domestic's involvement with a union organization. Other women, however, replace their role as daughter in origins with the concomitant implications of patriarchal control over their activities and identities for a similar situation in destinations. Turning away from domestics and towards manufacturing work still reveals the complex ways in which gender identities shape women's paid work after migration. Research by Lawson (1995; 1997) on women's garment work in Quito, Ecuador, found that women with substantial domestic responsibilities (often married women with young children) typically worked in different sites and under different employment relations from others in the industry. Ecuador's experience of neoliberal restructuring in the late 1980s and early 1990s has reverberated into the garment industry, sending substantial volumes of work to women in unregulated production. Despite these trends, women's access to this work is typically shaped by their domestic situation. Ludmilla's story provides an intimate view of the ways in which the workings of household gender relations and ideologies shaped her access to paid work in Quito after migration. In this case mobility has had little impact on Ludmilla's position within her natal or marriage households and her experiences of migration have been profoundly shaped by patriarchal gender ideologies. Ludmilla's experiences are overwhelmingly shaped by cultural constructions of wife, daughter-in-law, sister and mother in the Ecuadorian context. Ludmilla, a married woman of 29 years, migrated to Quito with her older brother at the age of 13. Their migration from a rural province was to allow him to attend university and become a professional:
We are eight brothers and sisters, they educated my eldest brother, he is an architect. That is life, each of us has our own destiny and mine was to go out, I was sent out [to work] to help in order that they [her parents] could give financial support to my brother (personal interview by author, 1992).

Ludmilla worked a double day of paid and domestic work in order to care for, and support financially, her brother as he pursued his degree. He is now an architect while Ludmilla operates a small, informal garment business which generates little income insufficient for her to be financially self-reliant. This arrangement is driven by gender

Victoria A. Lawson 49
ideologies in which women's social status is seen to be attained through marriage whereas sons must succeed and bring their own families to the highest social status they can attain. This gendered argument justifies not continuing a daughter's education in order to facilitate (or support directly) the success of a son. This case dramatically illustrates the ways in which the material consequences of migration can be powerfully gendered. Gender dynamics and ideologies continued to shape Ludmilla's destination experiences. Upon her marriage, Ludmilla moved into a household where, in her words, she has three husbands her spouse, her father-in-law and her brother-in-law, as well as her young daughter. Ludmilla is solely responsible for all the domestic work in this household and more. She commented regretfully in our interview that `I always thought when I was single that I would never go and live with my in-laws because I saw many problems amongst my friends in their marriages, and now look!' (personal interview by author, 1992). In addition, her business is located behind her father-in-law's candy store; she takes care of her child during the day, cooks lunches for everyone and runs errands for her father-in-law and her husband. When I asked her about the success of her business and her possibilities for making sufficient income she said:
he [her husband] prefers me to have my business here because this way I can care for my daughter . . . I have to look after my daughter, I have to cook, I have to go to the bank, and I have to deliver candies. Yes, at the moment I would prefer to work in another place because we lack the money to invest in this [business] and I would be able to work more (personal interview by author, 1992, emphasis added).

Extended family obligations, her roles as wife, mother and daughter-in-law, and her stage in the life course dramatically shape her access to paid work. In Ludmilla's case, household gender relations are central to understanding her postmigration experiences in the workforce. Indeed, many studies have found that women (with young children) who migrate to Latin American cities engage in part-time and/or informal jobs that allow them flexibility to care for their households (Beneria and Roldan, 1987; Chant and Brydon, 1989; Teltscher, 1993; Gonzalez de la Rocha, 1994). While these jobs may have advantages in terms of allowing women to balance roles, these jobs are often unstable, poorly paid and without benefits, such that they have clear negative implications for these women in terms of the material consequences of migration.

Concluding comments

This review essay has brought together empirical research on gendered migration in Latin America in order to extend our theoretical understandings of the consequences of migration. Current empirical research from Latin America highlights the importance of analysing gendered migration in the diverse contexts of intrahousehold dynamics, household divisions of labour and migrants' access to material and discursive power within this region. The case studies discussed here demonstrate that these expressions of gender difference (divisions of labour, access to household resources, gender identities) are themselves fluid and complex transforming in the context of shifting politicaleconomic and cultural forces impinging upon places and peoples. In order to answer the classic migration questions, then (for example, who migrates and why, what are the material consequences of migration, and how does mobility rework gender divisions of labour and gender ideologies and identities?), we must examine how household divisions of labour, gender identities and ideologies are being reworked in particular

50 Hierarchical households and gendered migration in Latin America

places and in the context of migration. Put another way, the consequences of gendered migration can only be understood in the context of the culturally produced causes of that migration. A key finding from this review essay is that feminist approaches to understanding the causes of migration involve analysis of how particular gender roles, ideologies, and identities (e.g., of mother, daughter, etc.) and household divisions of labour are historically and contextually produced and transformed. The various case studies described here detail the ways in which broader political-economic transformations reverberate in households to foment migration and to shape the consequences of mobility as these intersect with gender relations in households and labour markets. The case studies of rural out-migration in Peru and Bolivia demonstrate how, as peasant households have been unable to maintain historically important forms of production under commercialization, this has caused daughters to become migratory in new ways. In addition, the case study of MexicoUSA migration illustrates how the long-term out-migration of men from sending communities transforms women's gender roles and identities in origin places. Each of these case studies illustrates that household divisions of labour and gender identities and ideologies are fluid systems that interact with migration in complex and contextual ways. To understand gendered migration fully, then, research needs to take seriously the ways in which systems of gender transform, and are transformed by, mobility. I have also argued that the types of jobs into which women and men enter in their destinations have implications for household divisions of labour and gender identities among migrants. At the same time, impacts on migrants are always complex and ambiguous for example, migrant women may gain new economic independence through their move and this may change their gendered relations to other household members (i.e., evidence from Eastmond, 1993, and Hondagneu-Sotelo, 1994, on immigrant Mexican and Chilean women into the USA). Or, as in Ludmilla's story, migration may serve to reinforce patriarchal relations of control over women's labour and gender roles. This evidence prompts a range of research questions for Latin Americanist migration studies. Specifically, how does migration (despite labour market discrimination) affect patterns of paid work for women and men migrants? Are there significant differences in the work histories (in terms of status of jobs, pay level improvements, etc.) for women and men migrants? What reverberations on to household gender relations and identities emerge from women migrants' entry into Latin American urban labour markets?

Acknowledgements
This article is part of a larger project in which I am collaborating with Andrew Morrison of the International Development Bank and Richard Bilsborrow of the University of North Carolina. The project is funded by the National Science Foundation under grant SBR-9511129. Support from the NSF is greatly appreciated and input from my collaborators has been extremely valuable in the development of my thinking. I also gratefully acknowledge invaluable comments received from faculty at the Newcombe College for Research on Women at Tulane University, from my colleagues Lucy Jarosz and Rachel Silvey and from the anonymous reviewers.

Victoria A. Lawson 51

Notes
1. Drawing on an extensive literature, we define household to be an income-pooling unit that may or may not include kinship relations (for an extended treatment of the concept, see Bruce, 1989; Chant, 1992; Hart, 1992; Smith and Wallerstein, 1992). 2. The migration literature is vast and diverse and is impossible to review in depth within the space constraints of this article. Please refer to the references cited as more comprehensive examples of each perspective.

References
Allen, S. and Wolkowitz, C. 1987: Homeworking: myths and realities. Basingstoke: Macmillan Education. Arguello, O. 1981: Estrategias de Supervivencia: Un Concepto en Busca de su Contenido. Demografia y Economica 15, 190203. Beneria, L. and Feldman, S., editors. 1992: Unequal burden. Economic crises, persistent poverty, and women's work. Boulder, CO: Westview Press. Beneria, L. and Roldan, M. 1987: The crossroads of class and gender. Chicago, IL: University of Chicago Press. Bilsborrow, R. 1993: The feminization of internal migration in developing countries. New York: United Nations. Bilsborrow, R. and Peek, P. 1992: Peasant farming and labour migration: evidence from the Ecuadorian highlands. Paper presented at the IUSSP conference on `The Peopling of the Americas', Verzcruz, Mexico. Brown, L. and Lawson, V. 1985: Migration in third world settings, uneven development, and conventional modeling: a case Study of Costa Rica. Annals of the Association of American Geographers 75, 2947. Bruce, J. 1989: Homes divided. World Development 17, 97991. Carney, J. and Watts, J. 1990: Manufacturing dissent: work, gender and the politics of meaning in a peasant society. Africa 60, 20741. Chant, S., editor. 1992: Gender and migration in developing countries. London: Belhaven Press. Chant, S. and Brydon, L. 1989: Women in the third world: gender issues in rural and urban areas. New Brunswick, NJ: Rutgers University Press. Chant, S. and Radcliffe, S. 1992: Migration and development: the importance of gender. In Chant, S., editor, Gender and migration in developing countries, London: Belhaven Press, 129. Collins, J. 1986: The household relations of production in southern Peru. Comparative Studies in Society and History 28, 65171. Corden, W.M. and Findlay, R. 1975: Urban unemployment, intersectoral capital mobility and development policy. Economica 42, 14357. Eastmond, M. 1993: Reconstructing life: Chilean refugee women and the dilemmas of exile. In Buijs, G., editors, Migrant women. Crossing boundaries and changing identities, Oxford: Berg Publishers, 3554. Eaton, D. 1992: Rural-urban migration and underemployment among females in the Brazilian northeast. Journal of Economic Issues XXVI, 38595. Folbre, N. 1986: Cleaning house: new perspectives on households and economic development. Journal of Development Economics 22, 540. 1988: The black four of hearts: towards a new paradigm of household economics. In Dwyer, D. and Bruce, J., editors, A home divided: women and income in the third world, Stanford, CA: Stanford University Press, 24889. 1991: Women on their own: global patterns of female headship. In Gallin, R. and Ferguson, A., editors, The women and international development annual. Vol. 2, Boulder, CO: Westview Press, 89126. Fortuna, J.C. and Prates, S. 1989: Informal sector versus informalized labor relations in Uruguay. In Portes, A., Castells, M. and Benton, L., editors, The informal economy: studies in advanced and less developed countries, Baltimore, MD: Johns Hopkins University Press, 7894. Gilbertson, G. and Gurak, D. 1992: Household transitions in the migration of Dominicans and Colombians to New York. International Migration Review 26, 2245. Gill, L. 1988: Senora y Servientes: women and domestic service in La Paz, Bolivia. Paper presented at the XIV international congress of the Latin American Studies Association, New Orleans.

52 Hierarchical households and gendered migration in Latin America

Gisbert, M.E., Painter, M. and Quiton, M. 1994: Gender issues associated with labor migration and dependence on off-farm income in rural Bolivia. Human Organization 53, 11012. Gonzalez de la Rocha, M. 1994: The resources of poverty: women and survival in a Mexican city. Cambridge, MA: Blackwell. Greenwood, M.J. 1975: Research on internal migration in the United States: a survey. Journal of Economic Literature 13, 397433. Hanson, S. and Pratt, G. 1995: Gender, work, and space. London: Routledge. Hart, G. 1992: Imagined unities: constructions of the `Household' in economic theory. In Ortiz, S. and Lees, S., editors, Understanding economic process, Lanham, MD: University Press of America, 11130. Hondagneu-Sotelo, P. 1994: Gendered transitions: Mexican experiences of immigration. Berkeley, CA: University of California Press. Jelin, E. 1977: Migration and labor force participation of Latin American women: the domestic servants in the cities. Signs 3, 12941. Katz, E. and Stark, O. 1986: Labor migration and risk aversion. Journal of Labor Economics 4, 13449. Lauby, J. and Stark, O. 1988: Individual migration as a family strategy: young women in the Philippines. Population Studies 4, 47386. Lawson, V. 1995: Beyond the firm: restructuring gender divisions of labor in Quito's garment industry under austerity. Environment and Planning D: Society and Space 13, 41544. 1997: Tailoring is a profession, seamstressing is just work: resiting work and gendered identities among artisnal garment workers in Quito. Working paper. Seattle, WA: Department of Geography, University of Washington. Lowry, I.A. 1966: Migration and metropolitan growth: two analytical models. San Francisco, CA: Chandler. Lucas, R.E.B. and Stark, O. 1988: Migration, remittances and the family. Economic Development and Cultural Change 36, 46582. Manser, M. and Brown, M. 1980: Marriage and household decision making: a bargaining analysis. International Economic Review 21, 3144. Massey, D. 1994: Space, place, and gender. Minneapolis, MN: University of Minnesota Press. Mies, M. 1986: Patriarchy and accumulation on a world scale: women in the international division of labor. Atlantic Heights, NJ: Zed Books Mincer, J. 1978: Family migration decisions. Journal of Political Economy 88, 74973. Momsen, J. 1992: Gender selectivity in Caribbean migration. In Chant, S., editors, Gender and migration in developing countries, London: Belhaven Press, 7390. Mountz, A. and Wright, R. 1996: Daily life in the transnational migrant communty of San Agustin, Oaxaca and Poughkeepsie, New York. Paper presented at the University of Washington, February. Nash, J. and Fernandez-Kelly. M.P. 1983: Women, men, and the international division of labor. Buffalo, NY: SUNY Press. Nash, J. and Safa, H. 1986: Women and change in Latin America. New York: Bergin & Garvey. North American Commission for Latin America 1993: NACLA. Report on the Americas 27, 16. Orlansky, D. and Dubrovsky, S. 1978: The effects of rural-urban migration on women's role and status in Latin America. Paris: UNESCO. Radcliffe, S. 1990: Ethnicity, patriarchy and incorporation into the nation. Society and Space 8, 37993. 1991: The role of gender in peasant migration: conceptual issues from the Peruvian Andes. Review of Radical Political Economics 23, 12947. Roberts, B. 1978: Cities of peasants: the politicaleconomy of urbanization in the third world. London: Edward Arnold. Safa, H. and Antrobus, P. 1992: Women and economic crisis in the Caribbean. In Beneria, L. and Feldman, S., editors, Unequal burden. Economic crises, persistent poverty, and women's work, Boulder, CO: Westview Press, 4982. Sen, A. 1990: Gender and cooperative conflicts. In Tinker, I., editors, Persistent inequalities: women and world development, New York: Oxford University Press, 12349. Shrethsa, N. 1988: A structural perspective on labour migration in underdeveloped countries. Progress in Human Geography 12, 179207. Smith, J. and Wallerstein, I. 1992: Creating and transforming households: the constraints of the world economy. Paris: Cambridge University Press. Staeheli, L. and Lawson, V. 1995: Feminism, praxis and human geography. Geographical Analysis 27, 31238. Teltscher, S. 1993: Informal trading in Quito, Ecuador. Saarbruken: Verlag Breitenbach Publishers. Todaro, M. 1969: A model of labor migration and unemployment in less developed countries. American Economic Review 59, 13848.

Victoria A. Lawson 53
Westwood, S. and Bhachu, P. 1988: Enterprising women: ethnicity, economy and gender relations. London: Routledge. Wolf, D. 1992: Factory daughters. Berkeley, CA: Unversity of Calfornia Press. Young, K. 1980: The creation of a relative surplus population: a case study from southern Mexico. Society of Latin Amercan Studies Bulletin 32, 6189.