In the mid-l970s, Latin Americanist geographers joined other social scientists in producing a steadily growing current of research about women, although with somewhat less visibility than other social scientists. Given this chronology of events, it seems appropriate at this time to summarize the main themes and findings of studies about contemporary Latin American women published by geographers during the past decade and to discuss their implications for social science research in general. Because the number of such studies is quite small, the findings of each are described briefly here. It is hoped that these summaries will provoke thoughtful inquiry by other geographers along similar lines in their own work.

FEMINISM AND THE GEOGRAPHY OF WOMEN

Geographic studies of Latin American women have been undertaken from both feminist and non-feminist perspectives. However, feminism appears to have had only limited impact on geographers, both Latin Americanists and those of other specialties. A principal view of feminist philosophy is that the relationship between men and women, and the impact of that relationship on women's lives should be central objectives of investigation in social science research. For geographers, the spatial variations in this impact and in women's status and behavior are of particular importance. Feminist theory reminds researchers of certain essential questions to ask, by showing that women and men behave in different ways with results that are of potentially equal significance.

In contrast to this view, feminists argue that studying the geography of women merely by gathering and analyzing data without changing existing theory cannot be very productive (Women and Geography Study Group 1984). However, some feminists might argue that, at the very least, even non-feminist researchers should pay special attention to women in their investigations, if only for the sake of accuracy.

For example, when rural migrants are studied as a single homogeneous group, a great deal of inaccuracy can result in explaining or predicting their behavior. As a result, Latin Americanist geographers often desegregate migrant groups by educational background, age or location of origin, and they should also be disaggregated by sex. The same holds true for landowners, farmers, industrial workers and household heads. Although this approach can produce a "geography of women" which is distinct from a feminist geography (Women and Geography Study Group 1984), it does provide a base of information that can elucidate theory, and appears to be the most popular approach among Latin Americanist geographers studying women.

Geographers are producing a slow but steady stream of empirical work on Latin American women that should be of interest to anyone committed to accurate study of Latin America. The themes and methods of this work, coming mostly from British geographers, complement those of a larger body of empirical and theoretical work on Latin American women, although it is often emphasized that this kind of research does not fall neatly within disciplinary boundaries. In a most valuable annotated bibliography of work on Latin American women, Stoner (l989) explains that current studies focus on topics such as health, demography and urban and rural development rather than disciplines. Geographers, who have long prided themselves on reaching across disciplinary boundaries, should find themselves working comfortably with this approach. Accordingly, the geographic studies on Latin American women reviewed here are discussed in the context of non-geographic studies as a way of pointing to future research directions.

LATIN AMERICAN WOMEN AND POPULATION GEOGRAPHY

Rural-Urban Migration

A recent growing concern among Latin Americanist geographers is to differentiate between migration patterns of men and women in Latin America (Brydon and Chant 1989). Thomas and Muñiz (1981) called for a wider understanding of women's roles in migration, especially since it is now clear that women generally outnumber men in rural-to-urban migration, resulting in a growing gap in the sex ratios of rural and urban areas. Two geographic case studies show that the causes, destinations and consequences of migration by women can vary in crucially important ways from those characterizing migration by men.

Monk (1981) found that female migrants in Puerto Rico quite often migrated short distances within Puerto Rico, either to accompany their husbands or to marry, whereas male migrants were usually single and looking for employment, often in the United States. Rengert (1981) points out that Monk's results may be atypical because Puerto Rican women have access to economic resources such as welfare payments and factory work that are generally unavailable to rural women of other Latin American countries. According to Rengert, rural women of Mexico are likely to migrate if they come from large families in extreme poverty, and if they can become domestic servants as a way of providing themselves with transportation to a town or city and shelter, food and clothing upon arrival. Mexican men, however, have greater access to agricultural employment in their home villages (jobs that are unavailable to women), and male migrants tend to come from smaller families that apparently invest some of their resources in sending a son to work in agriculture in another rural district. Thus women and men respond differently to "push" factors of rural poverty.

Distinguishing between motives, origins and destinations of men and women who migrate is a neglected approach (Crummett 1987). This is surprising, because differences between male and female migrant behavior would seem to be of particular importance for policies aimed at either slowing or re-directing current migration paths in and from Latin America. The implication for those attempting to stem the tide of urban arrivals through rural development or modernization of agriculture is that such programs will probably not significantly reduce the stream of rural-to-urban migration unless they specifically offer women, including single women, viable opportunities to support themselves in the countryside (Deere and León 1987). It is also important to view women and men as members of a household with cultural and economic priorities of its own (Wood 1981). Radcliffe (1986) argues that migration should be understood as part of a household "livelihood strategy" as well as an individual behavior. [end p. 232] These few comments challenge geographers studying migration in Latin America to re-examine their migrant populations for significant patterns of male, female and "household" migration.

Health and Nutrition

Thomas and Muñiz (1981) remarked that Latin Americanist geographers of the l970s were more concerned with migration patterns than with fertility, mortality or other demographic features, and this seems to have been true of the l980s as well. However, sociologists and economists were actively investigating the relationship between fertility and women's employment and education (Davidson 1978; Powell 1976; Youssef 1982). It appears that fertility does not decline simply because the benefits of modernization trickle down to women but rather through programs geared specifically toward women's health (Betancourt 1988) and through the effect of certain types of employment (Smith 1981).

Efforts to tailor fertility or population control programs to the unique situations of women in different places could benefit greatly from studies of women's perceptions and sources of information about health, nutrition and medical care. In rural and in poor or traditional urban households, women are responsible for the family's physical well-being, nutrition and health. How they acquire information about diet or medicine proves to be just as important as the information or technology that is actually available to them. For example, Weil (1988) interviewed Bolivian mothers to identify their sources of information about health and illness and found that mass media and other formal channels of communication were less effective in conveying accurate information about medicine than were recommendations by mothers' female relatives or market traders. Geographers interested in the diffusion of information would find that this path of investigation could yield results with many useful applications.

LATIN AMERICAN WOMEN AND RURAL DEVELOPMENT

Rural Household Economies and Land Use

Sociologists, anthropologists and economists have been quite active in carrying out descriptive studies of women's work in rural areas and in explaining their work in theoretical terms (Deere and León 1987; Nash and Safa l985). Geographers have contributed to this effort by analyzing women and household economies in rural areas, but much remains to be done (Momsen and Townsend 1987).

Perhaps the single most valuable contribution of studies of women's work is the body of accumulated evidence showing that a significant twofold division of labor exists within households, a division between paid labor and unpaid labor, both of which contribute to household, village and regional economies, and a division between men's work and women's work. These divisions correspond roughly to the distinction often made by geographers between the public sphere, where men's work is performed, and the private sphere of women's work. A great amount of evidence has been collected to show that households in rural Latin America respond to a full array of economic and social opportunities available to both female and male household members rather than merely depending on wages or cash income (Benería and Sen 1981; Deere 1983; Deere and León 1987). Thus, the behavior of households can thwart or encourage rural development in ways that are simply overlooked if the distinction between different types of labor within households is not examined systematically.

Momsen (1987) has determined that small Caribbean farms owned or managed by women differ significantly from those operated by men. In Afro-Caribbean islands an especially high percentage of women are farmers (Henshall 1981). Conditions vary greatly from island to island, but women's farms tend to be smaller, on less fertile soils and used for producing subsistence crops or small livestock for household consumption. Men's farms are somewhat larger on average, have better soils, and produce export crops and cattle. To study or design policies for "the Caribbean small farmer" could therefore be quite misleading.

Momsen (1986) points out that many female farmers are household heads whose spouses have migrated abroad. On some islands, women traditionally do not till land and therefore cannot grow crops without men's labor. In other areas, women do perform some farming along with other chores but manage to produce only meager amounts of subsistence crops because they lack the labor of another adult.

Momsen's work helps explain, in part, two problems in the Caribbean that have attracted other scholars' attention. First is the thorny issue of why remittances from emigrant spouses do not produce much rural [end p. 233] development. Return migrants may use their wages to buy more land or to build a house for retirement but while they are away the remittances they send home to their families are seldom used for productive investment. In farming households where a male adult has emigrated, women recipients may spend remittances on temporary hired workers in order to maintain farm production levels. In other cases, cash is spent on food to replace what the family used to grow at home, and on consumer goods, children's clothes and medicines and other expenditures (Griffith 1985; Momsen 1986; Rubenstein 1983). The second problem is that the "feminization of agriculture" in the Caribbean has led to declines in farm productivity and food production (Momsen 1987; Chaney 1983). The future of domestic food supplies is thus in jeopardy in the Caribbean. As food producers, women are at the heart of this future.

In contrast to Momsen's study of a region where land use patterns have long been undergoing change, Townsend and Wilson de Acosta's study (1987) of colonization of Colombian rainforest examines how women and men adapt to a frontier economy. They found that men far outnumber women in a frontier village but not because few women move into the area. Rather, young girls growing up in the frontier migrate to cities rather than marrying locally and forming new households. The reasons for heavy female out-migration are both economic (women traditionally do not work in agriculture on the frontier and there are few alternative ways for women to earn wages) and social (frontier women have such heavy domestic responsibilities in caring for their own families and the household's hired laborers that even men help with some chores such as washing clothes, fetching water or cooking). This is another excellent example of the importance of examining the division of labor within households and how men and women respond differently to changing conditions in order to support themselves. This example also suggests a new line of inquiry for geographers working on frontier colonization.

Women's Role in Managing the Environment

Policies and projects designed to exploit or protect natural resources can flounder if women's activities in using resources are not fully understood and addressed. Again, examining the labor of individuals within rural households proves useful, if only because many such activities by women are not listed or defined in most labor statistics and are not taken into account in macro-level analyses.

For example, Bagchi (1987) summarizes a number of studies linking use of firewood to deforestation and points out the importance of examining the labor responsibilities of individual household members. As the Third World's supply of firewood dwindles relative to demand, the consequences include changes in cooking practices and diet, use of manure and crop waste for fuel instead of as fertilizer and increased time and labor for gathering firewood, usually a task performed by women and children. When women divert their time and labor from cooking, child care, food and craft production, market trade and health care, the entire household suffers. Bagchi suggests that agroforestry geared toward women might be a practical solution to this problem, if the tools, types of trees, storage methods and other essentials are viable for women to obtain, maintain and utilize. Women's day-to-day strategies for managing water, firewood, land, forests and fisheries are of crucial importance in formulating environmental and economic policy for rural development, yet very little is known about them.

How do geographic studies of women in rural development differ from and complement studies from other disciplines? From the work of anthropologists, sociologists and economists a strong body of feminist and Marxist theory has emerged to explain the findings of empirical studies (Cancian et al. 1978; Deere and León 1987; Nash and Safa 1985). This theory emphasizes the changes brought about by the introduction of capitalism into agricultural zones. Wage-paying agricultural jobs that usually go to men, channelling legal land titles, credit, technology and tools to men, and the availability of manufactured goods that are cheaper than local products all undermine or displace women's usual sources of income or subsistence. It is increasingly difficult for them to continue supplementing men's contributions to household income, much less support the households they head on their own as daughters, widows or single mothers (Rossini 1983). Thus many rural women give up and move to cities where they hope to find better opportunities. Studies of people and places by geographers have added weight to these arguments.

It is difficult to find many differences between geographers' empirical studies and those by other social scientists because all are still in the early stages of gathering basic facts and information. The very fact that geographic works have much in common with non-geographic works suggests that disciplinary boundaries are being crossed with great frequency, and that much could be learned from exchanging ideas and information. However, geographers' studies are seldom discussed or even cited in work by other social scientists on women and [end p. 234] development. In part this may be due to the fact that there are so few geographic studies about Latin American women and they are generally published in journals or proceedings that are apparently not routinely consulted by non-geographers. Considering the growing unease about the status of regional geography and the shrinking pool of Latin Americanist geographers, it might be worthwhile for geographers to make a special effort to publish their work in journals outside geography.

LATIN AMERICAN WOMEN AND URBANIZATION

Urbanization and Industrialization

As in the study of women in rural development, pre-l970 discussions of women in the processes of urbanization and industrialization suffered from a lack of terminology to accurately describe and evaluate women's work in cities. A great deal of discussion has developed around appropriate definitions of women's work, formal work and informal work, the merits of distinguishing between them and the importance of the international division of labor in constraining and in providing opportunities for urban women (Townsend and Momsen 1987).

It is generally agreed that women are usually paid less than men for equivalent work in Latin America (Humphrey 1985). But research has revealed that most often women and men do not perform the same work and that they are not simply interchangeable in the labor market, especially as national labor forces become increasingly integrated into a global division of labor. Cunningham (1987) links the changing global geography of manufacturing with conditions of work for individual women and the fates of their families and households. Her work on Brazil's manufacturing sector shows that the demand for women workers shifted abruptly from industry to industry in the l960s and l970s as multinational corporations responded to international variations in wage levels and government policies on trade unions. All other things being equal, cities or countries offering the lowest minimum wages and the least interference from labor unions see the greatest investment in manufacturing by large corporations. This has been particularly true in assembly industries and service industries, which employ the majority of women laborers in the formal sector.

Aside from the influence of the global economy, how and why urban women support households economically is affected by social and cultural factors related to their own households and to society in general. As Chant (1987) shows in her study of shantytown families in Querétaro, Mexico, the likelihood that a woman will engage in paid labor in either the formal or informal sector depends mostly on her marital status, her husband's attitude and the structure of her household. Women who are married and live in extended families are often freed from housework and child care by other adults, and/or forced to supplement their husbands' incomes in order to support extra household members. They, along with single women and female heads of households, are more likely to work than married women in nuclear families, whatever their age or number of children. Chant (1985) points out that rather than improving the lot of women by leading to the formation of nuclear families, the processes of urbanization and industrialization can instead make them and their children more dependent on one adult male's income even when wives desire to work and need the extra income to care for their families. This dependency can jeopardize a family's long-term financial stability, as Christopherson (1983) points out. Her study of Ciudad Juárez shows that many families respond to rapidly changing employment conditions by sending as many household members to work in as many different jobs as possible. Clearly, the unique cultural and economic contexts of specific places need to be considered when analyzing why and how urban women work.

Chant (1984) and Chant and Ward (1987) urge policymakers to adjust housing policy in Latin American cities where extended families are common and to accommodate or even encourage large households by providing for larger lots, flexible building construction and appropriate legal mechanisms for land ownership, among other suggestions. Geographers writing about these issues repeatedly call for more studies and more information about women, families and households in order to inform urban housing and economic policy.

The Informal Sector

It is well-documented that many Latin American urban women support or contribute to their households through work in the informal sector, especially as market traders and domestic servants (Babb 1989; Bunster and Chaney 1985). Latin Americanist geographers have produced many studies of periodic markets but few of them discuss market traders in detail (Jones 1982; Stadel and del Alba Moya 1988).

Benton (1987) found that campesina women in the Lake Titicaca area benefitted greatly from opportunities to earn cash through trade as their communities acquired greater educational opportunities, improved roads that reduced travel time to La Paz and a boom in the tourist trade. Women in lakeside villages now run shops and sell crafts to tourists, operate businesses selling fish in distant towns and villages and raise pedigree pigs for sale. Their educational and economic opportunities and increasingly frequent contacts with storekeepers, relatives and market vendors in La Paz have helped them take full advantage of their communities' modernization. But men have fallen relatively behind in earnings and in social status, having become increasingly dependent on women's incomes and initiative. One might ask what future problems this dependence could create.

Women recently involved in trade and business in Lake Titicaca seem to have avoided the problems of market trade in other Latin American cities that make it difficult for women to earn enough to support their families. In Cajamarca, Peru (Greenow and Muñiz 1988) and in other cities (Babb 1989) women traders in public markets pay credit charges, taxes and other costs that make it difficult to expand or, in some cases even maintain, their trading businesses. The organization of trade allows wholesalers to charge whatever they can while retail prices are controlled, canceling out any gains that traders could make by increasing their sales. If wholesalers have a similar effect on agricultural producers, the geographic implications of this problem are enormous. Poor women of small towns, especially those who are the principal source of support for their families, add significantly to the rural-urban migration stream as they search for stable incomes in the few types of employment that are feasible for those with children to care for and no skills for formal employment.

CONCLUSION

It is clear from these studies that it is not only necessary to distinguish between men's and women's experiences and needs as change occurs in Latin America. It is also necessary to distinguish among the many household, cultural and economic settings in which Latin American women operate (Henshall 1984). The need for expertise in working at different scales of analysis and in assessing the importance of location in women's behavior should place geographers at the forefront of this area of research. By asking appropriate questions geographers can continue to contribute to the growing stream of work on Latin America women. In particular, taking into account the differences between women's and men's migration patterns, and considering the role women play in affecting demographic trends as providers of food and medical care are two viewpoints that can enhance the kinds of work already being done by Latin Americanist population geographers. Asking specific questions about women in subsistence, commercial and frontier agriculture could explain in greater detail patterns of land use and environmental changes. And linking urban women's work patterns with household economies and housing or transportation needs appears to be a natural next step for Latin Americanist urban geographers. Above all, geographers must make full use of the wide body of literature already available from other disciplines and ensure that geographic viewpoints are disseminated to other Latin American specialists.

References