Current trends affecting women’s participation in the rural economy

Several concurrent and interrelated changes in the social and economic conditions facing rural Latin Americans help to explain the gender-specific shifts in population and labour force participation described in Section 2. We will examine in turn the effects of changes in female education, fertility and household formation; internal and international migration; agricultural liberalization and its impact on subsistence food production and the promotion of non-traditional agricultural exports.

Education, fertility and household formation

One of the most significant advances in gender equity in Latin America has been in the area of women’s education. For the region as a whole (including the Caribbean), female primary enrolment ratios are approximately 94 percent that of male ratios; women actually outnumber men in secondary education and they are near parity at the tertiary level as well (UNESCO, 2000). As might be expected, however, the absolute gains have been less for rural women than for urban women. For 13 Latin American countries that collected data disaggregated by gender and by geographic zone (urban/rural), the average level of educational attainment for urban women between the ages of 15 and 24 was nine years (the equivalent of completing secondary school), while for rural women it was six (the equivalent of completing primary school) (CEPAL, 2001).^[8]

Table 6 shows the average number of years of education for rural women and men by age group and by economic activity. Several trends bear mention. First, the educational levels of both rural men and women have consistently risen since 1980 - for women, on the order of a year or more on average (CEPAL, 2001). Second, in 10 of the 13 countries, younger rural women (15-24 years old) have higher levels of education than their male counterparts. Older women (25-59 years old) in about half of the countries are still less educated than their male peers, demonstrating a clear generation gap in rural women’s educational opportunities. Finally, economically active rural women are clearly at the upper levels of the female educational spectrum and in almost all cases have significantly more years of schooling than the male EAP. It appears, therefore, that rural women across the region are acquiring higher levels of basic human capital (including literacy and numeracy) and using these skills at least in part to enter the rural workforce in greater and greater numbers.

TABLE 6
Latin America: Average rural levels of education, by sex, age and economic activity

^a/ 1998
^b/ 1997
^c/ 1994

Source: CEPAL (2001).

The higher rates of education and economic activity for rural women are directly linked to the falling total fertility rates of every country in Latin America. For the region as a whole, the average number of children per woman has fallen from 6.0 in 1960 to 2.7 in 2000, and is expected to reach the replacement level (2.1) in 2025. As would be expected, there are fairly wide disparities both across countries and between urban and rural areas within countries. Bolivia, Guatemala, Honduras, Nicaragua and Paraguay, for example, all still have total fertility rates over four (CELADE/CEPAL, 2001). The most recent round of Demographic and Health Surveys also indicate significantly higher rates of fertility in rural as compared to urban areas (Table 7), although these are also falling over time. As rural families have fewer children, dependency burdens are reduced, and women are more able to engage in both agricultural work and off-farm employment.

TABLE 7
Latin America: Total fertility rates by geographic zone

Source: Demographic and Health Surveys.

Also associated with this demographic transition are changes in the patterns of household formation in rural Latin America: later marriages, greater incidence of informal consensual unions and, of particular interest here, increasing rates of rural female household headship. Rising rates of rural female family and household headship imply greater levels of economic responsibility and therefore higher rates of activity in the rural economy. Table 8 shows the most recent data on self-reported female headship, as well as households in which a woman is the primary income earner. For the 13 countries with information disaggregated by urban and rural areas, the unweighted mean rate of self-reported rural female headship is 18 percent, with a low of 13 percent in Brazil and a high of 25 percent in El Salvador. Using an alternative definition of headship, the data indicate that women in almost 23 percent of rural households are the principal economic providers.

TABLE 8
Latin America: Female household headship, 1999

^a/ 1998
^b/ 1997

Source: CEPAL (2002b).

While these rates are lower than those reported for urban areas, the true incidence of female headship may be substantially higher, due in part to the transitory nature of many informal male-female unions and also to the widespread existence of “nested” households (female-headed families residing within larger households). The former phenomenon implies that while at any given time a household may have an adult male residing within it (who is most often registered as the “head” in household surveys and censuses), the mother-child(ren) unit is the more stable formation. The issue of household nesting is explored in a rural Honduran context by Bradshaw (1995), who argues that potential female heads who are unable to survive on their own create “subfamilies” within extended households. These can take the form of adult daughters and their children residing with their own parents, or female heads living with members of their own generation (e.g. siblings). In Bradshaw’s rural Honduran sample, 11 percent of all households were made up of multiple sub-families, and the average duration of co-residence in these types of households (seven years) suggests that they are relatively stable formations.^[9]

Migration

A second important trend affecting the role of women in the rural economy is the migration of large numbers of men and women out of rural areas, both internally to urban centres and internationally within and outside of Latin America.

Gender-disaggregated data on internal migration are available only for some countries and for selected time periods. Table 9 summarizes the information for the 1960s, 1970s and 1980s, using the census survival ratio method.^[10] In Latin America as a whole, natural urban population growth now generally exceeds net migration, and in most countries for which there are data, women migrate internally in larger numbers than men do - a trend that has held since at least 1960.

What factors determine the internal migration of women, and are they different from those that influence men? Katz (2000), working with 1997 data from Ecuador, finds important gender differences in the interaction of demographic variables with the migration decision and examines the role of rural development policy in altering gender-specific migration flows. Contrary to expectations, high household dependency ratios reduce the probability that young men will migrate, while they have no statistically significant impact on women’s migration propensities.^[11] However, the gender composition of the household does not significantly affect migration probabilities, suggesting that the gender division of labour in highland Ecuador is relatively flexible; women do not necessarily have to be assured of replacement female labour in the household to free them up for migration, and likewise for men. Both male and female migrants are significantly more likely to be married than their non-migrant counterparts, and the effect of marriage on male migration probabilities is especially large in rural areas. A likely scenario here is that marriage is followed by a period of separation in which men “commute” between their community of origin and a destination area with better employment prospects; tied movers.once a “beachhead” is established in the city, wives rejoin their husbands as tied movers.

TABLE 9
Internal migration in Latin America

^a/ Data vary by country: Brazil 1990-1995; Chile 1982-1992; Guatemala 1984-1994; Mexico 1990-1995; Nicaragua 1985-1995; Uruguay 1986-1996.

Sources: United Nations (1996, 2001), Singelmann (1993), CEPAL/Habitat (2001).

Rural development appears to have very uneven effects on the migration behaviour of men and women. With regard to the labour market, the analysis in Katz (2000) suggests that women are almost 30 percent more likely to leave rural areas which offer non-agricultural jobs than from those that do not, implying that the development of off-farm employment opportunities has disproportionately benefited rural men. The effects of crop diversification and irrigation on women’s migration, on the other hand, are strongly negative, which can be explained by the fact that women’s labour contribution to crops such as tomatoes, cucumbers, peppers and tree fruits is relatively high compared to traditional crops like maize, wheat and beans. With regard to the effects of land ownership on migration, larger endowments of privately-owned land are more likely to restrict male migration, while access to common property resources acts as a migration deterrent for women.

International migration patterns are somewhat different from those of internal migration (see Table 10).^[12] Recent analysis of international migration within and from Latin America shows that, between 1970 and 1990, there was a tendency towards the “masculinization” of migration into the United States (made up primarily of Mexicans and Central Americans), but women made up a greater and greater percentage of intra-regional migrants (CELADE/CEPAL, 1999c; Villa and Martínez, 2001). The ratio of male to female Latin American migrants into the United States rose from approximately 90 in 1970 to 110 in 1990. In the case of Mexico, what had been a primarily female northward migration in 1970 strongly reversed itself over the next two decades, so that by 1990 approximately 123 Mexican men were moving to the United States for every 100 women. Similarly, male migrants from El Salvador increased their share relative to women from 68 percent in 1970 to 107 percent in 1990. However, between 1990 and 1996 there was a slight decline in the male to female migration ratio; and the data from the 2000 U.S. Census suggests that, at least for Mexico, while men continue to dominate the migration stream, there is some feminization of the migrant population (US Census Bureau, 2001).

TABLE 10
International migration sex ratios, Mexico and Central America

^a/ Excluding El Salvador for all years except 2000.

Sources: CELADE/CEPAL (1999c), US Census Bureau (2001).

As with internal migration, women are slightly more active than men in international migration within Latin America: for every 100 men that move within the region, 105 women change their country of residence. Some researchers attribute the difference between Latin America-United States migration patterns and those obtaining within the region to the distinct natures of the immigrant labour markets in the destination countries. International labour flows are predominately male into countries with high demand for foreign agricultural workers (such as Mexicans to the United States and Bolivians and Chileans to Argentina), while flows are more likely to include large numbers of women if the demand is primarily in the service sector, including domestic service (such as Colombians to Venezuela, Paraguayans to Argentina and Nicaraguans to Costa Rica) (Villa and Martínez, 2001; Portocarrero, 2001).

The large flows of migrants between Mexico and the United States - and the relatively greater degree of high-quality information about this migrant population - have generated particular interest in the gender-specific determinants and consequences of these international movements. Kanaiaupuni (2000) looked at 14 000 individuals in 43 Mexican communities between 1987 and 1997 to explore gender differences in the roles of human capital, socio-economic status, lifecycle, social networks and local economic opportunities in the decision to migrate. In this large sample, only 7 percent of women, compared to 41 percent of men, had ever migrated to the United States, and 74 percent of women with migrant partners had never migrated at all. Regarding the determinants of migration, she found that higher levels of education increased women’s but decreased men’s chances of migrating; she attributes this to the higher returns to female education in the United States labour market, relative to the Mexican labour market.^[13] Related to this, the likelihood of men migrating significantly decreased when communities had relatively high levels of female employment, implying that women’s local earnings can substitute for potential returns to migration for the household.

Finally, the data suggest that women are considerably less likely to migrate from households owning more than five hectares of land, while the effect on male migration is not statistically significantly different from zero. This important result regarding the deterrent effect of land ownership on female migration, which is replicated with different data in Donato (1993), suggests that, in the context of widespread male migration, women may be expected to remain in areas of origin to administer the family farm. Davis and Winters (2001), however, examining migration from the ejido sector, find a negative association between irrigated land owned and male migration, but a (very small) positive relationship between rainfed land and female migration. Their interpretation of this result is that, given the predominant gender division of labour in Mexican agriculture in which men are more likely to work directly on the farm, irrigated land, which is associated with higher returns, raises the opportunity costs of migration for men. These data do support the idea that livestock assets reduce female migration propensities, which is consistent with the gender division of labour in sending communities.

Food production and agricultural liberalization

Across Latin America, and especially in countries experiencing high rates of rural male outmigration, women are assuming larger and larger roles in the peasant or subsistence agricultural sector. Even where men are not physically absent for extended periods of time, trade and price liberalization affecting important food crops such as maize has led to widespread reallocation of household productive resources - including labour - to accommodate higher input and lower output prices for basic grains that form the foundation of the diet of the rural poor’s. The mobilization of women’s (mostly unpaid) labour is thus one way for households to achieve food security in the face of significant economic disincentives to invest in subsistence crops.

Household time-use data from 2 000 small-scale agricultural households involved in the production of primary food crops in 13 Latin American countries suggest that women are spending an average of six hours per day in agricultural or livestock activities (Kleysen, 1996). Looking at the gender division of labour by task within the major food crops, it appears that women’s participation is somewhat lower in the land preparation and field care phases of production (where an average of 40 percent of households report some level of female labour contribution) and higher for the planting and harvesting stages (where participation rates exceed 60 percent). Women in half of all households contribute to post-harvest processing of food crops, and about 40 percent play a role in commercialization. With regard to livestock production (for which data are available mostly for the Andean countries), women in approximately half of the households participated in the feeding and pasturing of both large and small farm animals. They are somewhat less likely to take an active role in the breeding and health and sanitation aspects of cattle raising, but make important labour contributions to the milking of dairy cows, as well as to the care of and product collection from other small livestock such as pigs, sheep and chickens (Kleysen, 1996).

In their analysis of an indigenous smallholding community in the Mexican central highlands, Preibisch, Rivera Herrejón and Wiggins (2002) find a strong association between the continued cultivation of maize and the feminization of subsistence agriculture in the late 1990s. Their argument expands upon other Mexican literature (see Barrera Bassols and Oehmichen Bazán, 2000) which considers male migration as the primary impetus to women’s increased participation in the rural economy, by showing how economic restructuring shifts agricultural work into the realm of reproductive activities. On the factor price side, as the costs of fertilizers and herbicides have risen, families are substituting (mostly female) labour for tasks such as weeding. On the output side, as the entry of United States maize into the Mexican market has substantially lowered the price that Mexican farmers receive for their crop, households consumed a greater portion and marketed a smaller portion of the maize they grew. And with this (re)conversion of maize from a cash-generating to a truly subsistence crop, men have increasingly left its cultivation and control over important production and use decisions to women, in favour of off-farm (often migratory) employment. For many rural women - especially those in communities with limited access to reliable off-farm employment - maize production is an important way for them to meet their responsibilities as food providers for their families. The crop also serves as a source of savings that can be sold in case of emergency, as a source of cash for small daily expenses and as a key source of cooking fuel and animal feed.^[14] For women who would otherwise have to rely on allowances or migration remittances from their husbands, maize is a small but steady source of cash.

The expansion of non-traditional agricultural exports

One of the major developments in the rural economies of many Latin American countries during the 1980s and 1990s has been the growth of non-traditional agricultural exports - generally high value, labour-intensive products such as fresh fruits, vegetables, flowers and ornamental plants destined for northern markets (see, for example, Barham et al., 1992, and Thrupp, 1995). A striking feature of the industries that have grown up around these new crops is their significant employment of women. Thrupp (1995) reports that 69 percent of non-traditional agricultural export production workers in Ecuador are female, and 30 percent of field labour and half of processing, post-harvest handling and greenhouse cultivation jobs in non-traditional agricultural export crops in Costa Rica, Guatemala and Honduras are occupied by women. This phenomenon of female participation in commercial agriculture has greatly increased the visibility of rural women’s work, creating new opportunities for paid employment and raising new concerns about the position of women in the flexible labour market that often characterizes these industries (Lara Flores, 1995; Aparicio and Benencia, 1999). Three case studies in the Dominican Republic, Colombia and Chile illustrate some of the central issues surrounding gender and employment in the non-traditional agricultural export sector in Latin America.

In the Dominican Republic, non-traditional agricultural exports include fresh, frozen and canned fruits and vegetables, as well as nuts, flowers and ornamental plants. The sector is geographically dispersed throughout the countryside, with only 16 percent of firms clustered around Santo Domingo (Raynolds, 1998). Data from 1990 suggest that approximately 40 percent of the non-traditional agricultural labour force is female, with women workers concentrated in the labour-intensive food industries such as canned tomatoes and fresh melon and pineapple production. Women predominate in the most rapidly-paced assembly line operations such as selection, washing, filling cans and packing fresh produce into boxes. Men, on the other hand, specialize in supervisory, mechanical, lifting and transportation aspects of the work. The work is mostly seasonal in nature and provides employment to women who were either previously out of the labour force, unemployed or self-employed in the informal sector. A firm-based survey suggests that one-third of female non-traditional agricultural workers are household heads, and fully three-quarters are mothers, implying that it is not predominately young, single rural women who are accessing these new jobs (Raynolds, 1998).

Cut flower production has constituted an important new source of female employment in rural Colombia. Concentrated in the central highland region (Sabana de Bogotá), the cultivation of fresh flowers for the export market has generated approximately 75 000 jobs, 60 to 80 percent of which are held by women (Meier, 1999). Flowers are a labour-intensive industry (wages account for nearly half of total production costs) which exhibits a marked gender division of labour: women are generally charged with planting, crop care, cutting, classifying, packaging and ancillary services such as cleaning and food service; men are more commonly employed in the construction and maintenance of greenhouses and other infrastructure, plant bed preparation, pesticide application, cold storage, transportation and security. Employers have demonstrated a preference for younger women (who are often subjected to illegal pregnancy tests) and for rural women (who, because of their limited alternatives, are believed to be less inclined to press for higher pay and improved working conditions). Earnings in the flower industry are generally set by the minimum wage, plus benefits such as social security payments. The jobs have been especially attractive to single mothers and recent migrants - groups with high demand for assistance with childcare. A 1996 industry survey, however, reported that only 14 (out of approximately 500) flower farms had onsite childcare services (Ascolflores, 1998, as cited in Meier, 1999). Another major source of concern regarding the flower industry has been the health risks posed by exposure to pesticides. While safety practices have improved since the industry began operating in the 1970s, systematic research on the short- and long-term health effects of employment in the cut flower sector has yet to be carried out.

In Chile, thousands of women have been integrated into the rural wage labour force, working to export fresh fruits such as kiwi, apples and table grapes to northern markets. National estimates are that approximately 150 000 women are employed on a temporary basis in the agro-export sector, constituting fully half of the total temporary agricultural labour force (Bee, 2000). The sector is primarily employing older, married rural women: various studies indicate that only 35 percent of the Chilean female temporary waged workers (temporeras) are single, and their average age is about 30. In the Guatulame valley in Chile’s Fourth Region (Norte Chico), increased feminization and seasonality of agricultural production is associated with the agro-export economy. In this region, many women have become involved as temporeras, working in the fields and packing plants of the grape industry. Many of these women also maintain work on the family farm producing tomatoes, beans, peppers, garlic and melons for the national market. Employment in the fruit sector is highly seasonal; over half of all the women interviewed by Bee (2000) from two communities in the Guatulame valley only worked during the peak harvest months of December and January, and another third did not work for more than six months out of the year. Women are concentrated in the post-harvest processing jobs, including, in the case of the grape industry, selecting, cleaning and packing. About one-third of the women in the sample participated in field tasks.