Women and Welfare: the Politics of Coping With New Social Risks in Chile and Uruguay

Women and Welfare: the Politics of Coping With New Social Risks in Chile and Uruguay Jennifer E. Pribble August, 2004 pribble@email.unc.edu Paper prepared for delivery at the 2004 Meeting of the Latin American Studies Association, Las Vegas, Nevada October 7-9, 2004. 1

It is widely recognized by scholars of economics, political science, and sociology that women make up a disproportionate share of the world s impoverished population. Latin America has not been immune to this feminization of poverty, and a recent ECLAC (2004) study found that nearly half of women older than 15 years of age do not have access to an independent income, compared to 20 percent of men in the same age group (translation by author, 139). Despite the clear gendered differences in Latin American poverty, analyses of the region s welfare states 1 have focused primarily on explaining differences in overall social spending levels. 2 This focus has neglected an important point of variation between Latin American social policy regimes: the gendered nature and outcomes of social policy. This paper aims to fill this hole in the literature. Through a comparative analysis of Chile and Uruguay, I consider the question of how and why the two countries vary in their ability to formulate policy that responds to the unique social risks faced by women. The design of social policy regimes has a profound impact on women s material well being as well as in the formation of societal definitions of a woman s role as a citizen, worker, care-giver, and wife. While some welfare regimes reproduce traditional gender roles, other systems provide women with an opportunity to reduce their dependency on both the market and the family. Additionally, some welfare states are able to address the specific social risks faced by women more effectively than others, producing lower levels of female poverty and male-female wage dispersion and higher levels of female labor-force participation. Ensuring equal social protection for women is important for many reasons. First, there is the normative expectation that democratic regimes will grant equal rights to all citizens regardless of sex, race, age, etc. Secondly, reducing women s exposure to risk is important for reversing recent growth in poverty and inequality in Latin America. The Chilean government estimates that between 1990 and 1996, 19.5 percent of households escaped poverty because of female entry into the labor force (ECLAC, 2000a: 51). This paper analyzes Chile and Uruguay s welfare regimes with regard to gendered outcomes. The comparison of the two countries is particularly insightful because despite similar levels of social spending, Chile and Uruguay have markedly different gendered outcomes. Employing a comparative historical analytic lens, I find that differences in gendered social policy regimes can be explained by the interaction of structural labor market conditions, the mobilizing capacity of women, regime legacies, and ideological constraints. New Risks and New Welfare: What are the Issues? Welfare states were originally designed to protect against risks that workers encountered in the market, namely the loss of income due to unemployment, sickness, disability, and retirement. In recent decades, however, the risk structure of the market has changed and new risks have emerged alongside the traditional issues. In Latin America, these new risks include the shrinking of the formal sector and subsequent growth of the unregulated informal market, and dramatic changes in the family structure. 1 I have chosen to use this word despite clear differences between the scope of Latin American social policy regimes and those of most advanced industrialized economies. While Latin American welfare regimes pale in comparison to their European counterparts, the concept is still useful. 2 See: Brown and Hunter, 1999; Dion, 2001; Huber, 1996; Kaufman and Segura, 2001; and Wibbels, 2003. 2

Women have been particularly affected by the emergence of these new risks because of their dual dependency on the family and the market. Indeed, changes in the family structure, namely growth in the number of marriages ending in divorce and an increase in solo-mother households, has increased women s exposure to risk. In conjunction with this shift away from the traditional family structure, the composition of the Chilean and Uruguayan labor-force has experienced a dramatic transformation in recent years. Women have increasingly moved out of unpaid domestic work and into the formal and informal labor markets in both countries, although to a lesser degree in Chile. At the same time, the economic crises of the 1970s and 80s resulted in higher levels of unemployment; 3 meanwhile competition for manufacturing work from lower-wage countries has resulted in a reduction of low-skill jobs. As a result of these changes, many workers have been forced into the informal sector. For the purpose of this paper, I define the informal economy as all incomeearning activities that are not regulated by the state (quoted in: Portes & Schauffler 1993: 6). The United Nations Economic Commission for Latin American and the Caribbean (ECLAC) estimates that as many as 50 percent of Latin American workers in urban centers function in the informal sector (quoted in: Schild 2002: 190). Because of the low-skill nature of informal work and flexibility that it offers, women are more likely to work in the informal sector than men. Victoria Contreras (2003) finds that, in 1998, 44.8 percent of working aged women in Chile functioned in the informal sector, while only 32.9 percent of men worked informally (60-64). A 2004 ECLAC study suggests that this same gendered division of informal sector employment also exists in Uruguay (162). In response to the emergence of these new social risks, citizens have, in some cases, urged politicians to respond with policy innovations. Chile and Uruguay have varied in their ability to address the new social risks faced by women. Despite similar spending levels, a shared democratic tradition, and historically important welfare states, the countries have divergent gendered social policy outcomes. Chile and Uruguay: the Numbers Chile and Uruguay have traditionally been considered to have two of the strongest and most universalistic welfare states in Latin America (Castiglioni, 2000: 1, Mesa Lago, 1994: 115 and 155). Both countries social policy regimes began to take shape in the early to mid 20 th century and slowly expanded to provide pensions, education, healthcare, and eventually family allowances and maternity leave. The evolution and roots of each state s welfare policies, however, differ greatly. Growth of the Chilean welfare state responded in large part to class-based demands. Social movements with roots in the working class generated organizational momentum and demanded social protection from the state (Castiglioni, 2000: 3). In Uruguay, however, the creation of the welfare state corresponded to political moves by the incumbent Colorado party to ensure electoral support from the urban working class (Filgueira, 1995: 8-9). Despite these different roots, Chile and Uruguay s social safety nets were strong by the end of the 1960s. Indeed, both countries were among the top four spending states 3 It should be noted that both Chile and Uruguay offer unemployment benefits. Uruguay s social insurance system, however, provides a longer benefit period and more generous payments than Chile s targeted program (Social Security Administration, 1999). 3

in the region for social security and welfare, health, and education as a percent of GDP (Huber & Stephens, 2003: dataset). 4 During their respective experiences under authoritarian rule, however, the two countries began to move in opposite directions. In 1973, democracy was interrupted in both countries. In Chile, the military ousted a democratically-elected government, while in Uruguay then-president Juan Bordaberry shut down the Congress, enacting a self-coup. The Chilean military junta was dogmatic in its economic reforms, and the liberalizing shock and stabilization efforts that were enacted by the Pinochet government were unmatched by the Uruguayan regime. Following the return to democracy, Chile s center-left Concertación government has worked to reconstruct a social welfare regime. The programs, however, do not represent a break with the Pinochet-era neo-liberal model, but rather the center-left governments have increased funding to expand means-tested programs to target the country s most vulnerable sectors. Indicators of Gendered Welfare Outcomes Female labor force participation is an important indicator of women s autonomy. When women are capable of obtaining an independent income, they reduce their dependency on the family and decrease their likelihood of living in poverty. Welfare regimes are fundamental in shaping women s access to employment. Social policies, such as subsidized childcare and universal education, support women s entrance into the labor force. Similarly, coordinating the timing of the school day with the official workday helps to reduce the burden of the double role of care-giver and bread-winner faced by women (Gornick, Meyers, & Ross, 1996: 2-7). Generous maternity leave policies might also strengthen women s attachment to the labor market, as such programs allow female workers to enter and exit the market with full job security. Finally, welfare states, especially those that provide services such as healthcare and education, can serve as an important source of employment for women (Kilkey and Bradshaw, 1999: 149). Uruguay has significantly higher levels of female labor force participation than Chile. 5 In 2000, 44.1 percent of working-aged women 6 were active in Uruguay s formal labor force, compared to only 31.4 percent in Chile. It is possible that Uruguay s low population growth and high employment levels may have generated a greater demand for female employment historically, thus accounting for some of the variation in female labor force participation rates between the two countries. It is unlikely, however, that this historical difference accounts for all of the variation. 7 Furthermore, Chile s levels of female labor force participation are among the lowest in the region, while Uruguay s are 4 Uruguay is the top spending country in the region for social security and welfare. It has occupied the top position since 1985. Chile ranks 3 rd or 4 th (depending on the year) for social security and welfare spending. In terms of health and education spending, Chile ranks second in the region, while Uruguay is the fourth highest spending country for this category (Huber & Stephens, 2003: dataset). 5 This measures only formal labor force participation. It is likely that if both the formal and informal sectors were considered, female labor force participation would be similar between the two countries. 6 The international standard for working-age is between 15 and 65. Female labor force participation is, therefore defined as the percentage of all women between the ages of 15 and 65 who are active in the formal labor market. 7 Uruguay s unemployment levels have surpassed those of Chile in recent years, yet the level of female labor force participation remains higher than Chile. This provides additional evidence that variation in female labor force participation cannot be fully explained by historical differences in the demand for female labor. 4

among the highest (Mizala, Romaguera, and Henriquez, nd: 1-3). This fact suggests that while historical discrepancies are important, policy makes a difference when it comes to incorporating women into the labor force. [TABLE 1] Of particular interest in Table 1 is the fact that growth in Chilean female labor force participation was most sluggish during the country s military regime, while in Uruguay the opposite occurred. Indeed, women s entry into the formal workforce grew dramatically during Uruguay s dictatorship. The existence of these opposing trends highlights an important difference in Chile and Uruguay s respective experiences under authoritarian rule. While the social conservatism of Chile s military regime institutionalized traditional gender roles and notions of women as wives and mothers, Uruguay s regime was more gender-neutral. Chile s low level of female labor force participation is even more dramatic if it is analyzed by income quintile. In 1990, the labor force participation in the formal sector among women in the top income quintile was 45.4 percent, while it was a mere 18.2 percent among women in the bottom quintile. 8 Similarly, in 1996 there was a 31.6 percentage point difference between female labor force participation in the lowest and highest income quintiles (ECLAC 2000a: 91). In contrast, male workers participate at nearly identical rates across income quintiles (ECLAC 2000a: 91). These data suggest that Chilean social policy is not only ineffective at encouraging female labor force participation relative to Uruguay, but the policy has a disproportionately negative effect on low-income women. Similar data are not available for Uruguay, but because of the country s more equal distribution of income, I expect it to have lower levels of labor force participation inequality between income quintiles. Much like the story of female labor force participation, the share of all femaleheaded households that live in poverty and extreme poverty is significantly higher in Chile than in Uruguay (ECLAC, 2000b: 279). 9 This trend has remained constant over time, with Uruguay consistently outperforming Chile. Indeed, ECLAC (2000b) data reveal that female-headed households in Chile are, on average, three to four times more likely to be poor or extremely poor than the same family type in Uruguay (279). Finally, Uruguay outperforms Chile in terms of women s earnings relative to those of men. According to the United Nations Development Program (2001, 2003), the ratio of estimated female to male earned income is higher in Uruguay than in Chile. In 2001, the female/male earned income ratio was.50 in Uruguay, compared to.36 in Chile. This difference is significant, and when considered with the other findings outlined in this section, it points to two over-arching conclusions about the differences between the Chilean and Uruguayan welfare states. First, Chilean social policy seems to foster higher levels of inequality between men and women than Uruguayan policy. 8 This is very important for the discussion of regime legacies. The divergent experiences of low-income and wealthy women have undermined cross-class alliances in the women s movement. 9 This variable is not the same as single mother households in poverty. The category female heads of household is comprised of single mothers and dual-parent homes in which the female earns more than the male. For this reason, the measure share of female-headed households in poverty differs only slightly from overall poverty levels. Due to lack of data regarding the number of single mothers who live in poverty and extreme poverty, this is the best available measure of women s relative well-being in each country. 5

Second, Uruguay appears more capable of reducing women s dependency on the family and the market than Chile. A Gendered Framework for Analyzing Social Policy The differences in Chile and Uruguay s gendered social policy outcomes beg an explanation, which in turn requires a gendered analytic lens. 10 Sainsbury s (1996) framework provides a gendered approach to the welfare state with a typology of two welfare regimes: the male breadwinner and the citizen model. The framework establishes specific dimensions of variation in both the demand and supply side of the welfare state. The male breadwinner ideal type is characterized by unequal provision of benefits and entitlements that are grounded in an individual s position in the labor market and/or family. The citizen welfare state, on the other hand, provides universal protection regardless of an individual s status in the market. Sainsbury s (1996) framework provides insight into the level of autonomy that women enjoy in each type of welfare regime. The framework is also useful because it allows for an analysis of aspects of social policy provision that may not be explicitly gendered, but which nonetheless have consequences for women s well-being. This aspect of Sainsbury s framework is important for my analysis, as some programs discussed in the following section were not adopted with gendered goals in mind, but still affected women s social rights. Finally, Sainsbury s (1996) framework is particularly useful because the specification of elements allows for a systematic comparison. 11 Breadwinner versus Citizen Welfare Sainsbury s (1996) framework, which was developed for the analysis of advanced industrialized welfare states, must be modified somewhat for research on Latin American social policy. Unlike in OECD states, the dichotomous distinction between breadwinner and citizen welfare regimes does not provide sufficient insight into differences between Latin American social policy regimes because no state in the region provides citizen-style welfare. Rather, Latin American social policy tends to vary with regard to the strength of the breadwinner model. For this reason, I have adapted Sainsbury s framework for my analysis of Chile and Uruguay. I use the same dimensions, but rather than considering the two categories of breadwinner and citizen, I create a continuous variable: strength of breadwinner. To discern the strength of the breadwinner model in Chile and Uruguay, I evaluate three policy sectors: family allowance, childcare, and maternity leave. I have chosen to focus on only three areas for two reasons. First, the space of a paper will not permit an analysis of the full welfare state. More importantly, however, is the fact that by limiting the study to these three policy areas, a meaningful comparison can be drawn. Mala Htun (2003) argues that, issues differ in how they are processed politically, the groups that weigh in on policy debates, and the ideas at stake in change To explain policy change, we must disaggregate gender issues (4-5). In this vein, I maintain that 10 See also: Orloff (1993) and O Connor, Orloff, and Shaver (1996). 11 For the purpose of this argument (and because there is simply not space in this paper) I have decided to omit five of Sainsbury s dimensions: the tax system, the unit of contribution, employment and wage policies, and caring work (paid or unpaid). The caring work dimension is relevant to the argument, but because the countries do not differ, it is not discussed. 6

the actors and ideas associated with these three policy-areas are comparable because all three are subsumed under the common issue of assisting women with the dual demands of motherhood and work. I provide a score for both countries on each dimension of the framework. For dimensions that are discussed separately (family ideology and sphere of care), each country receives one score for the dimension. When the dimensions are discussed across policy areas, the countries receive two scores (one for family allowance policy [FA] and another for maternity leave [ML]). The scores are displayed in table two. As mentioned previously, I operationalize the dependent variable on a continuous scale of breadwinner strength. My classification scheme denotes five categories that are scored in increments of one. The scale, therefore, ranges from one to five, with high scores denoting a weak breadwinner model. 12 I have also provided an average score for the overall nature of the welfare state. 13 The first dimension of Sainsbury s typology is familial ideology, which I maintain does not vary across policy sectors, but rather provides a common base upon which all social policy is constructed. Chile differs dramatically from Uruguay on this dimension because divorce did not exist in the country until March of 2004. Although the new law provides for a legal end to marriage, couples must be separated for at least one year before obtaining the divorce and both wife and husband must agree to the break-up, otherwise the waiting period is three years. It is clear, therefore, that while divorce does exist, the state continues to afford special legal protection to the institution of marriage. By contrast, divorce was permitted in Uruguay as early as 1907 (Garrido, 2000: 150). 14 Additionally, Chilean family law, specifically women s rights in raising children, women s partial control of marital property, and the rights of illegitimate children, remained extremely unequal until 1998 (Htun, 2003: 133). Similar laws granting women equal property rights in marriage and in child rearing were approved in Uruguay s 1946 Civil Code (Gobierno de Uruguay 1946). In light of these differences, I score Chile as a very strong breadwinner (1) and Uruguay as a breadwinner (3) on the dimension of familial ideology. 15 Family Allowances Chile and Uruguay both offer family allowances to impoverished families. In Chile, men and women have access to family allowances, but the entitlement requirements are not equal across the sexes. Chilean men and women qualify for the allowance if they have dependent children under the age of 18 (or 24 if in school). The entitlement changes, however, with regard to the spouse s status. Men qualify for a family allowance if they have a dependent wife, but women only receive an allowance 12 The full classification scheme is thus: very strong breadwinner (1), strong breadwinner (2), breadwinner (3), weak breadwinner (4), very weak breadwinner (5). 13 This average may not be the best way of establishing the full nature of the welfare state because it weighs all dimensions equally when in fact some dimensions may be more important than others in terms of influencing gendered outcomes. For the purpose of this paper, however, I assume that all dimensions are of equal importance. 14 Divorce at a woman s request was permitted beginning in 1913 (Garrido, 2000: 150). 15 Sainsbury (1996) specifies a second element of familial ideology: the gendered division of household labor. This characteristic is difficult to measure and since the two countries do not differ significantly, I have omitted any discussion of this aspect of familial ideology. 7

for their spouse if the husband is disabled (Social Security Administration, 1999: 77 & 378). In other words, working women in Chile do not receive the same level of family allowance benefits as working men. The Uruguayan family allowance system does not make this distinction, but rather pays a set amount for each dependent (regardless of sex). For these reasons I have scored Chile as a very strong breadwinner (1) and Uruguay as a very weak breadwinner (5) on these dimensions. The two countries family allowance systems also differ with regard to the basis of entitlement. In Chile, family allowances are available to employed persons and pensioners with one or more children or other eligible dependents (Social Security Administration, 1999: 77), while in Uruguay the benefit is extended to a wider range of individuals: employed persons, domestic workers, persons in receipt of unemployment benefits, newspaper and small rural products vendors, and pensioners (Social Security Administration, 1999: 377). 16 This diversity of entitlement base provides Uruguayan women with better access to family allowances since many domestic workers and small vendors are female. Additionally, on May 6, 2004, President Jorge Battlle signed a bill into law that extends coverage of family allowances to all households with incomes of less than three times the national minimum wage, including those headed by individuals working in the informal sector, (Radio Espectador, 2004). As I have mentioned previously, it is estimated that women are over-represented in the informal sector because of its flexibility and demand for low-skill labor. Thus, incorporating these categories of workers into the entitlement base boosts women s access to the benefit in Uruguay. Another factor that negatively affects the ability of Chilean women to secure access to family allowances (and all labor-based welfare transfers) is the two-tiered system of employment that exists in the country, which grants some workers access to social welfare benefits, while denying the same to others. Chilean workers who are subject to contract have access to all labor-market-based welfare benefits, while individuals who are paid on something called a boleta, or a receipt, do not enjoy the same privileges. It is important to note that boleta workers are formal sector employees, but their status is less than that of contracted workers. Access to fully contracted work is diminishing in Chile, while the boleta system has become more prevalent. The percentage of all salaried workers with contracts decreased between 1996 and 2000 from 80 to 77 percent for men and from 78 to 72 percent for women. High-income earners are much more likely to be contracted than low-income individuals. In 2000, 86.5 percent of females working in the formal sector with earnings in the top income quintile were contracted laborers, while only 39.7 percent of women in the bottom income quintile had a contract. Furthermore, the gap between the number of men and women who have a contract is largest among the poor. In 2000, the difference in contract coverage between male and female workers in the top income quintile was 1.5 percentage points. In the bottom income quintile, however, the difference in coverage was 18.6 percentage points (De la Luz Trautmann. 2002). It is clear, therefore, that a large proportion of low-income women are employed through the boleta system rather than by contract. This means that very few poor women have access to welfare benefits even when they work in the formal labor force. In short, Chile s two-tier employment system poses an important obstacle to women s entitlement 16 The Uruguayan family allowance system obtained this level of universality by the 1950s (Filgueira, 1995: 18). 8

to benefits. For all of these reasons, I have scored Chile as very strong breadwinner (1) and Uruguay as very weak breadwinner model (5) for the family allowance entitlement dimension. The recipient of the family allowance benefit in both the Uruguayan and Chilean case is the head of household (Social Security Administration, 1999: 77 & 378). Additionally, the unit of benefit in both cases is the family rather than the individual (Social Security Administration, 1999: 77 & 378). I, therefore, score Chile and Uruguay as breadwinner (3) on both the recipient of benefit and unit of benefit dimensions. Sphere of Care and Caring Work The Chilean government created a system of subsidized childcare in 1994. The benefit is provided to working mothers who qualify through the country s socioeconomic targeting scheme. 17 To qualify, the household must score no more than a minimum number of points on the targeting survey and the mother must show an employment contract and proof of working hours (Clert and Wodon, 2001: 60). The system, therefore, excludes three groups of potential beneficiaries: women working in the informal sector, women looking for work, and women who are not employed by contract. As I have previously demonstrated, low-income women who work in the formal sector are unlikely to be employed by contract. The exclusion of these groups, therefore, has a negative impact on the coverage of the childcare benefit among poor female-headed households. In addition to formal exclusion, Clert and Wodon (2001) argue that very few Chilean households are aware that they might be eligible for many of the means-tested social benefits. In a survey of impoverished Chilean homes, the authors found that 51 percent of households had never heard of any state-offered social entitlements (52). This lack of knowledge increased for certain programs. Of interest is the fact that 70 percent of female respondents had not heard of a national program aimed at reducing poverty in households headed by women (52). Uruguay also provides a form of state-subsidized childcare to a small number of low-income women. Beginning in the 1930s, but expanding dramatically during the 1980s and 90s, the Uruguayan state developed neighborhood-based Infant and Family Attention Centers (CAIF). The aim of CAIF is to improve the well-being, level of development, and social insertion of families in high socioeconomic risk groups through nutrition programs, health attention, education, and family assistance (quoted in: Midaglia, 2000: 63; translation by author). Childcare was not a focus of the CAIF program originally. With time, however, the centers have begun to function as a form of daycare. CAIFs operate during the eight-hour workday, and according to the staff, many women use the service as a form of childcare. The CAIF program, while successful, is limited in scope. 18 According to Midaglia (2000), the most common complaint about the centers is that many families do not have access to the service. One-third of CAIF centers report having a waiting list that represents 60 percent of its current coverage (67). The government uses a targeting scheme to determine which families have access to CAIFs. The targeting method relies 17 Some municipalities have also attempted to provide childcare, but I was unable to find systematic information about such services. 18 I was unable to find official estimates of the program s coverage. 9

on family income and the age of children in the household. Families with children under four and pregnant and nursing women are granted priority (Midaglia, 2000: 68). The effectiveness of this targeting method is buttressed by the fact that centers only operate in low-income neighborhoods and provide services to individuals living in the community. Women are not required to show proof of employment to access the service and therefore females in the informal sector and individuals looking for work are able to use the childcare services. The CAIF targeting scheme is very efficient at providing services to Uruguay s poorest families. According to Edmundo Murrugarra (2000), 80.7 percent of CAIF services are used by the lowest 20 percent of income-earners. The remaining 19.3 percent of users are from the second poorest income quintile (23). In summary, both Chile and Uruguay provide limited and targeted daycare for low-income families. Although the childcare efforts are similar in both countries, the targeting mechanisms used in Chile and Uruguay generate very divergent outcomes with regard to the potential autonomy enjoyed by women in each country. In Chile, the targeting process limits the services to low-income women who work in the formal sector and have a contract. This excludes informal sector workers, women employed without contract, and women looking for work. As has been previously noted, many low-income women fall into one of these three categories. In Uruguay, CAIF provides women with high-quality day-care, but the availability of services is quite limited. Families are selected based on income and the age/vulnerability of the child. This process does not, however, systematically exclude any one group of women. Additionally, the targeting scheme is very effective at delivering the service to the country s poorest families. For these reasons, I have assigned Chile the score of weak breadwinner (4) and Uruguay the score of very weak breadwinner (5) for sphere of care. Maternity Leave Both Chile and Uruguay offer paid maternity leave to women working in the formal sector. In Chile, women have the right to six weeks leave before and 12 weeks after the birth of the child. In Uruguay, the leave is six weeks before and six weeks after the birth of the child. Both countries provide women with 100 percent of their salary during the time of leave (Social Security Administration, 1999: 76 & 376-77). Additionally, Chile grants women a one-hour feeding break per day until the child is two years old, while Uruguay provides the break only until the child is six months old (Abramo, 2002: 3). At first glance, the Chilean system seems to be more generous than the Uruguayan system, offering greater flexibility for women to enter and exit the labor force. While this is true for some women, many female workers are denied access to the benefit because of the two-tiered system of employment and employer infractions of the law (Henríquez, Riquelme and Cárdenas, 1997). Entitlement to the maternity leave benefit is differentiated among spouses in both the Chilean and Uruguayan case. Neither country provides for the possibility of men to take paternity leave, as is the case in some advanced-industrialized welfare states. 19 I 19 A bill aimed a increasing the flexibility of maternity leave is currently under debate in the Chilean Congress. If approved, the legislation would allow women to take all 18 weeks of maternity leave after the child s birth (rather than six weeks before and 12 weeks after). A second portion of the bill proposes that male workers also have the possibility to take paternity leave. If approved, the legislation would allow families to decide if the father wants to take a portion of the leave while the mother returns to work. 10

have, therefore, scored both countries as breadwinner (3) on the entitlement dimension for maternity leave. The basis of the entitlement in both cases is as breadwinner, not as a citizen. Neither country offers financial support to pregnant women who have never been attached to the labor market. Uruguay, however, does provide a broader entitlement base. Rights to the maternity allowance are extended to self-employed individuals as well as women receiving unemployment benefits (Social Security Administration, 1999: 376). This is important because it provides protection to women who have recently lost jobs, which does not exist in the Chilean system. Furthermore, it should again be noted that in Chile the entitlement base is greatly reduced by the fact that only contract workers are eligible for the maternity allowance. For all of these reasons, I have scored Chile as very strong breadwinner (1) and Uruguay as very weak breadwinner (5) for the basis of entitlement to maternity leave benefits. The unit of benefit and the recipient of the benefit is the household in both Chile and Uruguay and I have therefore coded both countries as breadwinner (3) for those dimensions. Although maternity leave benefits in Chile appear quite generous, the coverage is limited because of the two-tiered employment system and a lack of enforcement. A 1995 study by the government s Labor Ministry found that 24 percent of female workers who qualified for maternity leave reported that they had been treated with hostility at the work place after announcing their pregnancy (14). Common complaints were that employers threatened to fire the worker or moved her to a different position (14-16). Although such treatment didn t occur in the majority of cases, it is important to note that almost one in four Chilean workers complained of unfair treatment with regard to their pregnancy. The same does not appear to be true in Uruguay, where I was unable to uncover instances of such abuse. 20 [TABLE 2] A Theory of Gendered Differences I have now established that the Chilean and Uruguayan welfare states differ in significant ways in terms of their gendered outcomes. Through an analysis of three policy sectors family allowances, maternity leave, and childcare I have demonstrated that the Chilean welfare state is a strong breadwinner model, while the Uruguayan state is a weaker version of the breadwinner type. I now turn to building a theory that explains such gendered differences. The question of why social policy regimes (in both gendered and non-gendered terms) vary cross-nationally is an issue that requires researchers to take a long-term analytic perspective. Such an approach is necessary because, in the words of Paul Pierson (2003), the process is big, slow-moving, and invisible (177). This is to say that the causes and consequences of variation in social policy outcomes become more visible over long time horizons. For this reason, I employ a historical and path dependent approach to explain variation in gendered outcomes. In addition to being historical and path dependent, variation in the gendered nature of social policy regimes is unique in that the outcomes may be both intentional and unintentional consequences of policy formation. Put differently, some policy, such as the Chilean adoption of the two-tiered labor market, or the Uruguayan law that incorporates 20 My research to uncover instances of violations of maternity leave law involved searching congressional documents, news sources, and the Labor Ministry s web page. 11

informal workers into the family allowance system, is gender-neutral in nature, but has important consequences for women s social rights. For this reason, an explanation of gendered social policy outcomes must consider both specifically gendered and nongendered historical processes. In light of these features, I argue that differences in the gender-friendliness of social policy regimes are best explained in terms of long-term path dependent processes that are both gendered and non-gendered. This is not to say that gendered social policy outcomes are pre-determined, or that states do not change policy directions. Rather, I maintain that the formation of gendered social policy regimes is a historical process that functions in the following manner: the nature and evolution of each country s labor market shapes the character of women s mobilization, which in turn influences the ideological power distribution and shapes the constraints on gendered and non-gendered social policy formation. The policies that emerge from this process (both gendered and non-gendered), exhibit a feedback mechanism, or regime legacy, that reshapes the nature of the labor market and the context in which women s mobilization is carried out. The process, therefore, is dynamic and self-reinforcing across time. Structural Labor Market Conditions In the case of Chile and Uruguay, an important starting point for conceptualizing variation in gendered social policy outcomes is the structural nature of the labor market and the difference in women s labor force participation. As displayed in table one, since the 1960s Uruguayan women have entered the work force at a much faster pace than Chilean women. Although female involvement in the labor force has grown throughout the world, the trend was likely intensified in Uruguay because of the high demand for labor. Historically, the country s small population and highly urbanized context created an excessive labor demand, thus providing Uruguayan women with an opportunity that did not exist in Chile. Finally, as noted previously in this paper, the different levels of social conservatism exhibited by each country s authoritarian regime influenced growth in women s labor force participation. As Uruguayan women gained increased access to the labor market (and unions as an organizing space), 21 they obtained a new level of political relevance that they had not previously enjoyed. Similarly, as they began to earn an independent income, they increased their economic power and mobility. In Chile, on the other hand, low levels of female labor force participation hindered the organizational and economic resources that women possessed to issue policy demands. In short, labor market conditions that encouraged women s entrance into the workplace granted women both economic and political relevance in Uruguay while the opposite held true in Chile. Women s Mobilization Different levels of female labor market participation in Chile and Uruguay, in turn, shaped the gendered distribution of power in each country. Changes in the power 21 Jaime Ensignia and Sonia Yáñez (1999) argue that although female unionization is low throughout the southern cone, the greatest barriers to participation are found in Chile (20). Furthermore, they note important achievements made by women in Uruguayan labor unions, such as the creation of the Tripartite Commission for Equal Opportunity and a special Women s Job Training Division in the Labor Ministry (62). 12

distribution were not automatic, but resulted from the mobilization of women s interests. Uruguayan women in the second half of the 20 th Century built on a strong tradition of mobilization. Indeed, by 1916 Uruguayan women s groups had demonstrated their growing economic and political power. Female workers led by Dra. Paulina Luisi founded the National Women s Council, which organized meetings and protests to demand voting rights. Although Uruguayan women were not granted suffrage until 1932, the Communist Party began efforts to incorporate women into the political organization beginning in the 1920s (Garrido, 2000: 149). The Communist Party was not alone, however, in its desire to attract females to the party base. In 1931, the Colorado party created the Batllista Feminist Committee, which also provided women with unique access to political resources (Garrido, 2000: 149). Prior to the formation of the committee, however, the Colorado party had displayed willingness to extend the role of the state beyond the public sphere and into the private, family domain. An example of policy that regulated household decision-making was the government s 1921 creation of the Women s University, which provided women with free access to higher-education, thereby decreasing young women s dependency on the family (Filgueira, 1995: 6). The formation of the Batllista Feminist Committee and the creation of the Women s University are particularly significant events in light of the nature of Uruguay s welfare state development. As explained previously in this paper, the Uruguayan welfare state originated as a top-down political project. Fernando Filgueira (1995) argues that, rather than reacting to demands for social justice, the [Uruguayan] state attempted to mold the groups that could formulate such demands (9). The Colorado party s move to make inroads with women, therefore, suggests that females were perceived as a potential constituency group by party leadership. Thus, it is possible that special efforts were made by the Batllista state to incorporate women s interests through social policy formation. This marks a difference with Chile, where welfare state formation came in response to pressure from below. The working class movements that prompted the development of Chilean social policy did not, however, provide any special space for the mobilization of women s interests. Throughout the second half of the 20 th century, women s growing political and economic relevance slowly opened spaces for the mobilization of gendered interests in national political institutions. Gender-relevant policies were, in turn, adopted in a progressive manner. By 1934, some maternity leave privileges had been established. These rights were expanded following the passage of a 1958 law that guaranteed paid maternity leave for female workers. In 1946, the Uruguayan state enacted the women s civil rights law and by 1968 the first female minister had been appointed (Garrido, 2000: 149-150). With the expansion of greater social rights, women obtained additional power resources. During the 1970s and 80s Uruguayan women participated actively in leftist political movements and later in opposition to the military dictatorship. Following the transition back to democracy, women have continued to create organizational spaces within political parties and labor unions to advance their interests. As a result, an equal opportunity employment law was approved in 1989 and further expanded upon in 1999. These policies granted women additional rights and facilitated greater entry into the labor market, which further intensified women s mobilization and allowed for additional altercations in the gendered distribution of power. This path dependent process moved forward largely because of activity by women in civil society groups and political 13

parties. Although moves to incorporate women into parties began with the Colorados in the 1930s, more recent attempts to secure the female vote have been exercised by the center-left Frente Amplio (FA) and Nuevo Espacio (NE) coalitions. In the 1999 Presidential conventions, only 15 percent of Colorado attendees and 8 percent of Blanco attendees were women. By contrast, 22 percent of FA convention participants and 30.6 percent of NE convention attendees were women (Laurnaga, 2001: 2). Uruguayan women have also assumed political positions in the Congress and in ministerial posts. In March of 2000, female deputies in Uruguay s Congress banded together to create a formal women s bancada, or legislative coalition. The group s primary aim is to promote legislation that expands women s rights (Garrido, 2000: 151). Thus far, the coalition of female legislators has pushed a bill through the House of Deputies that would legalize some forms of abortion. Additionally the bancada introduced debate over the adoption of a legislative quota system for women. Although these policy initiatives have subsequently been paralyzed, the fact that a formal alliance of progressive female politicians exists, provides Uruguayan women with a channel to pursue demands for more equal social rights. In sum, gendered social policy development in Uruguay demonstrates path dependent tendencies. At the core of this continuity, however, is the issue of women s mobilization. I maintain that there is a difference between women s mobilization and the women s movement. This difference is important for my argument. I define the women s movement as the combination of civil society organizations in a country that organize with the unique goal of furthering women s rights. Women s mobilization, on the other hand, is the organization and mobilization of women within any civil society group, regardless of whether or not the group has a feminist aim. This difference is important for my argument because in Uruguay it could be argued that the women s movement is not especially strong. Still, it is undeniable that women s mobilization is extremely high. In other words, while women s groups in Uruguay do not appear to possess special political clout, women as societal actors (by means of their mobilization in unions and parties) do hold a fair amount of political power. In short, Uruguayan women have used non-gendered engines of social change to further their own gender-specific goals. In fact, the high level of women s mobilization in Uruguay has been particularly effective because women have organized their interests within groups that have well-established links to the policy-making sphere. In a historical study of welfare state development and feminism, Christin Ehrick (2001) notes that [women s] participation in social assistance organizations gave them political experience and the confidence to confront state structures and to fight for institutional autonomy (139). In Chile, by contrast, there are numerous women s groups, but very few have access to the policy-making sphere and women s political mobilization remains comparatively low. The self-reinforcing cycle of increased political power that occurred in Uruguay, was not matched in Chile. Rather, Chile s path dependent process worked in the opposite direction and has therefore produced less equitable gendered social policy outcomes. As demonstrated in table one, female labor force participation in Chile has been consistently lower than in Uruguay, thus hampering women s access to a space within which to organize interests and develop power resources. The lack of mobilization has solidified 14

traditional gender attitudes and slowed the redistribution of power between men and women in Chile. Furthermore, characteristics of the Chilean women s movement, namely the existence of sharp class divisions and a hesitation to enter the policy-making sphere, have also undermined groups capacity to alter societal preferences about gender roles. In her work on the Chilean women s movement, Lisa Baldez (2002) argues that women s mobilization has, for the most part, unfolded outside the policy making sphere (12). Indeed, while it would be incorrect to label the Chilean women s movement as weak or non-existent, it is fair to say that the organizations have not mobilized their interests within groups that have access to the policy-making domain. This organizational strategy has limited the ability of Chilean groups to radically alter preferences about gender roles. Thus, while Uruguayan women obtained space within parties and unions to gradually advance their own interests, Chilean women s groups have remained in the political periphery. Regime Legacies It should be noted that the mobilization of women s interests in Uruguay is strengthened by regime legacies. Uruguayan social policy, in general, has been generous and based in a universalistic conception of welfare. These two factors have fostered a cross-class pro-welfare coalition, which has been further strengthened by low levels of inequality. Indeed, even as income inequality sky-rocketed throughout Latin America during the 1980s and 90s, Uruguay s Gini index a measure of income inequality held steady around 40 (Huber & Stephens, 2003: dataset). The relatively small gap between rich and poor in Uruguay has not only strengthened the pro-welfare coalition, but it has also provided a stable base for women s mobilization. Ehrick (2001) finds that middle and lower class women benefited from the early stages of Batllismo and were therefore invested in its protection and expansion (138-139). In short, the multi-class nature of the Uruguayan welfare state provided middle and lower class women with a sense of linked fate, which has in turn facilitated a united mobilization of gendered demands. 22 In Chile, by contrast, the relatively generous nature of the country s early welfare state was reversed during the 17 years of Gen. Agusto Pinochet s military rule. Pervasive income inequality and the introduction of targeted social policy provision, both of which were initiated by the dictatorship, undermined cross-class coalitions and further crippled the ability of the women s movement to mobilize unified demands (Blofield 2003). This process generated a negative policy regime legacy that was not specifically gendered in nature, but which has influenced the mobilization of women. It is important to note that from the 1920s onward, both Chile and Uruguay engaged in a comparatively advanced process of welfare state formation. This situation changed dramatically following both countries 1973 coups. Chile s experience under military rule marked a dramatic break with previous social policy trends. Upon consolidating its power, the Chilean military junta enacted a strict process of economic liberalization and cost-containment, dismantling many of the country s social welfare programs. Poverty and inequality levels soared under Pinochet s rule. The Gini index 22 It is important that Uruguay s feminist movement occurred simultaneous to the expansion of the country s welfare state. This is different from Chile and could be important in explaining differences in the nature of each country s women s movement (Ehrick 2001). 15