75835 MARCH 2013 • Number 109 Gender Equality and Economic Growth in Brazil Pierre-Richard Agénor and Otaviano Canuto This note studies the long-run impacts of policies aimed at fostering gender equality on economic growth in Brazil. After a brief review of gender issues in Brazil, this note describes a framework for quantifying the growth effects of gender-based policies in developing economies. The analysis is based on a computable overlapping generations (OLG) model that accounts for the impact of access to infrastructure on women’s time allocation, as well as human capital accumulation, inter- and intra-generational health externalities, and bargaining between spouses. The model is calibrated for Brazil and is used to conduct two experiments, the first involving improved access to infrastructure, and the second a reduction in gender bias in the marketplace. The key lesson of these experiments, among others reported in Agénor and Canuto (2013), is that fostering gender equality, which may depend significantly on the externalities that infrastructure creates in terms of women’s time allocation and bargaining power, can have a substantial impact on long-run growth in Brazil. In recent years, Brazil has been very successful in reducing cation. Although universal coverage in primary education poverty and income inequality. According to World Bank has been achieved, quality at both the basic and secondary data, relative poverty (based on a purchasing power parity levels remains a concern. Growth and private sector activity [PPP] US$2 per day metric) has fallen markedly, from 21 per- continue to be hampered by various barriers and regula- cent of the population in 2003 to 11 percent in 2009. Ex- tions, as well as inadequate infrastructure and a weak busi- treme poverty (based on a PPP US$1.25 per day metric) also ness climate. dropped significantly, from 9.8 percent in 2004 to 6.1 percent Gender inequality remains high, despite some significant in 2009. At the same time, income inequality fell significantly. improvements. In 2003, under his first presidency, Luiz Iná- Between 2001 and 2009, the income growth rate of the poor- cio Lula da Silva created a federal governmental body for the est decile of the population was 7 percent per year, while that purpose of addressing gender equality issues. This also led to of the richest decile was 1.7 percent; as a result, income in- the creation of a National Plan for Women’s Policies (NPWP), equality (as measured by the Gini index) fell significantly, which was adopted in 2004. The NPWP reaffirmed the com- from 0.594 in 2001 to 0.521 in 2011—a 50-year low. Key driv- mitment of the Brazilian federal government and other gov- ers of these achievements have been low inflation; sustained ernmental bodies to incorporating a gender perspective into economic growth, at a yearly average of 4.0 percent during public policies. The NPWP prescribed several concrete ac- 2002–8 and 5.1 percent during 2010–11; well-focused social tions to be implemented by different government sectors in programs; and real increases in the statutory minimum wage. cooperation with the private sector and addressed specific Despite these achievements, inequality remains at rela- needs of mothers, particularly health care before and during tively high levels, and there is still a large gap in access to edu- pregnancy and at birth, as well as child care and education. 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Brazil’s first female president, Dilma Roussef, also promised more women than men. The proportion of women in the when elected in 2010 to make gender equality a priority. At workforce rose from 52.8 percent to 57.6 percent between the same time, however, there has been limited effort to quan- 1998 and 2009, whereas the share of women in wage employ- tify the impacts of gender-based policies on gender inequality ment in the nonagricultural sector rose from 35.1 percent in and economic growth in Brazil. 1990 to 41.6 percent in 2007. The female to male labor force This note, which draws on Agénor and Canuto (2013), participation rate increased from 52.2 in 1990 to 63.9 in provides a quantitative analysis of the long-run impacts of 1995, 66.7 in 2000, and 73.3 in 2010. policies aimed at fostering gender equality on economic However, gender gaps in access to formal employment growth in Brazil, especially through their impact on women’s and market income still persist in Brazil. The proportion of time allocation and intrahousehold bargaining power. This women with formal jobs increased from 41.5 percent in 1999 analysis uses the gender-based OLG model described in Agé- to 48.8 percent in 2009, but it is still lower than that of em- nor (2012) and Agénor and Canuto (2012a). In the model, ployed men, which stood at 53.2 percent in 2009. At the women’s time allocation takes center stage. In addition, wom- same time, women in formal sector employment work less en’s bargaining power is endogenously related to time allocat- than men—an average of 36.5 hours a week in 2009, com- ed by women to human capital accumulation—and thus indi- pared to 43.9 hours for men. Even though there has been rectly to access to infrastructure, which has a direct impact on progress in the share of women employed in the nonagricul- time devoted to home production. This creates an important tural sector, their comparative advantage in education has not channel through which public policy can affect gender equal- been reflected in relative market wages—despite the average ity and economic growth. higher skill level of the female labor force. In 2008, women’s After a brief review of gender issues in Brazil, this note wages were 84 percent of men’s, and the gap increases at high- provides an outline of the model and its calibration. In addi- er levels of education.1 Among those with 12 or more years of tion, two policy experiments are discussed, both with poten- schooling, women earned merely 58 percent of men’s salaries. tially important direct and indirect effects on gender inequal- Brazilian women, even those working full time, continue to ity and growth: the first involves improved access to bear the brunt of time allocated to family chores (Bruschini infrastructure, and the second covers institutional reforms 2007); in 2008, women devoted an average of 25.1 hours per aimed at reducing gender bias in the marketplace. The results week to caring for their families and housekeeping, whereas of these experiments are used to inform broader policy les- men devoted an average of only 10 hours per week to such sons discussed in this note’s conclusion. tasks. The unemployment rate for females consistently exceeds Background: Gender Inequality in Brazil that of males by an average of 4–5 percentage points; the gap In the past two decades, Brazil has made significant progress is up to twice as high for those aged 15–24. According to the in reducing gender inequality. According to the results of a Gender Inequality Index (GII) introduced by the United Na- 2010 study by the Brazilian Institute of Geography and Sta- tions in 2011, Brazil’s rank is only 80 out of 187 countries, tistics (IBGS), illiteracy rates for women aged 15 years and with a score of 0.449—the same as in 2008. The Gender Gap older fell from 20.3 percent in 1991 to 13.5 percent in 2000 Index (GGI) developed by the World Economic Forum and and 9.8 percent in 2008. Brazilian women are now generally produced since 2006 gives similar results; in 2011, Brazil was more educated, with female participation in tertiary educa- ranked 82 out of 135 countries, with a score of 0.668, com- tion significantly exceeding male participation. As a result, pared to 0.654 in 2006.2 the share of the female labor force with tertiary education in- What explains the gender pay gap in Brazil? According to creased from 7.4 percent in 1992 to 8.5 percent in 1999 and some recent economic studies, only a small portion—between 11.9 percent in 2007, compared to 5.3, 6.2, and 7.3 percent 11 percent and 19 percent of wage differentials in the formal for males, respectively. Working women have an average of 8.8 labor force—can be attributed to differences between men years of schooling, while their male counterparts have an aver- and women in their endowments (such as education or expe- age of 7.7 years. This is important because, as discussed in the rience). For the most part, the wage gap appears to reflect dis- next section, educated mothers tend to have greater bargain- criminatory practices and social norms (see van Klaveren et al. ing power within the household over intrafamily allocation of [2009]). The model presented in the next section considers monetary resources, are more able to act on their preference both potential causes—in the form of gender bias in the work- for investing in children, and have a greater impact on family place and mothers’ time allocation between boys and girls. decisions regarding the allocation of children’s time to house- The model also considers a third potential cause, related to hold chores. Professions that traditionally were dominated by access to infrastructure services, and its implications for males, such as law, medicine, and engineering, are becoming women’s time allocation to household production, child rear- more balanced in terms of gender, and some already have ing, human capital accumulation, and market work. The as- 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise sumption under analysis is that poor access to core infrastruc- private capital ratio is relatively low, suggesting that public ture services forces women to allocate a large fraction of their capital remains a relatively scarce factor in Brazil, and that available time to family chores. As a result, they have less time women allocate 20.6 percent of their time to home produc- available to take care of their children—a productive use of tion, 11.7 percent to child rearing, 18.5 percent to human time if it helps to improve children’s health, and if health and capital accumulation (over their lifetime), and 42.1 percent productivity in adulthood depend on health in childhood; to market work. In line with the evidence reviewed earlier, the further their own education; and occupy formal sector jobs.3 degree of gender bias in the work place is estimated at 0.71, By implication, improved access to infrastructure services indicating that women engaged in market activity earn on av- may free women’s time in such a way that they could devote erage about 30 percent less than men. more time to building their own human capital. If bargaining Policy Experiments power between men and women depends (directly or indi- rectly, through wages) on relative human capital stocks of To illustrate the role of gender-related and gender-based pub- men and women, and if higher bargaining power for women lic policy on economic growth in Brazil, and how these poli- translates into more savings and more investment in girls’ ed- cies affect women’s time allocation and bargaining power, two ucation, then the positive growth impacts from closing the experiments were conducted: the first on impacts from im- gender gap could be fairly large in the long run. proved access to infrastructure and the second on impacts from a reduction of gender bias in the marketplace. The Analytical Framework Improved access to infrastructure To study the interactions between gender and growth, the ap- Consider the case of a public policy aimed at promoting ac- proach here draws on the gender-based OLG model of eco- cess to infrastructure by investing in rural roads, power grids, nomic growth described in Agénor (2012) and Agénor and and others: the direct effect of this policy is of course an in- Canuto (2012a, 2013), and summarized in Agénor and Ca- crease in the public-private capital ratio, which therefore pro- nuto (2012b). Key features of the model include: motes growth directly. In addition, this increase reduces (i) Home production combines women’s time allocated to mothers’ time allocated to home production and raises time that activity with infrastructure services; allocated to market work, human capital accumulation, and (ii) In families, fathers have a relatively higher preference for child rearing. The latter is also productive; it leads to im- current consumption, whereas mothers have a higher proved health in both childhood and adulthood. Thus, all of preference for children’s health, thus familywide prefer- these effects also help promote growth and health outcomes. ence parameters for consumption and children’s health Crucially, the increase in time devoted to human capital depend on women’s bargaining power; accumulation raises women’s bargaining power, which trans- (iii) Women allocate their time between four alternatives: lates into a higher family preference for girls’ education and market work, raising children, human capital accumula- children’s health, an increase in the average share of family tion, and home production; income spent on children, and a lower preference for current (iv) The gender gap in the workplace is captured by assuming consumption. The first two effects increase further the that women earn only a fraction of their marginal prod- amount of time allocated to education and child rearing, uct, and differences in economic outcomes between men whereas the last effect contributes to a rise in the savings rate. and women are fundamentally related to gender bias ex- Because increases in the level of income and in the savings rate perienced in the home during childhood; and raise private savings and the private capital stock, there is a (v) Women’s bargaining power depends on the relative levels positive effect on the growth rate of outputs. At the same of human capital of husband and wife, and thus indirect- time, female health in adulthood also improves—as a result of ly on access to infrastructure, which influences women’s receiving more care in childhood and higher government time allocation and thus the time that they allocate to spending on health. human capital accumulation. In quantitative terms, the numerical experiments fo- The model produces equilibrium values for women’s cused on a policy that takes the form of a budget-neutral in- time allocation and an explicit expression for the economy’s crease in government spending on infrastructure investment, growth rate in the long run. These static, long-run relation- from its current value of about 2.1 percent of gross domestic ships are then calibrated and simulated using data for Brazil. product (GDP) to 3.1 percent. Calculations suggest that this The calibration uses a variety of information sources, includ- policy could add between 0.5 and 0.9 percentage points to ing data from the 2009 National Household Sample Survey Brazil’s annual rate of output growth once direct and indirect (PNAD) and the calculations performed in Lopes Ribeiro and effects—most notably through changes in women’s time allo- Marinho (2012). The result of the calibration exercise shows, cation and their bargaining power over family resources—are for instance, that in the benchmark case, the initial public- accounted for. This positive effect could be even higher if the 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise increase in the share of spending of investment is accompa- infrastructure, given the complementarity between infra- nied by reforms aimed at improving the quality of such spend- structure and women’s time in home production. In that ing. Thus, a policy aimed at promoting access to infrastruc- sense, this analysis draws fundamentally on a macro theory of ture could have a substantial impact on the population’s bargaining power in the family. standards of living over a relatively short time horizon. The model was calibrated using a variety of data sourc- Reduction in gender bias in the marketplace es and two experiments were conducted, one on a policy Suppose now that the government introduces antidiscrimina- aimed at promoting access to core infrastructure, and an- tion laws that lead to a complete elimination of gender bias other on an “equal work, equal pay� policy aimed at reduc- against women in the workplace. Women’s “take-home� pay ing gender bias in the marketplace through active enforce- therefore increases, all else being equal. The direct effect of ment of antidiscrimination laws. The key lesson of these this policy (at the initial level of wages) is to raise family in- experiments, as well as other experiments reported in Agé- come. In turn, higher income leads to a higher level of private nor and Canuto (2013), is that fostering gender equality, savings and higher private capital stock, which have a direct which may partly depend on the externalities that infra- positive effect on growth and bring higher tax revenues. Be- structure creates in terms of women’s time allocation and cause changes in the degree of gender bias in the workplace bargaining power over family resources, may have a substan- affect tax revenues and private savings in exactly the same tial positive impact on long-run growth as well as human way, the public-private capital ratio is not affected, and wom- capital and health outcomes in Brazil. en’s time allocation is not affected either. Nevertheless, be- About the Authors cause higher tax revenues lead to higher public spending on health, there is a positive effect on health in childhood and Pierre-Richard Agénor is a Hallsworth Professor of international female health in adulthood. Thus, a reduction in gender bias macroeconomics and development economics at the Univer- leads to an increase in the growth rate of output and improved sity of Manchester, Co-Director, Centre for Growth and Business health outcomes. Cycle Research, International Research Fellow, Kiel Institute for In quantitative terms, the model-based calculations sug- the World Economy, and Senior Fellow, FERDI. Otaviano Canu- gest that an “equal work, equal pay� policy that would ensure to is Vice President of the Poverty Reduction and Economic Man- that women earn a wage that fully reflects their marginal con- agement (PREM) Network. tribution to market production could add up to 0.2 percent- age points to the country’s annual growth rate. Over a suffi- Notes ciently long period of time, this would also have a significant 1. There are also considerable gender differences across re- impact on the population’s standards of living. In addition, it gions. For instance, wage gaps tend to be lower in Rio de Ja- is important to note that the analytical framework from neiro than in São Paulo; the largest gaps are found in the which this estimate is derived does not capture the possibility Northeast. that gender gaps in access to managerial positions and employ- 2. For the GII index, see http://en.wikipedia.org/wiki/Gen- ment may also distort the allocation of talent and women’s der_Inequality_Index; for the GGI index, see http://www. incentives to invest in particular skills, thereby constraining weforum.org/. overall productivity growth. If these effects were to be ac- 3. Housework may affect market wages (and thus the gender counted for, the growth benefits of eliminating gender bias in wage gap) indirectly as well, by influencing women’s choices the marketplace would be even higher. regarding their selection of job characteristics (and thereby Concluding Remarks via job-related compensating wage differentials). Women who spend more time on household chores, particularly dur- Brazil has made significant progress in reducing poverty and ing the working week, may seek out jobs that offer more flex- income inequality in recent years. However, despite this prog- ible work arrangements, such as shorter commuting time or ress and the government’s continuing commitment to incor- greater flexibility in scheduling. Flexible working arrange- porating gender perspectives, gender inequality remains high. ments are likely to be costly to firms, and therefore wages may To determine the long-run impacts of policies aimed at foster- be lower in such jobs to compensate employers. ing gender equality on economic growth in Brazil, this analy- sis applied a gender-based OLG model that accounted for References women’s time allocation between market work, child rearing, Agénor, Pierre-Richard. 2012. “A Computable OLG Model for human capital accumulation, and home production. Crucial- Gender and Growth Policy Analysis.� Working Paper No. 169, ly, bargaining between spouses was assumed to depend on Centre for Growth and Business Cycle Research, University of relative human capital stocks, and thus indirectly on access to Manchester. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Agénor, Pierre-Richard, and Otaviano Canuto. 2012a. “Access to Bruschini, Maria C. A. 2007. “Work and Gender in Brazil in Infrastructure and Women’s Time Allocation: Evidence and a the Last Ten Years.� Cadernos de Pesquisa 37 (September): Framework for Policy Analysis.� Policy Paper No. 45, FERDI, 537–72. http://www.ferdi.fr/Documents-de-travail.html#P45. Lopes Ribeiro, Lilian, and Emerson Marinho. 2012. “Time Poverty ———. 2012b. “Measuring the Effect of Gender-Based Policies on in Brazil: Measurement and Analysis of its Determinants.� Economic Growth.� Economic Premise 85, World Bank, Wash- Estudos Econômicos 42 (April): 285–306. ington, DC. van Klaveren, Maarten, Kea Tijdens, Melanie Hughie-Williams, ———. 2013. “Gender Inequality and Economic Growth in Brazil: A and Nuria Ramos Martin. 2009. “An Overview of Women’s Long-Run Analysis� Policy Research Working Paper No. 6348, Work and Employment in Brazil.� Unpublished, University of World Bank. Amsterdam (December). The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise