Measuring Gender Inequalities
The study of gender as an analytical category in macroeconomics and macroeconomic policy requires appropriate tools—data, statistics, and modeling. But gender measurement issues have been addressed only in the past 30 years or so, and much work remains to be done in this area. Data collection methods are not always gender sensitive, for a number of reasons (World Bank 2001):
Managers, researchers, and technical staff may not be aware of or may lack experience with gender issues.
Surveyors usually interview the household head, who is in most cases a man.
Women may not be able or allowed to attend or speak at community meetings where gender-related issues are discussed, and formal interviews are not the best format for broaching sensitive topics (such as domestic violence).
Despite these issues, the case for measuring gender’s effects is strong. In the past 20 years, national statistics-gathering programs have sought to include gender-related data, and policy makers have begun to recognize the importance of gender analysis in the development and monitoring of public policy. For the most part, however, both groups have focused almost exclusively on gathering and analyzing social and demographic statistics on gender. In fact, gender is an issue related to all statistics concerning individuals—men and women alike—and all statistical departments should be required to collect data on gender. Doing so would require the commitment of top managers in statistics-gathering bureaus and the appointment of gender advisers who would report directly to the chief statistician.
Why Measure Gender?
The World Bank (2007) has identified gender equality as a development goal in its own right, with repercussions for the long-term growth prospects of countries. Figure 2.1 illustrates how equal opportunities in rights, resources, and voice lead to economic growth.
Data on key economic, social, and political inequalities suggest that gender inequality continues. Table 2.1 shows average values of education and health indicators for three country groupings, classified by high, medium, and low levels of human development based on net enrollment as a percentage of eligible population. The average woman’s life expectancy is higher than that of the average man, but the gap narrows in countries that score low on the HDI. Inequalities in health can be attributed to the higher mortality rate of girls and differences in life expectancy that do not accord with biological norms (Stotsky 2006). These “missing women” are a well-documented phenomenon. Comparing the sex ratios of populations with excess mortality among women to the ratios that would have prevailed without discrimination indicates that there were 90 million missing women in the early 1990s, mostly from Asian countries, particularly China and India (Klasen 1994).
Well-developed data sets are crucial to analyzing the links between levels of development and gender equality. Toward this end, a number of organizations — including the World Bank, the United Nations, and the Organisation for Economic Co-operation and Development — have constructed databases that attempt to isolate determinants of gender inequalities. Traditional economic analysis, which concentrated on the market and thus income-earning activities, ignored or underestimated the unpaid, yet valuable, work of many women, including domestic and volunteer work. The battle to include unpaid domestic production in the calculation of gross national product was hard fought, even with theoretical and empirical evidence from new household economics and the domestic labor debate. In the 1970s, however, academics, government representatives, and international organizations such as the International Labour Organization and the United Nations spearheaded a revaluation of women’s work, redefining “economically active” to include unpaid production. At the methodological level, many countries made a commitment to improve the accuracy with which women’s participation in the labor force was counted, and new techniques estimated the value of home production. In particular, structural adjustment programs of the 1980s resulted in a transfer of costs from the public sector to a private household. The underlying assumption was that women would be able to absorb these shocks by working more and “making do” on limited incomes (Elson 1993).
In summary, quantifying gender is a necessary step toward making gender-related work visible and ensuring that macroeconomic policy and models foster gender equality. Quantifying gender data warehouse facilitates a more accurate analysis of the unequal distribution of domestic work, productivity changes in unpaid production, shifts in domestic work and family welfare as a result of changes in family income and employment status of household members, and gross domestic product growth.
Gender Databases
The compilation and use of gender statistics have benefited greatly from the development of databases concerned with gender inequalities and gender issues, including the following:
• participation in decision making
• gender attitudes
• participation in elections
• entrepreneurship
• domestic violence
• poverty
• informal employment
• time use
• school attendance.
Gender-related data sets identify gender inequalities in terms of inputs and outcomes. A number of gender-related databases are updated continuously as issues are identified and data becoming. The World Bank maintains the Gender Stats database, which offers statistical and other data in modules at the regional and country levels. Modules include poverty; basic demography; human development (education, health and nutrition, and population dynamics); socioeconomic roles; access to economic resources; political participation; and effects of programs and policies. Coverage is sparse for some indicators. The database draws its statistics from national statistics bureaus, UN databases, and surveys conducted or funded by the World Bank. Gender Stats also includes categories for which limited or no data exist to illustrate the importance of collecting such data in the future. For example, in its educational access and attainment fields, Gender Stats includes variables such as progression to grade 5 (percentage of cohort) among boys and girls, primary school completion rate (percentage of age group) among boys and girls, and youth literacy rate (percentage of people 15–24) among young men and women, with a view to populating these variables when data become available.
The report suggests that detailed indexes, such as an individual disposable income index, be created by gender. The report also recommends modifying the GDI to reflect specific problems faced by women in developing countries—for example, access to nutrition, housing, and clothing—and to identify sources of income (labor and rent, for example). The review by Hailelul and others (2009) criticizes the GEM for not capturing gender empowerment at the household level and for failing to measure empowerment issues such as sexuality, religion, culture, and women’s rights. The GEM also fails to include some non economic dimensions of decision-making power, and it relies on international rather than national databases.
There are 162 countries in the GID database, although data on social institutions are available for only some of these countries. The GID focuses on gender-related differences; hence most of the variables are measured in terms of ratios. Figure 2.4 shows the regional indexes of discrimination against women for the four measures of social institutions identified in figure 2.3.
The World Bank, the ILO, and the World Health Organization provide the data on access to resources, economic development, and the economic role of women. The OECD also uses the GDI and GEM. Data on social institutions are not readily available and are therefore derived from many sources. They include both quantitative and qualitative variables from Amnesty International; BRIDGE, a research and information service of the Institute for Development Studies (IDS) that specializes in gender and development; the Women in Development Network (WIDNET); AFROL, a news agency that concentrates on Africa; and a study commissioned by the French Parliament (Lang 1998), among others. The database also draws on gender documentation from the various donor agencies, such as the Canadian International Development Agency (CIDA).
Failure to quantify gender relations can allow gender inequalities to go unchecked, with adverse implications for a country’s economic growth and development. In this chapter, we first make the case that fully appreciating gender as an analytical category requires appropriate data and tools. Gender measurement issues have been addressed only in the past 30 years, and databases measuring gender equality are far more recent. Although measuring gender remains a work in progress, significant efforts have been made in recent decades. Previously, economists simply disaggregated socioeconomic data by male and female. Doing so is just a first step toward measuring gender; it does not capture the effects of gender relations. The development of a number of databases marked another step forward. These databases incorporate key gender issues, such as participation in decision making, gender attitudes, participation in elections, entrepreneurship, domestic violence, poverty, informal employment, time use, and school attendance. This chapter shows how these databases identify gender inequalities in terms of inputs and outcomes and why this information is useful for policy makers. It concludes by depicting a number of potential approaches for the macroeconomic modeling of gender relations.
Macroeconomic Aggregates
Differences in the behavior of men and women may lead to different macroeconomic outcomes, particularly for aggregates such as private consumption, successful savings and investment, and the composition of government expenditure. As a result, gender budgeting has been an important issue in most countries over the past few decades. Chapter 8 examines the gender budgeting methods and outcomes in a number of countries.It explores gender’s influence on consumption and saving. We first contextualize the discussion by examining the linkages between household and its gender composition and macroeconomic aggregates.
The neoclassical model traditionally treated the household as a single decision-making entity. It could therefore not deal with the fact that individual members of the household make their own decisions. As a result, individual preferences within a household were modeled separately. In the 1960s, some economists began to view this model as outmoded and turned to new household economics. NHE brought greater flexibility to the structure of policies aimed at the household level despite the added complexity of the modeling process involved and the fact that it did not always yield policy responses. Contributed to initially by Schultz (1961, 1974) and Becker (1964, 1965), this theory treats the household as a multiple decision-making entity in which individuals make decisions to invest in both human capital (their offspring and themselves) and nonhuman capital.
Figure 3.1 illustrates the gender dynamic at the level of the household, outlining two approaches. One is a unitary model, based on a neoclassical framework in which resources and responsibilities within the household are pooled to achieve a common set of goals. In this model, gender has no effect on savings or consumption. The second is a collective model, in which individuals within the household differ in their preferences, rights, responsibilities, and resources, as suggested by NHE models. Here two possible outcomes are noted. Where there is a pooling of resources and responsibilities, we expect no gender effects on macro- economic variables. But if bargaining rather than cooperation characterizes household decisions—that is, if individuals within the household use the resources they control to pursue their own priorities—then policies that affect the distribution of resources within a household or shift the balance of power will have clear implications for gender equality, macroeconomics, and family welfare.
Deaton (1989) finds no gender bias in medical and education spending in Côte d’Ivoire and only a small but insignificant effect in favor of boys in Thailand. Across several developing countries, Glick, Saha, and Younger (2004) find no discrimination in health care or education spending among boys and girls in their review of the literature and analyses of developing country data sets. Alderman and Gertler (1997) find a gender bias in favor of boys with regard to demand for medical care in Pakistan. Macroeconomic policy that targets the exclusion of girls from education needs to consider the price and income elasticities of demand for education. As Stotsky (2006) notes, relative price increases for education would adversely affect girls, and price decreases would disproportionately benefit girls: “Higher income elasticity of demand for female education and health care implies that economic prosperity would disproportionately benefit women by expanding their access to these services, while recessions would have a disproportionately negative effect” (Stotsky 2006, 11). Sound macroeconomic policies that include an appropriately valued exchange rate are key to reducing gender disparities in education and health.
Lampietti and Stalker (2000) examine consumption expenditure and female poverty. Referencing more than 60 poverty tests carried out by the World Bank and other published and unpublished sources, the authors note the following:
• Poor women have higher fertility rates, higher maternal mortality rates, lower-birthweight babies, and less access to qualified or modern health care during pregnancy than nonpoor women do. These differences are found in both low- and middle-income countries.
• There is no conclusive evidence that poor women are worse off than poor men in terms of food allocation and anthropometric status (a measure of food intakes).
• Low- and middle-income countries vary in quantity and kind of educational opportunities they offer boys and girls. Several findings emerge from analysis of 22 poverty assessments that included analysis of gender and education. Girls are worse off than boys in Djibouti, Egypt, Kenya, Nigeria, Pakistan, and Yemen. Below the poverty line, girls are worse off than boys in Algeria, Bolivia, Côte d’Ivoire, India, Lao PDR, Malawi, Morocco, and Zambia; boys are worse off than girls in Lesotho. The findings are inconclusive or show no difference in Madagascar, Mauritania, Nicaragua, Rwanda, Sri Lanka, and Tanzania.
It is now generally accepted that women have a stronger preference than men for spending on goods and services that increase the human capital of their children. The macroeconomic implications of this statement are clear. Greater expenditure and investment in human capital will ultimately influence economic growth. Furthermore, expenditure on services such as education, health, and nutrition is less responsive to variations in income, thus bringing about greater stability in expenditure in economies in which women have greater control over household purchases. Policies that increase women’s control of household spending should therefore strengthen macroeconomic growth. At the individual level, gender affects consumption of education, health, and nutrition. A number of studies attest to the discrimination faced by women and girls in education. The chapter examines some policies that could increase parity in education for women and girls in developing countries.
Statistical Relationships
A number of measures of gender equality are positively correlated with economic growth, as measured by per capita income (table 4.1).
Stotsky (2006) separately plots both the UN Gender-related Development Index and the Gender Empowerment Measure against the log of per capita income for 41 countries chosen randomly from the sample of International Monetary Fund member countries. Both indexes show positive and nonlinear relationships with income, suggesting that increasing income leads to greater gender equality in both economic terms and political terms (figure 4.1). Several studies examine the empirical relationship between single indicators of gender equality and per capita income. Stotsky (2006) examines the relationship between the log of per capita income and the ratio of girls to boys in primary education, the ratio in secondary education, and life expectancy (see her paper for the list of countries and detailed results). She concludes that the relationship between single indicators of gender equality and per capita income is less clear than the relationship between income and indexes such as the GEM and GDI.
All measures of gender equality are positively correlated with per capita income, ranging from 28 percent for gender differences in secondary education to about 60 percent for women’s life expectancy minus men’s life expectancy, women’s economic rights, and women’s rights within marriage. The neoclassical/modernization approach has its roots in neoclassical theory, which relates economic growth to capital accumulation and savings, which in turn depend on the distribution of resources, income, and capabilities. As previous chapters have shown, gender inequalities affect these inputs at the micro economic level; gender inequalities can therefore affect economic growth at the aggregate level. As Stotsky (2006, 18) notes, “the neoclassical approach examines the simultaneous interaction of economic development and the reduction of gender inequalities. It sees the process of economic development leading to the reduction of these inequalities and also inequalities hindering economic development.”
Discriminatory social mores and norms may change in the long run as well, improving gender equality—or they may remain intact, working against economic modernization and growth to maintain inequality. For example, some observers maintain that “enduring patriarchal institutions will prevent gender equality even in the face of economic advancement” (Jütting and others 2006, 7, citing Marchand and Parpart 1995 and Parpart 1993). They point to Saudi Arabia—a high-income country with poor gender equality—as evidence. Others, such as Ramirez, Soysal, and Shanahan (1997), suggest that even in societies with strong patriarchal institutional legacies, there is some evidence that globalization has displaced traditional gender inequalities.
According to Walters (1995), the inclusion of the human capital element in endogenous growth theory paves the way for a consideration of gender in four ways:
• It opens the way for time to be incorporated into the production of labor inputs.
• It recognizes education and other influences on human capital accumulation and their relationship to growth.
• It allows for the possibility of trade-offs between government fiscal policies, including spending programs, and growth.
• It facilitates income distribution through the positive effect of human capital investment on growth.
Dollar and Gatti (1999) conclude that gender equality and growth are mutually reinforcing: “Societies that have a preference for not investing in girls pay a price for it in terms of slower growth and reduced income” (Dollar and Gatti 1999). They present a win-win scenario for gender inequality and growth: reduce gender inequality, improve growth. Using data from more than 100 countries over three decades, they investigate the relationships among gender inequality, income, and growth. They note that only 5 percent of adult women in the poorest quartile had any secondary education in 1990, only half the level of men. In contrast, 51 percent of women in the richest quartile had at least some secondary education, representing 88 percent the level of men (see table 4.3 for details).
Empirical results from India based on Volart’s model suggest that discrimination slows economic development. A 10 percent increase in the ratio of female to male managers in India would increase total output per capita by 2 percent, whereas a 10 percent increase in the ratio of female to male workers would increase total output per capita by 8 percent. The effects of gender discrimination are stronger in certain sectors of the economy, in particular sectors requiring higher skills. Lower ratios to male workers reduce output in both agricultural and non agricultural sectors, whereas lower ratios of female to male managers reduce output in non agricultural sectors only. More worrisome are the indications in Volart’s study that suggest that economic growth does not inherently reduce gender inequality. Even richer states continue to have lower ratios of female to male labor participation, suggesting the continued presence of discriminatory social norms. In such cases, targeted policies, such as those that encourage women’s education and in the labor market, are critical to change entrenched social norms and promote the development.
Dollar and Gatti (1999) consider whether gender inequality affects growth. They focus on the difference between male and female secondary attainment. The growth equation derived by Dollar and Gatti (1999) is typical in its attempt to explain income growth as a function of some initial conditions, including per capita income and policies that affect the business environment; it is atypical in that it includes gender as well.
Inequality in Labor Market
Gender inequality manifests itself most obviously in labor markets. Women face greater barriers than men to finding decent and productive work. Women’s labor force participation has risen in some but not all regions. Moreover, the quality of work and working conditions have not always kept pace with increases in participation. Persistent gender inequalities in wages suggest that the labor market is not operating freely.
Women continue to be over represented in unpaid and informal work. The agricultural and service sectors employ more women than men, yet they often pay women less for similar work. These inequalities are especially significant in developing economies, where women are more likely to be among the working poor—those who work but do not earn enough to lift themselves above the $1 a day poverty line. Indeed, women make up at least 60 percent of world’s working poor (ILO 2004).
The gender division of labor also remains sharp in industrial and urban societies. Worldwide most jobs are dominated by one gender or the other. For example, women make up 41 percent of the non agricultural labor force in the Organisation for Economic Co-operation and Development countries but are disproportionately shunted into service jobs: women constitute 62 percent of service workers, compared with only 15 percent of production workers (Anker 1998, 171). The following sections present data that highlight gender inequalities in the labor market and show the progress that has been made in some areas and regions. The discussion extends beyond the data to include a commentary on how gender inequality in labor markets affects macro economic policy. The chapter also highlights the vulnerability of women in the world of work and includes a discussion of the third Millennium Development Goal (MDG 3), which focuses on redressing gender disparities and empowering women.
Collier (1994) considers possible reasons why gender inequalities arise in the labor market (in Africa) by focusing on four processes that lead women to face constraints on their economic activity that are different from those facing men. First, women may encounter discrimination outside the household. According to the International Labour Organization (2007), the segregation of occupations by sex is changing, albeit slowly; additional investment in women’s education and training is needed to keep change moving forward. Between 1996 and 2006, the share of women in wage and salaried work increased from 42.9 to 47.9 percent (ILO 2007). Still, fewer women than men have paid work, especially in the world’s poorer regions. Interestingly, time-use studies suggest that women work more total hours than men (Stotsky 2006).
Table 5.1 shows the change in the LFPR for men and women over a 10-year period. The total female labor force worldwide was 1.2 billion in 2006, up from 1.1 billion a decade earlier. Over the 10-year period, the female LFPR declined to 52.4 in 2006, from 53 percent in 1996. The ILO attributes this decline to the increasing numbers of young women receiving education and the increasing share of older women in the labor force. More recent data indicate that the LFPR in 2008 was 52.6 percent for women and 77.5 percent for men (ILO 2009). Although the gap is narrowing, it still stands at 25 percentage points. Women made up 40.5 percent of the global labor force in 2008, up from 39.9 percent a decade earlier (ILO 2009).
Table 5.2 shows the regional LFPRs for men and women for 2007 and 2009. Globally, the LFPR is estimated to have remained the same for women and declined marginally for men. Across regions, the estimated increase in the LFPR for women is highest in Sub-Saharan Africa, the Middle East and North Africa, and Latin America and the Caribbean (a 0.6–0.7 percentage point increase). The increase for the other regions is marginal (0.1 percentage point), with a decrease projected in East Asia (0.3 percentage points). For men, LFPRs fell, except in Southeast Asia and the Pacific, the Middle East and North Africa, and Sub-Saharan Africa, which all showed marginal projected increases of at least 0.1 percentage point.
As the ILO (2007, 3) notes, “taken on their own, rising or high labor force participation rates do not necessarily mean that labor markets are developing positively for women.” Is gender equality in the labor market improving? Are wage gaps and discrimination reducing LFPRs? Are women finding the type of work they want? Does the fact that more women are in school account for the lower female LFPR? What are the characteristics of women’s work compared with men’s work? The following sections shed some light on these questions.
Although data across countries and regions are limited, it is clear that wage gaps between women and men persist, even in occupations dominated by women. One study of six occupational groups estimates that in most economies, women earn 90 percent or less of what their male counterparts earn (ILO 2007). Wage inequality is found across all occupations, predominantly in low-skilled occupations but also in highly skilled ones. Corley, Perardel, and Popova (2005), quoted in ILO (2007), note that the average female wage is only 88 percent of the average male wage in occupations such as accounting and computer programming. Oostendorp (2004) shows that in developing countries, globalization has led to an improvement in wages in low-skilled occupations, in which women are more highly represented, but a widening gender gap in wages among high- skilled occupations, in which men are more highly represented.
Figure 5.7 examines wage data for six occupational groups in industrial economies, Central and Eastern Europe and the CIS, and developing economies. Male and female hotel receptionists in Central and Eastern Europe, the CIS, and developing economies enjoy wage parity, and female nursing professionals in industrial economies, Central and Eastern Europe, and the CIS earn slightly more than their male counterparts. Wage rates are significantly higher for female first-level teachers, computer programmers, and accountants in Central and Eastern Europe and the CIS than for men in the same professions; this gap reflects the historically greater wage equality in the planned economies of Central and Eastern Europe and the CIS, which may narrow or disappear once the few women who have successfully managed the transition process retire (ILO 2007). In all other professions and regions compared, women earn less than men. Persistent inequalities in wages indicate that the labor market is not operating freely. Differences between men’s and women’s outside obligations—and, by extension, bargaining power—may help explain this ongoing problem. For example, the reservation wage—the lowest wage a worker will accept for a particular job—is often lower for women than it is for men. This may reflect the fact that women’s family obligations decline with their mobility; unable to move for a better-paying job, they end up accepting lower wages close to home. As noted above, in developing economies, for example, women receive lower pay than men for the same agricultural work.
At the global level, male and female shares of employment by sector follow a regular pattern, with women employed predominantly in agriculture and services (figure 5.8). Out of the total number of employed women in 2006, 40.4 percent worked in agriculture and 42.4 percent in services. Of all working men, 37.5 percent worked in agriculture and 38.4 percent in services. This pattern was unchanged in 2008, with 35.4 percent of women employed in agriculture compared with 32.2 percent of men. The rates for service employment were 46.3 percent of women and 41.2 percent of men (ILO 2009). Meanwhile, 17.2 percent of all women worked in industry in 2006, a proportion that increased marginally to 18.3 percent in 2008.
One indicator of progress toward MDG 3 is an increasing share of women in wage employment in the non agricultural sector. Although no target has been set, this change should accompany economic development as people move from being contributing family workers and own-account workers to being wage and salaried workers (ILO 2007, 10). Figure 5.9 examines the change in the share of women in non agricultural wage employment (and the proportion of seats in parliament held by women) by region.
Data disaggregated by gender on status of employment have only recently been made publicly available. The status of employment indicator places workers in one of three categories: wage and salary workers (employees), self-employed workers, and contributing family workers. In doing so, it provides a measure of the progress of female workers.
Globalization and Gender Relations
Globalization affects the living standards of men and women through several transmission channels. Trade liberalization, for example, increases the flow of goods and capital across countries and contributes to economic growth. Between 1970 and the late 1990s, trade in goods and services as a proportion of world gross domestic product (GDP) increased by about 50 percent, thanks in large part to an increase in exported manufactured goods from developing countries in South and East Asia. The increase coincided with rapid GDP growth in these countries.
The opening of an economy’s borders has direct implications on the flow of labor both within and across economies. There are three main hypotheses as to why and when women find more employment opportunities under globalization. The buffer or reserve army hypothesis holds that more jobs for women are available during periods of labor shortages following economic expansions; these jobs are lost during recessions. According to the segmented market hypothesis, women find more employment when output in the sectors in which they are over represented rises more rapidly than output in the rest of the economic sectors. The substitution hypothesis posits that women’s employment opportunities increase when they gradually replace men in jobs previously considered “male jobs.”
Low labor costs, especially for women, may partially explain the outsourcing of services to developing countries such as China and India. Multinational corporations that relocate to developing countries provide employment opportunities for women, exploiting their willingness to work for low wages. To prevent multinational corporations from moving out of their country, some developing governments do not adequately enforce labor regulations, reinforcing lower wages for women.
Globalization also affects the flow of labor across economies. Historically, skilled labor migrated only across developed countries. Globalization, however, has resulted in a marked increase in skilled labor migration from developing countries to developed countries. Women’s lower skill levels make them less mobile than men. That said, they are migrating in larger numbers than ever before. They tend to have lower-paying jobs, however—in fields such as domestic work and hospitality services—and earn less than work men, even given the same education and skills.
Men tend to migrate abroad because of high unemployment rates locally, and they tend to remit larger amounts than women. These remittances go primarily to their wives for taking care of their children while they are away as well as to local income-generating activities. For their part, women migrate to acquire a greater level of power within the household, often remitting a greater proportion of their income than men. These remittances go not only to their immediate family but also to a wide range of relatives, with all of the money earmarked for family needs, including education and health.
The World Bank estimates that in 2004 migrants officially remitted almost $124 billion—twice as much as official development assistance. About 40 percent of Ghanaian and Nigerian migrants in the United States; 50 percent of all Latin American and Central Asian migrants; and 75 percent of Filipino and Indonesian migrants in Southeast Asia who remit money are women. Studies of the pattern and motivation of remittances indicate a positive impact for the household in terms of increased consumption, property investment, and better education and health care (Kireyev 2006).
The average cost of sending $200 has been reduced from an average of 15.0 percent prior to 2000 to 5.6 percent in 2006 (figure 6.1). Costs have fallen as a result of increased competition in the industry, which has stemmed in part from an increase in the number of money transfer organizations entering the market. The level of literacy among people sending remittances is very low; customers have very little information on the best money transfer organizations in their locality and the services they offer. Further cost reductions would stem from greater transparency in the industry, making it easier for customers to compare services.
Trade-related gains in employment for women in developing countries have occurred in export-processing zones, in larger firms (that subcontract work), and in the informal sector. In Ecuador, for example, women benefited as flower exports increased. In Bangladesh about 2 million jobs were created in the textiles and apparel industry, most of them filled by women. Women still suffer from long hours, job insecurity, unhealthy working conditions, and low pay. Table 6.1 provides data on the share of women employed in export-processing firms in developing countries during 2003.
The monetary and fiscal policies that countries use to respond to recessions can also disproportionately affect women. For example, decreases in tax revenues and official development assistance lead to cuts in public sector budgets. Reductions in spending on health and education reduce women’s and girls’ access to basic services. Girls may withdraw from schools to help with household work during times of economic crisis, reinforcing gender gaps in education. Higher unemployment and lower household incomes force women to turn to vulnerable and informal employment (including care giving, which is provided mostly by women and girls). These coping strategies undermine long-term economic development.
Among the 193 million people unemployed in 2008, 112 million were men and 81 million were women. The International Labour Organization (ILO 2009a) estimated that the economic crisis would increase the number of unemployed women by up to 22 million over 2007–09. The unemployment rate for women was 6.1 percent in 2008 and was estimated to rise to at least 6.7 percent in the most optimistic scenario and to 7.3 percent in the most pessimistic scenario in 2009 (table 6.2). In most regions, particularly Latin America and the Caribbean, South Asia, and the Middle East and North Africa, where women often face higher barriers in the labor market, unemployment as a result of the economic crisis was expected to hit women harder than men. Only in East Asia, Central and Southeastern Europe, and the Commonwealth of Independent States was the opposite true. In East Asia unemployment among men was estimated to reach 5.3 percent, 1.3 percentage points higher than the 4.0 percent rate predicted for women in the worst-case scenario for 2009.
Notes
1. The relationship between the formal and informal sectors can be caused by (a) the dualistic features in developing countries, where the formal and the informal sectors exist separately, with no direct links between the two; (b) structural factors, where the subservient informal sector is a source of cheap labor and goods for the wealthier and powerful elites; and (c) the high burden of regulations, which cause enterprises to resort to informality as a response to over regulation and control by bureaucratic governments that enact rules and regulations but do not enforce full compliance.
2. See Stotsky (2006) for a review of studies showing the positive impact of trade liberalization on women’s employment.
3. Although export-oriented firms are concentrated in many industries, they are dominant in clothing and textiles.
Gender and Finance
Access to finance or credit is crucial for improving economic well-being; gender discrimination has hindered access for women in this area. The literature on this topic is slim and falls under umbrella of the micro foundation approach to macroeconomics. It is concerned with the micro level behavior of individuals that leads to macroeconomic effects of growth and development.
The economic potential of women entrepreneurs is curtailed when barriers—social values, religious norms, cultural heritage, and institutional practices—hinder their access to finance. Although women entrepreneurs have a key role to play in strengthening and broadening the base of the macro economy and in bolstering international competition, they may face discrimination in obtaining financing to grow their enterprises.
A number of studies point to the economic and social benefits of facilitating women’s access to credit. Figure 7.1 examines the pathways by which increased gender equality in households, markets, and society translates into current and future economic growth. Increased gender equality in households, markets, and society increases women’s access to markets, education, and health and gives them greater control over decision making in the household. A number of studies identify the significant positive relationship that extra income in the hands of women has on child welfare. Thomas (1990, 1997) shows that extra income for women in Brazil increased child survival by a factor of 20 compared with extra income for men and that the effects on child nutrition were four to eight times as great. Narain (2009) notes that a mother’s access to credit directly affects the household, with positive implications for children’s health and education.
Access to finance and credit empowers women. This empowerment is manifest in the form of increased earning capacity and control of household assets, resulting in greater autonomy and decision making within the household (Kabeer 1998; Khandker 1998; Pitt and Khandker 1998). The spin-offs from greater earning capacity, control over household assets, and voice in decision making are felt at the macro-scope level. The provision of credit for women helps reduce poverty by enhancing the productivity of their enterprises and the profitability of other enterprises in which they invest (Kaur 2007). Some studies find that greater empowerment arising from borrowing enhances a woman’s ability to sell assets without asking her husband’s permission; such empowerment increased husbands’ acceptance of their wives’ participation in market-based economic activities (Agarwal 1997). Women’s increased labor force participation, productivity, and earnings enhance consumption and income, which lead to city growth and poverty reduction.
In many parts of the developing world, access to finance and capital markets is considered a major obstacle to the inception or economic growth of women-owned micro- and small enterprises. Table 7.2 provides a taxonomy of possible providers of financing for such enterprises.
Women use their own or their spouse’s savings to finance their businesses. In Tanzania, for example, 66.0 percent of women used their own funds to start their business, 32.8 percent borrowed money from their spouse, 21.1 percent used credit from other family and friends, 8.6 percent received credit from a microfinance institution, 3.9 percent used bank credit, and 0.8 percent used credit from a moneylender. As businesses grew, additional capital needs were met by owners’ savings (78 percent), micro-finance credit (25 percent), and bank credit (10 percent) (ILO 2005c) (sums are greater than 100 because borrowers used more than one source of financing). A 1999 survey of MSEs in Kenya finds that the main sources of start-up capital are personal savings and family funds (90.4 percent of initial capital and 80.0 percent of additional capital).
Semiformal financing is often carried out through rotating savings and credit associations (ROSCAs). A typical ROSCA comprises 5–10 members. At regular meetings all members contribute a fixed amount; every member then gets a turn as a recipient of the contributions. In some cases, outside cash is brought in through a loan from a microfinance institution and an interest rate is charged to borrowers. The ILO (2005a, 23) notes that ROSCAs “do not work as well when external capital is introduced, as this tends to make the lump sums too large in relation to the members’ capacity to repay.” In Kenya an estimated 76 percent of ROSCA clients are women.
Nongovernmental organizations (NGOs) play a key role in providing access to finance, both directly as lenders and indirectly as investors in microfinance institutions. The Women Economic Empowerment Consort (WEEC) in Kenya offers financial services (savings and loans) and training in business skills to women in the Kajiado District of the Rift Valley and two districts of Central Province (ILO 2005b). NGOs are also involved in microlending in Tanzania. Most members of the Tanzania Gatsby Trust (TGT) and the Zanzibar Fund for Self-Reliance are women. TGT’s client base is 80 percent women; of the approximately 4,000 members of the Zanzibar Fund for Self-Reliance, 70 percent are women (ILO 2005c, 32).
At the policy level, politicians and the development community have embraced microfinance, with the predictable result that some of its merits have been oversold. In reality, most microfinance institutions are weak, heavily donor dependent, and unlikely to ever reach scale or attain independence. To achieve their full potential to serve poor households, they need to fully integrate with the mainstream financial system. Doing so requires financially sound, professional organizations capable of competing with commercial banks, accessing commercial loans, collecting deposits (which requires a license), and growing significantly. Globally, women constitute a disproportionately large percentage of microfinance institutions’ client bases. Women make up more than 80 percent of client membership of the 34 largest micro credit institutions in the world (Morrison, Raju, and Sinha 2007).
As part of a broader effort to raise gender awareness and mobilize women, micro credit institutions could provide an entry point to strengthening women’s networks and mobility and increasing their knowledge, self-confidence, and status in the family. Understanding of the benefits of access to microfinance remains insufficient, and it is unclear whether gender discrimination limits women’s access to finance. Indeed, the issue of access is difficult to separate from related issues, such as gender differences in usage of formal and informal financing, titling, access to collateral, the power of the household head, education level, and credit history.
In contrast, Essel’s (1996) study of the Kakum Rural Bank in the central region of Ghana finds that men had more access to credit from this rural bank than did women. Essel reports that institutional and cultural factors played a role in this gender bias in credit allocation. Institutional factors include banks’ rigid demands for collateral. Social and economic factors include (a) women’s fear of taking risks (as perceived by women themselves); (b) women’s lack of awareness of credit (perhaps because of their low education levels), leading to reduced demand for credit; and (c) skewed ownership of traditional resources (which can be used as collateral) in favor of men.
Among the social and economic factors that woman entrepreneurs identify as limiting their access to credit is their inability to accumulate the savings required to start a business. Women with low levels of education are unlikely to have accumulated savings from previous employment; women who do not have property rights are unlikely to meet collateral requirements.
Governments’ new emphasis on gender equality, commonly called gender mainstreaming, ensures that the goal of gender equality stays central to all activities, policy developments, research, advocacy, dialogue, legislation, resource allocation, planning, and program implementation and monitoring (Sarraf 2003). As the definition indicates, gender main streaming goes beyond allocating funds to gender-specific programs or projects under a government department dedicated to women’s affairs. It refers to a set of policy guidelines and analytical tools that all government ministries can use to generate feasible gender-aware policies. As Stotsky (2006, 1) notes, “To be more useful, gender budgeting should be integrated into gender budget processes in a way that generates tangible improvements in policy outcomes.”
Governments’ new emphasis on gender equality, commonly called gender mainstreaming, ensures that the goal of gender equality stays central to all activities, policy developments, research, advocacy, dialogue, legislation, resource allocation, planning, and program implementation and monitoring (Sarraf 2003). As the definition indicates, gender main streaming goes beyond allocating funds to gender-specific programs or projects under a government department dedicated to women’s affairs. It refers to a set of policy guidelines and analytical tools that all government ministries can use to generate feasible gender-aware policies. As Stotsky (2006, 1) notes, “To be more useful, gender budgeting should be integrated into gender budget processes in a way that generates tangible improvements in policy outcomes.”
Consider the sample objective of “expand primary education.” The GRGB approach challenges the assumption that an expansion in primary education will be distributed equally across population. A gender responsive budget therefore responds to the inequalities in primary education for girls. In practice, the GRGB approach demands many resources, including pre- and post budgeting tools and the training of budget officers. Moreover, the gender surveys necessary for planning are limited in their scope and coverage. These concerns represent serious limitations for developing countries. Although many international organizations facilitate the process by providing technical and financial help, “[such programs] cannot cover the cost of training of a vast number of staff in the government budget preparation process, especially at the level of spending agencies of central government and several local governments as well as the cost of conducting meaningful surveys” (Sarraf 2003, 11). Sarraf (2003) emphasizes the importance of pre- and post budgeting tools for the introduction of GRGB (box 8.2). Pre budgeting tools include gender cost-benefit analysis and emphasize a participatory approach to budget preparation, reflecting the influence of government expenditure on the female population. Post budgeting tools focus on the impact of government programs for the female population and the contribution of such programs to improving gender equality; these tools also influence future pre budgeting tools.
Box 8.2
Analytical and Technical Tools of Gender-Responsive Government Budgeting
The analytical and technical tools of gender-responsive government budgeting include the following:
• Gender-aware medium-term economic policy framework. This tool assumes and advocates for the presence and integration of a strong gender mainstreaming policy in developing the medium-term fiscal policy framework, through engaging local government authorities, traditional rulers, civil society organizations, NGOs, community-based organizations, and donor agencies on both the revenue and expenditure sides. The tool is based mainly on the evolving notion of participatory budgets, which involve beneficiaries and affected groups in the design and implementation of policy, programs and projects, decentralization of financial authority, the empowerment of local communities, and cooperation with key stakeholders.
• Gender-aware policy appraisal. This tool serves to appraise from a gender perspective, the policies and programs funded through the budget, which asks, “In what ways are the policies and their associated resource allocations likely to reduce or increase gender inequality?” The tool refers mainly to the annual budget preparation process by government officials engaged in the budget management system.
• Gender-disaggregated beneficiary assessment. This tool is a means by which the voice of the citizen can be heard. In these exercises, the actual or potential beneficiaries of public services are asked to assess how far public spending is meeting their needs, as they perceive them. A gender-disaggregated beneficiary assessment can be conducted through opinion polls, attitude surveys, group discussions, or interviews.
• Public expenditure incidence analysis. This tool estimates the distribution of government expenditures among men and women directly involved in government operations or immediate beneficiaries of government programs.
• Gender-aware budget statement. This tool is the government report that reviews the budget using some of the above tools and summarizes its implications for gender equality with different indicators, such as the share of expenditure targeted to gender equality, the gender balance in government jobs and training contracts, or the share of public service expenditure used mainly by women.
Source: Sarraf 2003, 9.
Since the mid-1980s, governments and civil society in more than 40 countries have tried gender budgeting, with international organizations playing a supporting role. These efforts have focused predominantly on expenditure, although some have focused on revenue. Reforms differ across countries, reflecting circumstances. Stotsky (2006) references Budlender and Hewitt (2002) as providing the most comprehensive survey on gender-budgeting initiatives. Women’s Budget Group in the United Kingdom comments on the fiscal policies in each annual budget. In Mexico NGOs work with federal and state governments to support gender equality and poverty reduction with academic research.
This volume provides a framework for considering the intersection of gender relations and macroeconomic policy: engendering macroeconomics. The subject was first broached only recently; this volume presents the progress that has been made in a few years. The analysis is timely given the focus on Millennium Development Goal 3, which deals with promoting gender equality and empowering women.
In explaining why engendering macroeconomics is important, we discussed the data and tools available to examine gender relations. We noted that statistics and modeling are critical in making explicit the inequalities facing men and women. Only by having access to these data are policy makers fully equipped to inform policy, measure its effectiveness, and monitor the progress being made. Progress has been made since the 1970s in reevaluating women’s work to account for unpaid work and voluntary work. Gender-related databases have been used to identify gender inequalities in terms of inputs and outcomes.
Engendering macroeconomics remains a work in progress. Despite significant research on the interrelationship between women’s empowerment and economic growth and development, there is a need for further analysis. Persistent gender inequalities—particularly in developing economies and with respect to labor market indicators—represent a brake on economic growth and development. Women’s empowerment policies need to be mindful of household dynamics and their broader macro economic impact. Insufficient attention has been directed to these issues.
Much has been made of the improvements in child welfare stemming from women’s empowerment within the household—justifiably so. But much remains to be learned about the role of other causal factors, such as parents’ education level; the longevity of improvements; and their implications for growth and development. Most studies on gender bias in savings focus on developed countries. By contrast, very little information on gender-based differences in savings is available on developing economies. The lack of information has implications for women’s access to finance and their ability to borrow. More information and better analysis of the information already available are needed.
More broadly, domestic work and child-related activities are largely unaccounted for in national income statistics, making it hard to interpret cross-country gross domestic product statistics in terms of overall welfare. Future research focusing on ways to account for women’s unpaid work would be helpful in addressing these shortcomings. One must bear in mind, however, that data compilation is expensive, in terms of both time and resources; gathering data for its own sake is not cost-effective.
To conclude, current data and resource constraints leave much room for fruitful research. Many of the empowerment studies are based on case studies. Although these studies have been beneficial in suggesting correlations, they are restricted by time and place. Recent work in development economics using randomized control trials may prove important for engendering macroeconomics. Although such trials are not without their critics, their methodology appears tailor-made to address many of the important gender-based questions that remain unanswered.
