65931 Opportunities for Men and Women: Emerging Europe and Central Asia Sarosh Sattar THE WORLD BANK WASHINGTON, D.C. © 2012 International Bank for Reconstruction and Development / International Development Association or The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpre- tations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or accep- tance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Contents Acknowledgments ix Preface xi Executive Summary xiii Links to World Development Report 2012: Gender Equality and Development xiv Main Messages xv Policy Priorities xviii Concluding Remarks xix Introduction 1 Chapter 1. Gender Issues in Human Capital 3 Introduction 3 Gender Differences in Education 4 Demography 11 Health 14 The Implications for Policy Design 26 Chapter 2. Women in the Labor Market 31 Introduction 31 Description of Labor Markets 32 The Gender Wage Gap 52 The Implications for Policy Design 56 Chapter 3. Women in Entrepreneurship 63 Introduction 63 The Involvement of Women in Entrepreneurship 64 Women Entrepreneurs and the Characteristics of Their Firms 69 iii Constraints on Women’s Entrepreneurship 71 Do Woman-Owned Firms Perform Comparably with Man-Owned Firms? 74 The Implications for Policy Design 76 Chapter 4. Summary of Findings 81 Human Capital Endowments 81 Labor Markets 83 Entrepreneurship 86 Policy Implications and Critical Knowledge Gaps 88 Annexes 91 Annex A: Data 92 Annex B: Selected Government-Led Initiatives in Europe and Central Asia 97 Annex C: Selected Donor Initiatives 98 References 99 Figures Figure EX1 The Gender Gap in Human Capital xv Figure EX2. Employment Opportunities xvii Figure 1.1. Primary Education: A Regional Comparison 5 Figure 1.2 Primary Enrollment Rates and Working Children by Gender, Selected Countries 6 Figure 1.3 Secondary School Enrollment and Academic Performance, 2009 7 Figure 1.4 School Enrollment Rates among 15- to 17-Year-Olds by Income Quintile, 2009 9 Figure 1.5 Tertiary Education by Gender, 2009 10 Figure 1.6 Population Pyramids, Selected Countries and the Region, 2000, 2025, 2050 13 Figure 1.7 Women’s Advantage in Life Expectancy, Selected Countries and Regions, 2005–10 15 Figure 1.8 Changes in the Gender Gap in Life Expectancy, 1990–2010 16 Figure 1.9 Trends in Life Expectancy, Selected Countries and the Region, 1990–2015 16 Figure 1.10 The Evolution in Mortality Rates among the 20–59 Age-Group 17 Figure 1.11 Maternal Mortality Rates, 1990–2008 18 Figure 1.12 Maternal Mortality Rates by Country, 2000 and 2008 19 Figure 1.13 Infant and Under-5 Mortality Rates, 1990–2009 20 Figure 1.14 Under-5 Mortality Rates, 2000 and 2009 20 Figure 1.15 Infant Mortality Rates 21 Figure 1.16 The Evolution of the Total Fertility Rate, 1990–2009 21 Figure 1.17 The Variation in Total Fertility Rates across Countries, 2009 22 Figure 1.18 Changes in the Total Fertility Rate across Countries, 1990–2009 23 Figure 1.19 Abortion Rates across Countries, 1999 and 2008 24 Figure 1.20 Adolescent Fertility Rate, Selected Regions, 1998–2009 25 Figure 1.21 Adolescent Fertility Rates across Countries, 2009 25 Figure 1.22 Sex Ratio at Birth, 2008 26 iv Opportunities for Men and Women: Emerging Europe and Central Asia Figure 2.1 Correlation among Per Capita GDP Growth, Labor Force Participation, and the Wage Gap 32 Figure 2.2 Labor Force Participation Rates 33 Figure 2.3 Labor Force Participation Rates by Subregion 35 Figure 2.4 The Gender Gap in Labor Force Participation by Age, 2009 36 Figure 2.5 Labor Force Participation among Men and Women, 1980–2009 37 Figure 2.6 Female Employment Rate by Number of Children and the Age of the Youngest Child, 2008 38 Figure 2.7 Gender Gaps in Unemployment 40 Figure 2.8 Youth Unemployment and Long-Term Unemployment, by Gender, 2008 40 Figure 2.9 Self-Employment 41 Figure 2.10 Sectoral Employment, by Gender 43 Figure 2.11 The Gender Gap in Management 46 Figure 2.12 Part-Time Workers, by Gender 47 Figure 2.13 The Difference in Time Use between Men and Women 47 Figure 2.14 The Distribution of Household Chores among Men and Women 48 Figure 2.15 The Gender Wage Gap 53 Figure 2.16 Reasons Men and Women Work Less Than 30 Hours a Week, European Union, 2007 59 Figure 2.17 Childcare Enrollment Rates among Under-3-Year-Olds 60 Figure 2.18 Share of 0- to 3-Year-Olds Enrolled in Childcare, 2008 61 Figure 3.1 Women and Men Employers 64 Figure 3.2 Women’s Participation in Firm Ownership 67 Figure 3.3 Women’s Participation in Top Management 68 Figure 3.4 Women’s Ownership of Firms by Firm Characteristics, 2008 69 Figure 3.5 Sole Proprietorships and Firm Characteristics, 2008 70 Figure 4.1 Secondary School Gross Enrollment Rates: The Region’s Gender Advantage is Disappearing, 2009 82 Figure 4.2 Gender Gaps Emerge Significantly at the Tertiary Level 83 Figure 4.3 Women’s and Men’s Labor Market Outcomes 84 Figure 4.4 Working Women’s Constraints 85 Figure 4.5 The Wage Gap 86 Figure 4.6 Employers and Owners 87 Figure 4.7 Firm Characteristics of Woman-Owned and Woman-Managed Firms 88 Figure A1. Percentage of the Population above 60, 2009 and 2050 92 Figure A2. Percentage of the Population above 80, 2009 and 2050 92 Figure A3. Relation between the Total Fertility Rate and the Mean Age at Birth 95 Tables Table 1.1 Demographic Profiles and Fertility 11 Table 2.1 The Sectoral Structure of Employment 44 Table 2.2 Maternity Leave Legislation 57 Table A.1 Life Expectancy, by Country 93 Table A.2 Age-Specific Fertility Rates 94 Table A.3 Age at First Birth 96 Contents v Boxes Box 1.1: Russia’s Growing Health Crisis 17 Box 1.2: Health Access among Roma Women 27 Box 2.1: Most Teachers are Women 45 Box 2.2: The Informal Labor Market in Europe and Central Asia 50 Box 2.3: Migration and Remittances in CIS Countries 51 Box 2.4: Family Policies in Romania 59 Box 3.1: Entrepreneurship: Opportunity or Necessity? 66 Box 3.2: Encouraging Property Ownership through Stamp Duty Reduction 72 Box 3.3: Equity Funds 75 Box 3.4: Capital Seed Program in Chile 76 Box 3.5: Training Women in Nontraditional Sectors 76 Box 3.6: Networking for Success: Examples of Women’s Business Associations 77 Box 3.7: Finance for the Missing Middle 78 Box 3.8: A Small and Medium Enterprise Finance Facility for Women 78 Box 3.9: A Government-Led Initiative in India to Track Gender-Disaggregated Data on Access to Finance 79 vi Opportunities for Men and Women: Emerging Europe and Central Asia Abbreviations CIS Commonwealth of Independent States EU European Union GDP gross domestic product MFI microfinance institution OECD Organisation for Economic Co-operation and Development PISA Programme for International Student Assessment (OECD) SME small and medium enterprise Vice President: Philippe H. Le Houerou Sector Director: Yvonne M. Tsikata Sector Manager: Benu Bidani Task Leader: Sarosh Sattar vii viii Opportunities for Men and Women: Emerging Europe and Central Asia Acknowledgments T his report has been the work of the World Bank’s Europe and Central Asia Region. The task was led by Sarosh Sattar. The main contributors to the various chap- ters are as follows: Josefina Posadas and Sara Johansson (human capital); Josefina Posadas and Luca Flabbi (labor markets); and Sushma Narain and I. Elaine Allen (en- trepreneurship). The chapter on entrepreneurship could not have been written without the analysis carried out by Shwetlena Sabarwal and Katherine Terrell, who prepared, as background, the World Bank Policy Research Working Paper, “Does Gender Matter for Firm Performance? Evidence from Eastern Europe and Central Asia.� The team would like to acknowledge the cooperation of the following persons: Thokozani Kadzamira, Gregory Kisunko, Huailu Li, Claudio Montenegro, Saumik Paul, Shaha Ali Riza, Lars M. Sondergaard, and Ricky Ubee. The peer reviewers are Pierella Paci and María Beatriz Orlando. Administrative support has been provided by Helena Makarenko. The report has been prepared under the guidance of Benu Bidani (Sector Manager), Yvonne Tsikata (Sector Director), and Indermit Gill (Chief Economist). ix Preface T his report has been motivated by the desire to understand how men and women have benefited from the long economic boom that began in the region in 2000. During the subsequent decade, the countries of the Europe and Central Asia Region achieved robust economic growth and macroeconomic stability, and many at- tracted foreign capital and investors. This favorable situation at the national level raised the obvious question of who was benefiting from growth and how widely shared were the wealth and the new opportunities created. During early consultations for this report both inside and outside the World Bank, there was a strong interest in understanding women’s and men’s gains over the last several years, especially in the labor market and in the private sector. However, while the analysis was under way, the global financial crisis hit the Eu- rope and Central Asia Region particularly severely, leading to sharp falls in gross domes- tic product and rising unemployment. It also delayed the work on the report while the World Bank was responding to the immediate needs faced by the countries in the region. Though the worst of the crisis has passed in some countries, and many countries have seen a resumption of growth, the data on the crisis years are still not available so that we might tell the story of how men and women fared as economic actors during the crisis. At the early stages of the drafting process of the report, there was a growing aware- ness in the World Bank of the importance of advancing women’s economic opportuni- ties, in addition to improving women’s health and education. In 2007, the Gender Ac- tion Plan was launched to improve women’s access to jobs, land rights, financial services, agricultural inputs, and infrastructure. This expansion in the issues that might benefit from a deeper understanding of the opportunities and challenges faced by women and men separately led over the next four years to an increase in gender-disaggregated analy- sis and projects. Building on the analysis and the experience gained through the Gender Action Plan, World Development Report 2012 focuses on gender equality and development (World Bank 2011a). It looks at the facts and trends regarding the various dimensions of xi gender equality in the context of the development and enjoy freedom of choice. This shortcoming is process. The report provides a useful framework caused by the quantitative focus of this report and based on three pillars—endowments, access to eco- the lack of regionwide data. However, agency is an nomic opportunities, and agency—to understand important topic, and encouraging work in this area is the sources of differences in outcomes among men essential to obtaining an in-depth understanding of and women. Within this framework, the report the reasons behind the outcomes observed. (Chap- covers a broad range of issues, from health, educa- ter 4 summarizes the main findings of our report.) tion, and fertility to land ownership, labor markets, This report is a small contribution to our and access to new technologies. understanding of gender issues in the Europe and Aspects of the framework of the World Devel- Central Asia Region. It does not pretend to cover all opment Report guide the analysis in our report. Chap- important topics even in the areas of human capi- ter 1, on human capital, looks at a narrower range of tal, labor markets, and entrepreneurship. Rather, it endowments rather than the broader range of assets aspires to set the groundwork for more in-depth such as land. Chapters 2 and 3, on the labor mar- analysis at the country level of the differing out- ket and on entrepreneurship, respectively, provide comes among men and women and to encourage a information on access to economic opportunities. dialogue about the related medium- and long-term However, our report does not cover agency, that is, implications for sustained economic development women’s capacity to make independent decisions and welfare improvements. xii Opportunities for Men and Women: Emerging Europe and Central Asia Executive Summary T he countries of Central and Eastern Europe and Central Asia have a long history of striving for gender equality, especially in the public sphere. Not only was this an important goal during the socialist era, but governments continued to pursue gender equality even during the difficult years of transition. The governments in the region allocated substantial resources toward the health and education of both women and men. They also adopted legislation that treated women and men equally in the la- bor market and they provided child care services. During much of the last century, the region surpassed countries—both developing and developed—in establishing the equal treatment of women and men. Much of the first decade of the 21st century was defined by robust growth. The new member states of the European Union saw major structural changes, while the Commonwealth of Independent States bounced back from the financial crisis in the Russian Federation at the end of the previous decade. Thus, for several years, the econ- omies of the region experienced robust growth. However, this period of prosperity was followed by a global financial crisis, which hit the region severely, especially in 2009 and 2010. The question remains whether men and women benefited equally over the course of the last decade, with its major economic developments, or whether the gaps between men and women and girls and boys changed significantly. This report reviews changes in gender inequalities in Europe and Central Asia over the last decade, with a particular focus on economic opportunities both in labor mar- kets and in entrepreneurial activity. In addition, given the importance of health and education in opening up opportunities for men and women in the economic sphere, the report discusses the changes in human capital endowments in this area as well. Un- fortunately, the lack of data prevents us from covering the role of other assets, such as property or financial assets, in broadening opportunities for women. There is a significant body of work on gender equality to build upon. Most re- cently, the World Bank published World Development Report 2012: Gender Equality and Development, which examines the progress in women’s lives and provides policy recommendations to advance gender equality (World Bank 2011a). The report also xiii provides a useful framework for understanding Gender disparities still remain in many areas, the underlying causes as well as consequences. and even in rich countries. The most persistent Gender in Transition, an earlier World Bank report and egregious gaps include excess deaths of specific to the region, also reviewed the gender- girls and women…, disparities in girls’ school- based differences in economic opportunities, es- ing…, unequal access to economic opportuni- pecially in the labor market and human capital, ties…, [and] differences in voice in households over the first decade of transition in the 1990s and society. (World Bank 2011a, xxi) (Paci 2002). The remainder of this executive summary is Income growth by itself does not deliver structured as follows. The next section, A, reviews greater gender equality on all fronts… Gen- the main aspects of the World Development Re- der gaps persist where girls and women face port on gender and development. The following other disadvantages… Markets, institutions, sections, B and C, discuss the main messages of this and households can also combine to limit report and policy priorities, respectively. Section D progress… globalization can help…, but [the] makes some closing remarks. impact [of the related forces] will be muted without effective domestic public action. (World Bank 2011a, xxi) Links to World Development Report 2012: Gender Equality and Development As the report suggests, there is an important role for policies targeted toward reducing the most World Development Report 2012: Gender Equality costly gender disparities that are not responsive to and Development examines various dimensions of growth, and closing the most egregious gender gaps gender equality in the context of the development has become more urgent now than it was two or process. In particular, gender equality is viewed three decades ago. through an economic lens, with a focus on the re- An overlap exists between World Develop- lationship among (1) the accumulation of endow- ment Report 2012 and this report. The methodol- ments (such as human capital and financial assets), ogy of examining endowments and the access to (2) access to economic opportunities and their re- economic opportunities and their returns is ap- turns, and (3) agency, or the ability to take actions plied in the analysis of gender issues in Europe and and make choices that impact the individual, house- Central Asia. Our report discusses issues of hu- hold, or public spheres. These factors interact and man capital, such as education and health, which affect one another significantly. is the most important asset of the majority of the The report’s main messages are as follows: population. In our report, the issues of economic opportunities and their returns are discussed in Gender equality is a core development objec- the context of labor markets. The entrepreneur- tive in its own right. . . . Greater gender equal- ship chapter (chapter 3) touches on women’s ity can enhance productivity, improve devel- business assets and on economic opportunities. opment outcomes for the next generation, However, our report does not address agency, a and make institutions more representative. topic extensively addressed in World Development (World Bank 2011a, xx) Report 2012. The concept of agency—the ability to exer- Development has closed some gender gaps. cise choice over consequential decisions—is im- The disadvantages faced by women and girls portant in furthering our understanding and in- that have shrunk most rapidly over the past corporating into analysis, particularly where there quarter century include educational enroll- are large discrepancies between men and women ment…, life expectancy…, and labor force in the private (household) and public spheres. In participation. (World Bank 2011a, xx) Europe and Central Asia, women and men appear xiv Opportunities for Men and Women: Emerging Europe and Central Asia Figure eX. 1 The Gender Gap in Human Capital a. Gender gap, secondary gross enrollment rate b. Gap in life expectancy, women and men percentage points number of years, 2009 15 10 9 10 8 7 5 6 5 0 4 3 –5 2 1 –10 0 ECA LAC MNA SSA EAP SAS ECA LAC MNA SSA EAP SAS 1999 2009 Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: The gender gap measures the difference in the gross secondary enrollment rate or life expectancy of males and females. eAP = east Asia and the Pacific. eCA = europe and Central Asia. LAC = Latin America and the Caribbean. MNA = Middle east and North Africa. SAS = South Asia. SSA = Sub-Saharan Africa. All regional data are for developing countries only. to have equal agency because of factors such as Main Messages the gender-blind legal system, women’s compara- ble labor force participation rates, and girls’ access This section consolidates the analysis in our main to education. However, if equal agency is to exist, report to identify the overarching messages. This men and women may need different types of in- approach allows policy makers and civil society stitutional support given the differences between alike to see the significance of individual gender them. gaps within the larger context of economic devel- Agency in the context of Europe and Cen- opment and prosperity. Presented below are the tral Asia has changed over the last two decades report’s three take-away messages. in concert with the changes brought about by the First, the region’s advantage in gender equal- contraction of the role of the state and, more gen- ity has eroded, with the result that the region now erally, the transition to a market economy and the looks more similar to the rest of the world. The development of large informal markets. Across gender gaps in school gross enrollment rates have the board in the region, there has been a reduc- fallen further in Europe and Central Asia over the tion in fertility-related benefits, especially the past decade, but so have the world average gaps, contraction in childcare services. The scarcity of given the substantial progress achieved in improv- childcare availability has narrowed women’s op- ing educational outputs (see figure 1).1 Globally, portunities, especially in the labor market. Agen- female enrollment rates have risen more rapidly cy has also been circumscribed by low incomes than male enrollment rates at both the primary and on the one hand and, on the other hand, by a lack secondary school levels so that global gender gaps of the public services (for example, childcare) in 2009 had fallen to 4 and 2 percentage points, re- that allow women to function independently. spectively, compared to Europe and Central Asia’s However, women have much greater agency in the region because of human capital endow- ments, independent sources of income, and the 1 Measured using the difference between male and legal framework. female gross enrollment rates. Executive Summary xv 1 and 3 percentage points. In terms of health out- menia, Azerbaijan, and Georgia, the corresponding comes, the region exhibits health indicators that number is 16, 15, and 11, respectively. The miss- are better than the world averages, especially the ing girls at birth reflect covert discrimination in maternal mortality rates, which, in the region, are the household, resulting from the combination of the lowest among all developing countries. How- strong preferences for sons, declining fertility, and ever, in terms of female life expectancy, the region the spread of prenatal sex determination. The sex ranks below Latin America and the Caribbean, but imbalance is reversed among the population group is comparable to East Asia and the Pacific and the ages 50 years and above by the early demise of men, Middle East and North Africa. Male life expectan- especially in Kazakhstan, Russia, and Ukraine be- cy in the region is unusually low and comparable to cause of poor health, alcoholism, and accidents. the average of Sub-Saharan Africa. Second, the structural changes in the econo- The region’s gender gap in education is similar mies of the region have opened up economic and to the gaps in other predominantly middle income re- employment opportunities for women and reduced gions. One of the key areas where the gender pro- some avenues of prosperity for men. Women play an file of the region is similar to that of the rest of the important role in these economies; they comprise 46 world is in the area of education, especially in pri- percent of the labor force in the region, which is above mary education and, to a lesser extent, in secondary the world average of 40 percent. The labor force education. Net primary enrollment rates in the re- participation rates of women and men in Europe gion are comparable to the world average, with no and Central Asia average 51 and 69 percent, respec- significant gender gap. However, even in the area of tively (figure 2). While female participation rates are secondary education, selected regions such as Latin comparable to the world average, male participation America are quickly converging toward Europe rates are well below the global average of 78 percent. and Central Asia’s high net female and male en- This may be a reflection of many factors, including rollment rates and modest gender gap. Among the the types of jobs being created. Over the last decade, countries in the region that have made significant the region has experienced an expansion in the ser- progress over the last decade, Turkey stands out be- vice sector, while the manufacturing and agriculture cause of an increase of 20 percentage points in fe- sectors have shrunk significantly. The contraction of male secondary enrollment rates. However, Turkey the manufacturing sector in particular has had a dis- continues to show a large gender gap at the second- proportionate and adverse impact on men, while the ary level, as do some minority communities (such growth in the service sector has opened up relatively as the Roma) in Central and Eastern Europe. At the more opportunities for women, who constitute half tertiary level, female enrollment rates exceed male of all workers in the service sector in the region. enrollment rates in most parts of the region, and However, women’s gains in the service sector may the level of inequality appears to increase with de- be only short term. Service sector jobs are diverse velopment. Next to the high-income countries of in terms of occupations and productivity. Service the Organisation for Economic Co-operation and jobs may be found in both the public and private Development, the region has the largest gender gap sectors. In the region, women appear to be dispro- in tertiary education, at 11 percentage points. portionately represented in the public sector, as Health indicators in the region are superior to indicated by several country studies. For example, those in other middle- and low-income countries, ex- if we look at the education sector (which is mostly cept for a sex imbalance among young children that in the public sector), women constitute 91 and 72 has begun to emerge and unusually high mortality percent, respectively, of all primary and secondary rates among prime age men in selected countries. Ar- teachers. However, the fiscal pressures on the large menia, Azerbaijan, and Georgia show the highest public sector are growing because of the need for sex imbalance at birth in the world after China. adjustment following the economic crisis, which hit Thus, in China, 18 more baby boys than baby girls the region severely. Moreover, the role of women in are born per 100 live female births, while, in Ar- the private sector as employees, but also as employ- xvi Opportunities for Men and Women: Emerging Europe and Central Asia Figure eX. 2 Employment Opportunities a. Labor force participation rates, 2009 b. Raw gender wage gap percentage male wages in excess of female wages, % 100 50 45 45 80 40 39 36 35 60 30 25 40 20 18 15 10 20 10 5 0 0 ECA LAC MNA SSA EAP SAS EU10 Central Asia Russian S. Caususes W.Balkans Federation Female Male Sources: For labor force participation rates: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/ data-catalog/world-development-indicators/. For the wage gap: various; see chapter 2. Note: eAP = east Asia and the Pacific. eCA = europe and Central Asia. eu10 = the 10 european union (eu) countries of Central and eastern europe, that is, Bulgaria, the Czech republic, estonia, Hungary, Latvia, Lithuania, Poland, romania, the Slovak republic, and Slovenia. LAC = Latin America and the Caribbean. MNA = Middle east and North Africa. SAS = South Asia. russian Federation, S. Caucasus = southern Cau- casus. SSA = Sub-Saharan Africa. W. Balkans = western Balkans. ers lags significantly with respect to the role of men. Third, the dramatic demographic changes For example, women are much less likely to engage in the region have different implications for men in entrepreneurship, which has long-term implica- and women, which must be taken into account if tions for women’s wealth accumulation. economic growth is to be sustained in the medium The wage gap between men and women remains to long term. Though the total population of the large in the region, and, if human capital endow- region is not expected to change during the first ments are taken into account, the gap is even wider. quarter of this century (a projected −0.1 percent- The wage gap in the region is comparable with the age point decrease), there will be a dramatic change gap observed in other low- and middle-income in the age structure of the population in coming countries. The data indicate that women’s hourly years. The share of the population above 60 years wages are 22 percent less than the hourly wages of of age will rise sharply in 2009–25, from 15 to 25 men, on average, in the region. Moreover, because percent of the population, and women will con- women work fewer hours, the monthly wage gap is stitute 57 percent of this age group. Moreover, 17 larger, indicating that women earn 29 percent less countries in the region will suffer from population than men per month. Women earn systematically implosion, which will result in a shrinking labor less than men even if one controls for differences in force. These countries can ill afford this outcome human capital. The size of the gender wage gap var- given their levels of income and investment. ies greatly within the region. The differences across There will be more elderly people, especially wom- countries are large within the region. The western en, and these people will be vulnerable to old age pov- Balkans (with the exception of Albania), the EU10 erty. A variety of factors increase the probability that countries, Moldova, and Turkey have a gender gap of about 20 percent.2 Other countries show gender 2 The EU10 refers to the 10 European Union (EU) gaps that are among the largest in the world (for ex- countries of Central and Eastern Europe, that is, Bul- ample, Tajikistan, with a gap of about 65 percent), garia, the Czech Republic, Estonia, Hungary, Latvia, while still others (Albania and Russia) have gaps in Lithuania, Poland, Romania, the Slovak Republic, and the intermediate range of about 35 percent. Slovenia. Executive Summary xvii women are more likely than men to fall into poverty. removing the impediments women face in contrib- Women without independent sources of income in uting to the economy. The gaps to fill in the labor old age must rely on the incomes of family members. market in the economies of the region are large and However, even women who work in the formal sec- will become larger in the future. The countries might tor and possess pension rights tend to be more vul- also benefit from the more effective use of women nerable to poverty. Pension rules in many systems because of their higher levels of human capital. in the region allow women to retire early, but this The first policy recommendation: take measures decision has consequences for the level of benefits. to facilitate women’s entry into the labor force so as to Moreover, women live longer, on average, than men, meet the challenges of demographic changes, especially meaning that their low benefits must stretch farther, the aging of the population. Women’s labor force par- likely resulting in a divergence between average pen- ticipation is important for several reasons, includ- sions and average wages over time. Finally, the infor- ing to help ward off old age poverty and to assist mal social safety net may be weakened because older the region’s economies adjust to the contraction women will have fewer children to rely upon. in the labor force. However, at the same time, it is Many countries in the region will have to find a also important that labor market participation not way to encourage more women to participate without impede women in having children. These multiple decreasing women’s fertility. Many of the countries objectives might be accomplished by (1) increasing in Europe and Central Asia with aging or declin- the retirement age of women so as to achieve par- ing populations could experience an economic ity with the retirement age of men; (2) reevaluating contraction because of the shrinking labor force. maternity-related benefits in terms of their effec- To prevent a sharp decline in economic activity, tiveness in increasing fertility, for example, mater- one may adopt the option of trying to maintain the nity leave, child allowances, lump-sum payments size of the labor force by attracting temporarily or for the birth of a child; and (3) providing childcare permanently inactive women or by retaining work- given that its absence may account for the choices ing women in the labor force for a longer period women make in their fields of specialization in ter- of time by delaying retirement (so that the retire- tiary education and in their occupation because of ment age of women is equal to the retirement age compatibility with raising children. of men). Yet, to achieve higher female labor force The second policy recommendation: adopt educa- participation without adversely impacting fertil- tional reforms to reduce the gender imbalances at the ity, the supporting institutional infrastructure (for secondary and tertiary levels. The secondary school example, childcare and appropriate parental leave gender gap (in favor of males) appears, at best, to benefits) needs to be in place, and pension rules be closing only slowly. Country-specific policies are should be amended. needed given that, depending on the country, the cause of the gap may vary significantly across the region, from income to minority status or ethnic- Policy Priorities ity. For example, Roma children (especially girls) are less likely to attend secondary school. At the Policy recommendations that address gender equal- tertiary school level, significantly more women ity in the region could have two main objectives. The than men pursue higher education: the average first goal may be to achieve equal opportunity among female and male gross enrollment rates are 60 and men and women by taking into account their differ- 45 percent, respectively. The lower probability that ent circumstances. Women differ from men in terms men will participate in tertiary schooling may be of their roles in the private sphere, their greater linked to the lower perceived returns to education vulnerability to physical insecurity because of their among men (either in absolute terms or relative to longer life spans, and their fertility, among other fac- what men might earn through migration). Also, tors. The second goal may involve helping societies given the real concerns of the relevance of educa- meet their medium- to long-term economic needs by tion discussed in Skills, Not Just Diplomas: Manag- xviii Opportunities for Men and Women: Emerging Europe and Central Asia ing for Results in Education Systems in Eastern Europe Concluding Remarks and Central Asia (Sondergaard and Murthi 2012), young men may decide to forgo education so as to This report analyzes various markets through a gen- acquire additional work experience and earnings. der lens. It thereby quickly makes two main find- The third policy recommendation: address pock- ings. First, our knowledge of why men and women ets of health disparities on a country-by-country ba- behave differently is limited. For example, why do sis. The region is not free of health disparities, al- women pursue certain fields in education that men though no one health challenge is regionwide. To tend to avoid or why do more women than men address health problems, a comprehensive national migrate from some countries. Second, average eco- program is needed that operates at all levels of gov- nomic indicators can be misleading because they ernment. The countries will need to ensure that an hide the differences in behavior across large groups appropriate strategic, legislative, and policy frame- (in our case, males and females). The importance of work exists, along with adequate institutional ca- the adoption of a gender lens in the development of pacity. In addition, the identification and delivery policies in the region derives in no small part from of specific programs would be required and, to be the high stakes resulting from the dramatic demo- effective, may call for a multifaceted approach that graphic changes and mounting labor resource needs. includes components such as information cam- Moreover, such a lens helps countries identify the paigns to promote the more effective enforcement necessary and appropriate policy and institutional of existing laws (such as road safety campaigns) to framework to allow them to take advantage of un- the more effective provision of health services (for derexploited opportunities (for example, bringing example, emergency care). This is true of the effort more educated women of prime age into the labor to reduce maternal mortality rates, increase male force). Ultimately, the use of disaggregated data to life expectancy, and address the imbalance in the undertake analysis will likely lead to better poli- sex ratio. cies that further the long-term objective of shared growth. Executive Summary xix Introduction E urope and Central Asia have suffered a setback in economic growth because of the recent global crisis, which revealed fundamental structural weaknesses pre- viously hidden by the prosperity before the crisis. The major weaknesses are the large savings deficits, the lagging reforms in the social sectors, and the deterioration in competitiveness. Policies can address these weaknesses by taking into account the role of the behavior of firms, public spending on health and education, the consequences of demographic pressures, particularly on pension systems, and the bottlenecks created by skilled labor force shortages.3 To overcome the current difficulties, the countries of Europe and Central Asia will need to deepen policy changes across the social and economic spheres. Social sector reforms that reach all groups within society can promote inclusive growth by increasing employment and access to high-quality public services. Furthermore, many reforms are needed to reverse the deterioration in competitiveness. Some of the most important ac- tions will be those that enhance productivity and skill development in the region, which already possesses a highly educated population. Any reform agenda ultimately affects a country’s population. An agenda that seeks to promote inclusive growth will benefit from the identification of the different groups within society, an understanding of the related constraints, and consideration of wel- fare issues. Adding a gender perspective to policy making can be particularly pertinent because men and women may behave differently in the economic and social spheres. These differences mean that reforms can have dissimilar and unintended impacts on men and women. It may also mean that, by adopting gender-informed policies, coun- tries are more likely to achieve desired outcomes. This report reviews the performance of women and men during the past decade in three spheres: human capital, labor markets, and entrepreneurship. Similar to World Development Report 2012 (World Bank 2011a), it looks at the dimensions of endowments and economic opportunities. However, it does not cover issues related to agency, that is, 3 The diagnosis and the solutions are proposed in World Bank (2011b). 1 the ability of actors to make independent choices a quantitative analysis can add value by providing and exercise control over their own actions. The some measure of the degree of differences in the data are analyzed to determine whether women and outcomes observed. Yet, it is a weakness as well be- men are performing well compared with each oth- cause the outcomes are measured, but not always er, but also how they fare in a global context. The explained. Consequently, more work is needed in analysis is primarily quantitative and mines various this area, especially qualitative analysis that is fol- data sets. This is a strength of the report because lowed up by targeted quantitative surveys. 2 Opportunities for Men and Women: Emerging Europe and Central Asia Gender Issues in 1 Human Capital Introduction A nation’s stock of human capital influences economic growth, productivity, and, ulti- mately, poverty reduction. Investing in the building blocks of human capital—educa- tion and health—has both direct and indirect effects on economic growth, but also on poverty reduction. The socialist legacy of investment in education and health for both men and women in Europe and Central Asia has provided an important foundation for the majority of the countries as they have embarked on a significant transformation and liberalization of their economies and on global integration. This stock of human capital is all the more valuable now because of the demographic transition occurring in the region and the decline in population in 21 of the 30 countries. The shrinking popula- tion underscores the importance of using available human resources more effectively in terms of the share of the population that is economically active, as well as the productiv- ity of the population. Recent analytical work has laid the foundation of knowledge on the broad edu- cation and demographic issues facing the region. The report “From Red to Gray: The Third Transition of Aging Populations in Eastern Europe and the Former Soviet Union� (Chawla, Betcherman, and Banerji 2007) analyzes the implications of the aging popula- tions and relatively weak institutions in low- and middle-income countries. The study “Skills, Not Just Diplomas: Managing for Results in Education Systems in Eastern Eu- rope and Central Asia� (Sondergaard and Murthi 2012) discusses the lagging perfor- mance of the education sector, primarily in terms of the quality and flexibility of cur- ricula to facilitate lifelong learning, an important characteristic of education systems in rapidly changing economic environments. The analysis in this chapter adds a gender 3 emphasis to some of the conclusions reached in Gender Differences in Education these reports. The chapter contains three main messages, as Education plays an important role in raising the follows: quality of human capital and helping people be- come more productive in their personal and profes- ◆ The countries of Europe and Central Asia sional lives. The countries of Europe and Central have achieved gender parity in primary edu- Asia invest heavily in the education of their citizens cation and, to a lesser degree, in secondary at all levels of schooling. The governments and the education. However, there are gender gaps in citizenry in these countries have an established tra- tertiary education that are large and show the dition of educating both girls and boys. In the past, potential to widen. this set the region apart from the rest of the world ◆ The demographic transitions in the region through narrow gender gaps in literacy rates, but are dramatic. The aging of the population will the situation is now changing, and the gender gaps lead to large numbers of elderly women. In in education are narrowing in all regions.4 countries with growing populations, there This section provides a description of the will be proportionately more women of successes and challenges in gender equality in the working age, which should provide an oppor- educational systems in the region. There have been tunity to these countries to expand the pool significant and broadly based successes, but subtle of educated labor. challenges are also emerging. In the discussion, we ◆ Several health indicators in the countries of review the status of male and female outcomes at the region are better than the correspond- the primary, secondary, and tertiary levels. In ad- ing indicators in many developing countries dition, several of the countries in the region par- at similar levels of income, especially the in- ticipate in an international assessment of students dicator on maternal mortality. Yet, other in- that allows us to examine the presence of a gender dicators show a lag, such as male life expec- gap in capabilities. Finally, we focus especially on tancy, which is exceptionally low in selected developments in tertiary education, including en- countries considering the income levels. trenched and emerging trends. Moreover, a gender imbalance has begun to emerge in some countries among young chil- dren, and this is cause for concern. Primary Education The region continues to perform well in providing The rest of the chapter is structured as fol- primary education to both girls and boys. Though lows. The next section provides an overview of there are concerns about selected countries (for primary, secondary, and tertiary outcomes in edu- example, Azerbaijan and the Kyrgyz Republic) in cation. The third section discusses demographic which almost a fifth of the relevant population is transitions. The fourth section reviews key health not enrolled in primary school and other countries indicators, such as life expectancy, mortality rates, in which completion rates are not universal (for ex- fertility, and missing women. The health and edu- ample, Albania), these problems do not appear to cation indicators cover the current situation, but reflect a gender bias. Despite the importance placed also trends over time so as to capture changes in on education in the region and the significant eco- the gender gap. Where possible, a comparison is nomic growth experienced for much of the last made between Europe and Central Asia and other regions of the world, including the high-income countries of the Organisation for Economic Co- 4 Literacy rates among men and women in Europe and operation and Development (OECD). The final Central Asia are 97 and 99 percent, respectively, com- section offers an analysis of the implications for a pared with the corresponding world averages of 79 policy response. and 83 percent. 4 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.1 Primary Education: A Regional Comparison a. Primary net enrollment rates, 2009 b. Enrollment rates by income quintile percentage %, by gender 100 100 90 90 80 80 70 70 60 60 50 Bottom Second Third Fourth Top 40 Population quintile by consumption per capita ECA OECD LAC MNA SAS SSA WLD ECA Bulgaria ECA Bulgaria Female Male Female Female Male Male Sources: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/; europe and Central Asia Data (database), World Bank, Washington, DC, http://data.worldbank.org/region/eCA. Note: eCA = europe and Central Asia. LAC = Latin America and the Caribbean. MNA = Middle east and North Africa. OeCD = high-income OeCD countries. SAS = South Asia. SSA = Sub-Saharan Africa. WLD = World. All regional data are for developing countries only. Data for South Asia is for 2008. The europe and Central Asia region average enrollment rates by income quintile are based upon the latest available data. Data for Bulgaria and Hungary are for 2007, and data for Albania, Azerbaijan, and Lithuania are for 2008. The rest of the countries are Kazakhstan, Latvia, Moldova, FYr Macedonia, Poland, the russian Federation, Tajikistan, and ukraine (all 2009). decade, there is a substantial share of working chil- Even in countries with below average primary dren, especially boys, in some countries (the former enrollment rates, the gender gap is small. In many Yugoslav Republic of Macedonia and Ukraine). The countries, low primary rates are driven by a lack scarcity of data makes it difficult to draw conclu- of female enrollments; however, this is not the sions on the prevalence of this problem or the nega- case in Europe and Central Asia. The countries tive long-term consequences for children who carry with the lowest net primary enrollment rates the dual burden of work and school. among girls in 2009 were Albania, Armenia, Primary enrollment rates among girls and boys Azerbaijan, the Kyrgyz Republic, FYR Macedo- are high in the region and comparable with the world nia, Moldova, and Uzbekistan (83–86 percent average. The region’s net primary enrollment rates range). Primary enrollment rates among boys in are comparable with the world average (see figure these countries fall in the same low range. The 1.1). In the region, the net primary enrollment rates average gender gap in primary enrollment rates among girls and boys are 92 percent; they show no is negligible, and the standard deviation is small. significant gender gap. The region’s traditional ad- The countries with the largest gaps are Tajikistan vantage in this area is eroding because the gap in and Uzbekistan, though the differences are small net primary enrollment rates by gender has been (between 2 and 4 percentage points). At the oth- closing in most parts of the world, including in low- er extreme are Armenia and the Czech Republic, income regions. The world averages of net primary enrollment rates among girls and boys are 87 and 89 percent, respectively, indicating that the next 5 In countries where many children enter school late or generation of women and men in many parts of the repeat grades, gross enrollment rates are high (and can world will be as well educated as their counterparts exceed 100 percent). In Europe and Central Asia, net enrollment rates are high, but gross enrollment rates in Europe and Central Asia. Only the quality and do not indicate whether most children in school are at the relevance of education will set any country or the age-relevant grade levels. Net enrollment rates are region apart from others.5 reported less often. Gender Issues in Human Capital 5 Figure 1.2 Primary Enrollment Rates and Working Children by Gender, Selected Countries a. Primary completion rates, 2009 percentage point difference in male and female rates b. Economically active children 8 40 35 6 30 4 25 20 2 15 0 10 5 –2 0 Turkey Kazakhstan Azerbaijan Kyrgyz Republic Uzbekistan Serbia Macedonia, FYR Tajikistan Bosnia and Belarus Ukraine Albania Georgia Herzegovina –4 –6 Armenia Estonia Macedonia, FYR Turkey Lithuania Tajikistan Georgia Latvia Female Male Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: Panel a: data are for 2009 for all observations except 2007 for Belarus and 2008 for Turkey. Panel b: data are for the most recent year available (2005 or 2006). where the gender gap of 3 percentage points is a whole (figure 1.1). This equitable trend across caused by lower male enrollment rates. genders and income groups occurs in most coun- The average gender gap in primary comple- tries; the exceptions are Albania, Bulgaria, and tion rates in the region is small, though, in a hand- FYR Macedonia. For example, Bulgaria exhibits ful of countries, there does appear to be some both income disparities and gender disparities. disparity between boys and girls. The average pri- Almost 20 percent more 7- to 14-year-old girls mary completion rates for girls and boys were in the highest income quintile attend school rela- 95 and 96 percent, respectively, in the region tive to the lowest quintile. Albania exhibits the in 2009. This gap is comparable with the gap opposite trend, with 13 percent more girls in the in Latin America and the Caribbean. Though lowest quintile attending school relative to the the average difference between the completion top quintile. To understand the causes of these rates among girls and boys is small, Georgia and variations, one must see whether these children Latvia stand out in terms of the gender gap, who were out of school were economically active, which is about 6 percentage points.6 In Lithu- and one must also examine the attitudes of the ania, Tajikistan, and Turkey, boys lead girls by parents of these children toward schooling. about 4 percentage points in primary school Young boys are more likely than young girls completion rates, while, in Armenia, the oppo- to be economically active. The data on child em- site is true (figure 1.2). Though the gap in any ployment are limited, and only a few countries year is small, this inequity will lead to illiteracy report these statistics. The largest gender gaps among a limited, but significant share of the are found in Albania and FYR Macedonia (fig- adult female population. ure 1.2). The countries in the region with the The region shows little variation in school en- highest proportion of economically active chil- rollment rates among young children across income groups. The difference in enrollment rates be- 6 However, this should be seen against the backdrop of tween the bottom and top deciles among boys strong improvements over the last decade among both and girls (ages 7–12 years) is small, at 1 and 2 boys and girls: the rates have grown by 16 and 12 per- percentage points, respectively, for the region as centage points in Georgia, respectively. 6 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.3 Secondary School Enrollment and Academic Performance, 2009 a. Gross enrollment rates by gender b. PISA, Europe and Central Asia percent scores 120 480 460 100 440 80 420 400 60 380 40 360 340 20 320 0 300 ECA OECD EAP LAC WLD Reading Math Science Female Male Sources: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/; PiSA 2009 Profiles by Country/economy (database), Programme for international Student Assessment, OeCD, Paris, http://stats. oecd.org/PiSA2009Profiles/#. Note: eAP = east Asia and the Pacific. eCA = europe and Central Asia. LAC = Latin America and the Caribbean. OeCD = high-income OeCD countries. WLD = World. For a description of PiSA, see the text. Panel b: PiSA scores are for the eu10 (the 10 european union countries of Central and eastern europe, that is, Bulgaria, the Czech republic, estonia, Hungary, Latvia, Lithuania, Poland, romania, the Slovak republic, and Slovenia), Albania, Azerbaijan, Croatia, Kazakhstan, the Kyrgyz republic, Montenegro, russia, Serbia, and Turkey. dren (ages 7–14 years) are Georgia and Ukraine, nities of students and, consequently, for the income where 32 and 17 percent of children are work- and household welfare of these young people. ing, respectively.7 Over 90 percent of children Girls are falling behind in secondary enrollment who are economically active are also studying, in Europe and Central Asia relative to the high-income except in Turkey, where only 60 percent of OECD countries. The decades of the 1990s and these children also attend school. In most coun- 2000s saw an increase in secondary enrollment rates tries for which data are available, boys are more among girls, though this was outpaced by the growth likely to work than girls; the exception is Ta- in enrollment rates among boys. The gender gap was jikistan.8 Yet, this may be an incomplete picture relatively small as measured by the ratio between the since girls may be engaged in household chores respective gross secondary enrollment rates among and other unpaid family work elsewhere as well. girls and boys, which was 0.96, meaning that, for every 100 boys in school, there were 96 girls. This compares unfavorably with the high-income OECD Secondary Education countries, in which gross enrollment rates are 101 Secondary education enhances a population’s social and 102 for girls and boys, respectively (figure 1.3). and economic opportunities. It provides students Though the gender gap in the region is not large with marketable skills and increases their productiv- and the changes over a decade appear small, they ity in the workplace. Economies that are more so- have accumulated over time rather than reversed. A phisticated in terms of the production or the com- broader concern is the fact that school enrollments plexity of institutions require higher levels of human capital input. Secondary education also prepares 7 This assertion is based on 2005 data, which are the lat- students to continue on to college and university. est available data. Thus, a gender gap at the secondary school level has 8 In recent years, this gap has likely widened because of large implications for the long-term career opportu- migration and poverty in rural areas. Gender Issues in Human Capital 7 in general are not rising to converge with the high- average adolescent fertility rates; however, more re- income OECD countries among boys or girls. search needs to be conducted to determine the main Secondary enrollment rates among girls and boys correlates. are high in the countries of the region relative to other Among 15- to 17-year-olds, though poor children low- and middle-income countries, and the gender gap have lower enrollment rates than richer children, the in secondary schools is relatively small, which is similar average gender gap by income quintile is low in the re- to other regions of the world. The gross secondary en- gion. The gender gaps in each income quintile are rollment rates in the region are 87 and 91 percent negligible in the region, indicating that the disad- for girls and boys, respectively. Though these rates vantage of poor children is the same irrespective of are below the gross enrollment rates among girls in whether they are girls or boys (see the case of Alba- Latin America and the Caribbean, this is somewhat nia in figure 1.4). This is, indeed, largely true across misleading. When we compare net secondary school the region, though there are exceptions. In three enrollment rates in Europe and Central Asia and in countries, there is at least a 10 percentage point dif- Latin America and the Caribbean, we find that not ference between the enrollment rates of girls and only does Europe and Central Asia perform com- boys in the first quintile: Bulgaria, FYR Macedonia, paratively better, but that the gender gap narrows and Tajikistan. The gender gap in Bulgaria and Ta- additionally.9 The net enrollment rates among girls jikistan—where poor girls are less likely than boys and boys in Europe and Central Asia are 80 and 82, to attend secondary school—is the opposite of the respectively, compared with 76 and 71 percent, re- gender gap in FYR Macedonia. However, the gen- spectively, in Latin America and the Caribbean. der gap is not at the same magnitude, if it exists at The low-income countries in Central Asia show all, in the top quintile in these same countries. higher secondary enrollment rates among girls relative In the region, girls perform better than boys in to other low-income countries. In the Kyrgyz Republic learning outcomes in reading, but have similar scores and Tajikistan, enrollment rates among girls were in mathematics and science. The OECD Programme 85 and 78 percent, respectively, in 2009. These are for International Student Assessment (PISA) sur- significantly higher than the average female second- veys test 15-year-olds in three subject areas. In the ary gross enrollment rate in low-income countries, 2009 round, 65 countries were surveyed, includ- which was 34 percent in 2009. Furthermore, the ing 18 countries in Europe and Central Asia. These average ratio of male to female enrollment rates in were a diverse group of low- and middle-income low-income countries was 1.36, whereas, for the countries.10 The average score of girls on reading Kyrgyz Republic and Tajikistan, the ratios were 0.99 was 11 percent higher than the corresponding aver- and 1.15, respectively, indicating a greater degree of age for boys, a significant difference. The average gender equality in these countries. scores for boys and girls on mathematics and sci- Tajikistan and Turkey are the two outliers in ence were similar.11 This differential pattern in gen- terms of the gender gap in secondary enrollment rates. der scores is not unique to the region. In almost all the countries in the region, there is a strong correlation between male and female sec- 9 Net enrollment rates are available only for Europe and ondary school enrollment rates, with the exception Central Asia, Latin America and the Caribbean, and of Tajikistan and Turkey. The gender gap in both the Middle East and North Africa. of these countries is between 7 and 10 percentage 10 The countries in the region participating in the 2009 points in favor of boys. In Tajikistan, net secondary PISA were Albania, Azerbaijan, Bulgaria, Croatia, school enrollment rates among girls and boys are the Czech Republic, Estonia, Hungary, Kazakhstan, 77 and 88 percent, respectively, while, in Turkey, the Kyrgyz Republic, Latvia, Lithuania, Montenegro, Poland, Romania, Serbia, the Slovak Republic, Slove- the respective rates are 70 and 77 percent. In Tajiki- nia, and Turkey. stan, the lower enrollment rates among girls may be 11 The ratios for boys to girls in mathematics and science caused by the lack of an adult male presence (aris- were 0.99 and 1.02, respectively. The difference is not ing from out-migration), large families, and above statistically significant. 8 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.4 School Enrollment Rates among 15- to 17-Year-Olds by Income Quintile, 2009 a. By gender b. Gender gaps, bottom and top quintiles percent of 15- to 17-year-olds percentage points 100 30 90 20 80 10 70 0 60 –10 50 –20 40 –30 Bottom Second Third Fourth Top Macedonia, Moldova Latvia Albania Bulgaria Tajikistan Population quintile by consumption per capita FYR Population quintile by consumption per capita ECA Bulgaria ECA Bulgaria Female Female Male Male Bottom quintile Top quintile Source: europe and Central Asia Data (database), World Bank, Washington, DC, http://data.worldbank.org/region/eCA. Note: The gender gap is the difference between the enrollment rates of boys and girls. eCA = europe and Central Asia. The data are the latest available data between 2007 and 2009. Panel a: eCA average includes Albania, Azerbaijan, Bulgaria, Hungary, Kazakhstan, Latvia, Lithuania, Moldova, FYr Macedonia, Poland, russia, Tajikistan, and ukraine. Tertiary Education ception are Tajikistan, Turkey, and Uzbekistan, in Tertiary education extends the period of educa- which the ratio of female to male rates ranges from tion and learning among students. It allows young 41 to 70. On average in the region, tertiary gross en- adults to acquire more specialized knowledge, rollment rates among women and men are 60 and while continuing to develop their cognitive and in- 44 percent, respectively, and the average ratio is tellectual skills, all of which combines to improve 129 women for every 100 men attending colleges their productivity in the work force. Yet, one gener- and universities. This pattern of disproportionately ally does not strive to make tertiary education, un- higher female enrollment in tertiary education is like the primary and secondary levels, universally global, including in the high-income OECD coun- available to the population, nor is the content of tries. However, women are less likely than men to curricula uniform. pursue postgraduate work. Tertiary enrollments are growing at a rapid pace, The richer the country, the more likely women will raising the question of the impact on quality and on be disproportionately more well represented than men in the returns to education, especially among women. tertiary education. In the region, the tertiary enroll- During the decade of 1998–2009, there was a rap- ment rates grow as gross domestic product (GDP) id increase in tertiary enrollment rates in general per capita increases (figure 1.5). However, relative to in the region. Gross tertiary enrollment rates grew, the male enrollment rate, the female enrollment rate on average, by 25 and 15 percentage points among is more strongly correlated with a country’s GDP per women and men, respectively. Though it appears capita. As figure 1.5 reveals, there is significant varia- that women are investing more in education over tion among countries, and some clustering at the up- time (in both absolute and relative terms), it is per end of the distribution. This pattern also occurs not clear whether the resources and the quality of at the global level. In low-income countries, tertiary tertiary education can keep pace with the rapid in- enrollment rates are low, and female enrollment crease in enrollments. rates are much lower than male enrollment rates. In In the vast majority of countries in the region, fe- lower-middle-income countries, the gender gap in male tertiary enrollment rates outstrip male tertiary enrollment rates is negligible, while it grows dramati- enrollment rates. The three countries that are the ex- cally among upper-middle-income countries. Gender Issues in Human Capital 9 Figure 1.5 Tertiary Education by Gender, 2009 a. Enrollment rates by GNI per capita b. Students %, by gender % of students, by gender 120 30 26 100 25 80 20 60 15 15 15 14 40 10 10 7 9 20 5 4 0 3.0 3.5 4.0 4.5 0 Science Social Educ, health, Other Log of GNI per capita, PPP & eng sciences welf Female Male Female Male Sources: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/; uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: gNi = gross national income. PPP = purchasing power parity. eng = engineering. educ = education. Welf = Welfare. Social sciences includes business and law. Other = agriculture, the humanities, services, and unspecified fields. Panel b: the data are for 2009 or the latest available year. The gender gap in tertiary education has been engineering students are women. In Europe and widening steadily over the last decade. The differen- Central Asia, the average is 46 percent; the high- tial in gross enrollment rates by gender in the early est participation rates are in Albania, Estonia, and years of the transition appeared to be relatively nar- Latvia. Almost half the women in the region pur- row, though this cannot be verified because of the sue degrees in the social sciences, business, and law; sparseness of the data. However, since 1998, time health care, education, and welfare-related subjects series statistics on tertiary enrollment rates disag- are the second largest category. gregated by gender have been available for 23 coun- The gender differential in fields of study related tries in the region. Over the period 1998–2008, the to education has an effect on occupational segregation gender ratio (in favor of girls) grew from 113 to and the gender wage gap. A recent study by Flabbi 128 in the region. This was the result of the higher (2011), which was prepared as a background paper pace of tertiary enrollments among women rela- for World Development Report 2012 (World Bank tive to men (7 versus 5 percent per annum). This 2011a), examines the impact of the choice of the feminization of tertiary education is occurring at a field of study on future labor market outcomes in more rapid pace in Europe and Central Asia than 13 European countries, including the Czech Re- in high-income OECD countries, where the ratio public and Estonia. Three of the main conclusions increased from 116 to 129 over this same period. of the study are relevant to Europe and Central In only three countries—Azerbaijan, Tajikistan, Asia. First, women tend to choose education, the and Uzbekistan—did male tertiary enrollments humanities, and health care as their field of study, outpace female tertiary enrollments. while men choose the sciences, mathematics, ag- At the tertiary level, women constitute almost riculture, and veterinary medicine. In the social half the science and engineering students in Europe sciences, business, and law, the gender differen- and Central Asia. This regional average is compa- tial varies by country: for the two countries in the rable with women’s participation in these fields in Europe and Central Asia region that are included the United Kingdom and the United States, where in the study, there is no gender differential. Sec- 50 and 43 percent, respectively, of all science and ond, the gender differential in the field of study 10 Opportunities for Men and Women: Emerging Europe and Central Asia is not explained by (observable) individual char- sectors of health and education, but also in physi- acteristics. The only difference between men and cal infrastructure. women is the effect of ability on the probability of The dramatic demographic changes in the choosing the sciences or mathematics. Persons of majority of countries in Europe and Central Asia high ability (measured by top grades in secondary have brought this issue to the forefront. Many pol- school) are more likely to pursue study in these icy makers are aware of these changes, but much fields, and the effect of ability is larger among men remains to be done at the national level to incor- than among women. Third, the field of study is im- porate the implications of the demographic transi- portant in future labor market outcomes, particu- tion for public expenditures and policies. Medium- larly in the choice of occupation and in the gender and long-term national economic development gap in earnings. plans and strategies should reflect this information, though it is not clear that this is being done con- sistently in any country in the region. This section Demography discusses some of the key changes pertaining to women and men as separate groups. Demographic Though the pace of changes in population profiles changes can have different implications for men may appear relatively slow, the changes have large and women given their frequently different roles in implications for the societies in which they occur. the private and public spheres. Demographic shifts can easily be overlooked be- The demographic transition represents the cause the age and gender profile of a population shift from high fertility rates and high death rates remains relatively stable over the period in which to low fertility rates and low death rates over a pe- most policies are designed and implemented. How- riod of decades or centuries (see table 1.1). The ever, this prevents policy makers from reorienting countries in Europe and Central Asia are experi- public expenditures gradually, but systematically encing different stages of the demographic transi- to meet the transformation in the needs of citizens. tion, though at a far more rapid rate than Western An understanding of demographic shifts can help Europe and North America. The demographic governments shape public expenditures across a changes in the region have different implications wide range of areas, most obviously in the social for men and women, which are magnified because Table 1.1 | Demographic Profiles and Fertility High fertility low fertility Sex large increase in Moderate increase in and window of and window of imbalance the share of elderly the share of elderly Profile opportunity opportunity at birth women women Young population; Kyrgyz Republic, population pyramid Tajikistan, with large base and Turkmenistan, skinny top Uzbekistan Aging population; Armenia; Bosnia Armenia; population pyramid and Herzegovina; Azerbaijan; with balanced age Kazakhstan; Macedonia, structure Macedonia, FYR; FYR Moldova Aged population; Belarus, Bulgaria, Croatia, Czech Republic, population pyramid Estonia, Hungary, Latvia, Georgia, Montenegro, with small base and Lithuania, Russian Poland, Romania, Serbia, wide top Federation, Ukraine Slovak Republic, Slovenia Source: Chawla, Betcherman, and Banerji (2007). Note: Window of opportunity refers to an increase in the number of the potentially economically active population (between the ages of 20 and 59). Gender Issues in Human Capital 11 many of these societies—even the wealthier soci- tive or in low-productivity jobs, such as subsistence eties—have incomes per capita that are relatively farming. One key area of support in this high-fertility modest. Each stage of the demographic transition subregion would be the provision of some type of poses challenges that need to be addressed in the childcare services, combined with social assistance near future to benefit from the window of opportu- for low-income single-parent households. nity or to manage more effectively the consequenc- In a diverse group of countries in Europe and es of large structural changes.12 Central Asia, there will be more working-age women Yet, the various demographic transitions will and fewer children in absolute terms. The countries lead, in the next two decades, to a higher proportion in the demographic window of opportunity are Al- of elderly persons, particularly elderly women. Dur- bania, Armenia, Azerbaijan, Bosnia and Herzegovi- ing the next several decades, though the total pop- na, Kazakhstan, FYR Macedonia, and Moldova. For ulation of the region is not expected to change (a this group of countries, the number of children will −0.1 percentage decrease), there will be a dramatic fall by 1.9 million during 2000–25, while the num- change in the age structure of the population. The ber of women between 20 and 59 years of age will share of the population above 60 years of age will increase by 0.98 million during the same period. rise sharply, from 15 percent in 2009 to 25 percent Though these societies will experience declining in 2025 and 35 percent in 2050. Moreover, this fertility, the needs of the growing share of work- change in the age structure will not be gender neu- ing women for family-friendly policies and services tral (figure 1.6). (for example, childcare) will remain equally rel- The demographic transition in most of Central evant compared with Central Asia and, possibly, Asia is characterized by growth in the share of the even more relevant depending on the strength of population of prime age, providing an opportunity for informal family safety nets. greater economic growth. The four countries that The decline in fertility in some of the countries fall into this category are the Kyrgyz Republic, showing the demographic window of opportunity has Tajikistan, Turkmenistan, and Uzbekistan. These been accompanied by a substantial sex imbalance at countries are moving rapidly from the population birth. Three of the countries characterized by a explosion phase, with a proportionately large share shift toward the demographic window of opportu- of children, to the window of opportunity phase, nity show unbalanced sex ratios at birth. In Arme- with more working-age persons. The proportion nia, Azerbaijan, and FYR Macedonia, the number of the population between 20 and 40 years of age of missing girls is extremely high and comparable will rise from 30 to 33 percent between 2000 and with the levels in China, India, and the Republic of 2025. Consequently, the demands on key public Korea. Missing girls at birth reflect covert discrimi- services such as maternal and child health, as well as the continued demand for education services at all levels will continue. The working population be- 12 Demographers classify population pyramids according tween 40 and 60 years of age will more than double to four types, as follows: (a) demographic explosion in among men (110 percent) and women (114 per- young populations, (b) demographic window of oppor- tunity, (c) demographic implosion, and (d) demographic cent) during 2000–25. hourglass. The demographic explosion is the pyramid Yet, for Central Asia to benefit from the demo- with a large base and skinny top, while the demographic graphic window of opportunity, women of prime age implosion is the opposite, a pyramid with a small base need government support to facilitate their entry into the and a wide top. The demographic window is character- labor market, and especially into productive jobs. Sev- ized by a high proportion of the working-age popula- eral of the Central Asian countries have seen large tion while the demographic hourglass has parts of the working-age population missing and a large elderly pop- out-migrations of working-age men and the resulting ulation. Countries described by the demographic hour- feminization of the population. However, because glass show a rising dependency ratio, which reduces the of the lack of incentives and economic opportuni- potential of economic growth and increases the vulner- ties, many women remain either economically inac- ability of households to poverty. 12 Opportunities for Men and Women: Emerging Europe and Central Asia nation in the household, resulting from a combina- In the countries in the window of opportunity tion of strong preferences for sons, declining fertil- phase of the demographic transition, the aging of the ity, and prenatal sex determination. This problem population has begun, and this has important conse- has not been sufficiently studied in the region and quences for elderly women. In 2009, the proportion calls for greater attention. of the population above the age of 60 years was 11 Figure 1.6 Population Pyramids, Selected Countries and the Region, 2000, 2025, 2050 a. ECA – all countries but Turkey, 2000 b. ECA – all countries but Turkey, 2025 c. ECA – all countries but Turkey, 2050 100+ 100+ 100+ 90–94 90–94 90–94 80–84 80–84 80–84 70–74 70–74 70–74 60–64 60–64 60–64 50–54 50–54 50–54 40–44 40–44 40–44 30–34 30–34 30–34 20–24 20–24 20–24 10–14 10–14 10–14 0–4 0–4 0–4 –20,000 –10,000 0 10,000 20,000 –20,000 –10,000 0 10,000 20,000 –20,000 –10,000 0 10,000 20,000 d. ECA – young population countries, 2000 e. ECA – young population countries, 2025 f. ECA – young population countries, 2050 100+ 100+ 100+ 90–94 90–94 90–94 80–84 80–84 80–84 70–74 70–74 70–74 60–64 60–64 60–64 50–54 50–54 50–54 40–44 40–44 40–44 30–34 30–34 30–34 20–24 20–24 20–24 10–14 10–14 10–14 0–4 0–4 0–4 –3,000 –2,000 –1,000 0 1,000 2,000 3,000 –3,000 –2,000 –1,000 0 1,000 2,000 3,000 –3,000 –2,000 –1,000 0 1,000 2,000 g. ECA – aging population countries, 2000 h. ECA – aging population countries, 2025 j. ECA – aging population countries, 2050 3,000 100+ 100+ 100+ 90–94 90–94 90–94 80–84 80–84 80–84 70–74 70–74 70–74 60–64 60–64 60–64 50–54 50–54 50–54 40–44 40–44 40–44 30–34 30–34 30–34 20–24 20–24 20–24 10–14 10–14 10–14 0–4 0–4 0–4 –3,000 –2,000 –1,000 0 1,000 2,000 –2,000 –1,000 0 1,000 2,000 –2,000 –1,000 0 1,000 2,000 Female Male (Continues to next page) Gender Issues in Human Capital 13 Figure 1.6 Population Pyramids, Selected Countries and the Region, 2000, 2025, 2050. (Continued) j. ECA – aged population countries, 2000 k. ECA – aged population countries, 2025 l. ECA – aged population countries, 2050 100+ 100+ 100+ 90–94 90–94 90–94 80–84 80–84 80–84 70–74 70–74 70–74 60–64 60–64 60–64 50–54 50–54 50–54 40–44 40–44 40–44 30–34 30–34 30–34 20–24 20–24 20–24 10–14 10–14 10–14 0–4 0–4 0–4 –15,000 –10,000 –5,000 0 5,000 10,000 15,000 –15,000 –10,000 –5,000 0 5,000 10,000 15,000 –15,000 –10,000 –5,000 0 5,000 10,000 15,000 Female Male Source: World Bank staff calculations based on uN (2009a). Note: Young countries include the Kyrgyz republic, Tajikistan, Turkmenistan, and uzbekistan. The aging countries are Albania, Armenia, Azer- baijan, Bosnia and Herzegovina, Kazakhstan, FYr Macedonia, and Moldova. Aged countries are Belarus, Bulgaria, Croatia, the Czech republic, estonia, georgia, Hungary, Latvia, Lithuania, Montenegro, Poland, romania, russia, Serbia, the Slovak republic, Slovenia, and ukraine. each bar represents a five-year age-group. Women are indicated in blue; men are indicated in red. percent in the region, of which almost two-thirds care and appropriate parental leave) needs to be (59 percent) were women. Because of the acceler- in place, and there need to be changes in retire- ated aging, the proportion of the elderly will rise to ment laws. 22 percent, of which 57 percent will be women by 2025. The higher proportion of elderly women has potential fiscal ramifications through the pension Health and social protection systems. The majority of countries in Europe and Cen- Along most health indicators, the region is converging tral Asia are aging rapidly, and the number of em- toward the high-income OECD countries. In most of ployable persons is falling. A total of 17 countries in the indicators on demography and health, the re- the region are experiencing a demographic implo- gion performs well compared with Latin America sion.13 The number of persons between the ages or East Asia. Though the region lagged behind the of 20 and 60 years will fall from 182 million to OECD at the beginning of the transition, most in- 162 million between 2000 and 2025, equal to an dicators have converged during the past decade. 11 percent decrease. To prevent a sharp decline in However, there is significant diversity in the region: economic activity, one might try to maintain the selected countries show health indicators similar to size of the labor force by attracting temporarily or those in other, less-developed regions. permanently inactive women and retaining work- ing women in the labor force for a longer period of time by delaying retirement (so that it is equal 13 The countries are Belarus, Bulgaria, Croatia, the Czech to men’s retirement age). Yet, to achieve higher Republic, Estonia, Georgia, Hungary, Latvia, Lithuania, female labor force participation, the supporting Montenegro, Poland, Romania, the Russian Federation, institutional infrastructure (for example, child- Serbia, the Slovak Republic, Slovenia, and Ukraine. 14 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.7 Women’s Advantage in Life Expectancy, Selected Countries and Regions, 2005–10 averages, years 14 12 10 8 6 4 2 0 Sub-Saharan Africa South-Central Asia Northern Africa Eastern Asia Serbia Azerbaijan Macedonia, FYR Turkey Montenegro Bosnia and Herzegovina Tajikistan Western Europe Czech Republic Albania Uzbekistan Armenia Latin America and the Caribbean Croatia Aging Populations Young Populations Georgia Bulgaria Romania Slovenia Kyrgyz Republic Moldova Europe and Central Asia Central and Eastern Europe Slovakia Hungary Commonwealth of Independent States Turkmenistan Aged Populations Poland Latvia Estonia Ukraine Lithuania Belarus Kazakhstan Russian Federation Source: World Bank staff calculations based on uN (2009a). Life Expectancy and Mortality the Russian Federation. The large gap in life expec- Two drivers are determining the forecasted demo- tancy may be driven by the high adult male mortal- graphic transition: changes in mortality rates and ity rate observed in some of these countries. changes in fertility rates. Mortality rates and health The gender gap in life expectancy in the coun- outcomes have dramatically improved in the re- tries of the region needs to be narrowed. The average gion; in particular, during the last decade, the mor- change in the gap in life expectancy in the region tality rate has somewhat improved among men, has closed by one year; however, the variability in leading to a contraction of about a year in the gen- the change in the gap in life expectancy is large der gap in life expectancy. However, the high male (measured by statistical variance) (figure 1.8). In mortality rate relative to the female mortality rate Bosnia and Herzegovina, the gap has closed by 12 is persisting in many countries. This is associated years, mostly because of a recovery in male life ex- with a variety of factors, including alcohol abuse. pectancy. At the opposite extreme, the gender gap The gender gap in life expectancy in the region is in life expectancy in Belarus has increased in the larger than the global average, including Western Eu- past five years: while male life expectancy has in- rope, and varies greatly across countries within the re- creased (by 0.3 years), female life expectancy has gion. As a general average, women can be expected increased even more (1.8 years).14 to outlive men by four years, but, in Europe and During the last decade, the negative trend in Central Asia, this difference is seven years (figure male life expectancy that has been common in certain 1.7). In all the countries in the region, the differ- countries of the region has reversed. Although the ence in life expectancy surpasses the world average average trend in life expectancy in the region has of four years. In the countries in which the gap is been positive for men and women since the beginning smaller (Azerbaijan, FYR Macedonia, Montenegro, Serbia, and Turkey), the difference is only slightly more than four years. However, the gap is three 14 Figure 1.8 shows changes in the gender gap in life times larger in Belarus, Kazakhstan, Lithuania, and expectancy; annex A, table A.1 shows the values. Gender Issues in Human Capital 15 Figure 1.8 Changes in the Gender Gap in Life Expectancy, 1990–2010 4 2 0 –2 –4 –6 –8 –10 –12 –14 Bosnia and Herzegovina Azerbaijan Latvia Aging Populations Tajikistan Montenegro Croatia Western Europe Hungary Czech Republic Central and Eastern Europe Europe and Central Asia Sub-Saharan Africa Georgia Estonia Slovakia Slovenia Poland Kyrgyz Republic Young Populations Aged Populations Serbia Armenia Commonwealth of Independent States Romania Bulgaria Latin America and the Caribbean Uzbekistan Turkmenistan Macedonia, FYR Moldova Russian Federation Albania Lithuania Turkey Ukraine Northern Africa Eastern Asia Belarus South-Central Asia Kazakhstan Source: World Bank staff calculations based on uN (2009a). of the transition, Belarus, Kazakhstan, Montenegro, below. During the past decade, all these countries have Russia, and Ukraine experienced a negative trend in seen an increase in male life expectancy, despite the life expectancy during the 1990s. Average male life continuing high adult mortality rates in Belarus, Rus- expectancy fell from 64 to 62 years during the 1990s sia, and Ukraine (figures 1.9 and 1.10). in this set of countries. This negative trend seems to The male mortality rate among adults between have been associated with alcohol abuse, as discussed 20 and 59, although still higher than the correspond- ing female rates, has fallen from the high values ob- served in the 1990s. In the region, the average male mortality rate for adults 20 to 59 years of age fell Figure 1.9 Trends in Life Expectancy, from 223 in 2000 to 205 in 2006. The female adult Selected Countries and the mortality rate in the region is considerably lower Region, 1990–2015 and has been more stable since the transition at the 85 beginning of the 1990s. However, the male adult 80 mortality rate decreased noticeably in the last de- 75 cade. This decline occurred everywhere except the 70 Kyrgyz Republic and Ukraine, which has shown an 65 increase in the mortality rate equivalent to the one 60 observed in the 1990s. The high male mortality 55 rate countries—Belarus, Lithuania, and Russia— 50 1990–1995 1995–2000 2000–2005 2005–2010 2010–2015 showed some improvement, but remain among the Female - ECA countries with the highest rates compared with the Male Russian Male Federation Montenegro rest of the region (figure 1.10). Male - ECA Male Male Possible determinants of the high mortality rate Ukraine Kazkahstan are associated with alcohol abuse. The high male mortality rate in many of the countries in the region Source: World Development indicators Database, World Bank, Wash- ington, DC, http://data.worldbank.org/data-catalog/world-develop- is associated with alcohol abuse and other, related ment-indicators/. risks. A recent World Bank report concluded that 16 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.10 The Evolution in Mortality Rates among the 20–59 Age-Group a. Europe and Central Asia, by gender, 1990–2006 b. Males, selected countries, 1990–2007 350 600 300 500 250 400 200 300 150 100 200 50 100 0 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Male – ECA Average Male – Selected Countries Female – ECA Average Female – Selected Countries Lithuania Belarus Ukraine Russian Federation Source: World Bank staff calculations based on uN (2009a). Note: Selected countries are Belarus, Bulgaria, Czech republic, estonia, Hungary, Latvia, Lithuania, Poland, romania, russia, Slovak republic, Slovenia, and ukraine. Russians, especially men, suffer disproportionately Box 1.1: Russia’s Growing Health Crisis from (1) injury and violence, including traffic ac- Every day, 100 people die in traffic accidents in Russia, and over cidents, domestic and other violence, and suicide; 100 people die of acute alcohol poisoning. The excessively high (2) cardiovascular disease; and (3) cancer (see box levels of mortality, ill health, and disability among the working- 1.1). All these causes are affected by excessive alco- age population have far-reaching demographic, financial, and social consequences. (The evolution in mortality rates in the hol, tobacco, and other drug use. The relationship 20–59 age-group is shown in figure 1.10, for example.) A recent with changes in economic growth suggests that the report of the World Bank’s Europe and Central Asia Region con- cluded as follows: lack of economic opportunities and the ensuing mental stress continue to play a crucial role. • Economic stress, coupled with a tradition of unhealthy life- The high male mortality rate is not a new phe- styles and unhealthy environments, has led to a reversal in life expectancy, especially among adult men. This is worsening nomenon in Russia. The average mortality rate the demographic trends in a rapidly aging population. among adult Russian males is now on a par with • Noncommunicable diseases, notably cardiovascular diseases and cancer, as well as injuries, are the main causes of death the rates in Ethiopia, Haiti, and Sierra Leone. in Russia. The mortality rates associated with these diseases The Russian male adult mortality rate has shown and with injuries are three and five times higher, respectively, several spikes over the years, usually associated than the corresponding rates in the European Union. Traffic injuries, suicide, alcohol poisoning, and violence account for with the economic and political situation (Nolte, the main share of lethal injuries. McKee, and Gilmore 2004). The slight increase • Alcohol and other drug abuse, as well as heightened tobacco consumption, increases the probability of these diseases, as in male mortality during the 1960s, 1970s, and do poor diets and stress caused by worsening socioeconomic 1980s reversed in the mid-1980s concurrently conditions. with political changes. The male adult mortality • The ill health and high mortality rates result in fewer produc- tive workers. rate then fell between the mid-1990s until the • Regional disparities and even national security risks are Russian economic crisis in 1998. Since then, how- growing. Health care costs and the costs associated with the loss of production because of absenteeism or low productiv- ever, the male mortality rates for ages 25–59 have ity are high. been rising again, despite high economic growth. Finally, 2007 may turn out to be the beginning of Source: World Bank (2005a). a new decreasing trend. Gender Issues in Human Capital 17 Reproductive Health and Maternal Mortality Kyrgyz Republic, Tajikistan, and Turkmenistan are At the beginning of the transition, women across the three countries with the highest rates, while Po- the region enjoyed relatively good health and ac- land, Serbia, and the Slovak Republic are the three cess to basic health services. The only major area of countries with the lowest rates (figure 1.11). concern was the lack of access to modern forms of Improvement in the maternal mortality rate has contraception and the heavy reliance on abortion. been achieved in selected countries in the region. De- The change in the trend that began in the 1990s spite the convergence in the 1990s, two countries in persisted, and a convergence toward OECD coun- Europe and Central Asia still showed higher mater- try levels is observable today. nal mortality rates in 2000 than the average in Latin The maternal mortality rate in the region has con- America or East Asia. In 2000, the maternal mortali- verged to the level of the rates in the OECD countries. ty rate was 91 in Turkmenistan and 120 in Tajikistan At the beginning of the 1990s, maternal mortality (figure 1.12). However, significant improvement was rates in Europe and Central Asia, though relatively observed in other countries of Europe and Central low compared with Latin America and East Asia, Asia, including Azerbaijan, Kazakhstan, Latvia, Ro- were high compared with developed countries. In mania, Russia, Turkey, and Ukraine. The improve- 1990, the maternal mortality rate in the region was ments in these countries were mainly responsible almost two-thirds higher than the rate observed in for the convergence of the rates in the region toward OECD countries: 49 versus 30 deaths per 100,000 the rates in the countries of the OECD. live births, respectively, in Europe and Central Asia However, it is not clear why improvements in the and in the OECD.15 However, during the last two maternal mortality rate were not achieved in all coun- decades, the maternal mortality rate has converged tries. In a region in which almost all births are attend- toward the levels of the rate in the OECD (figure ed by health personnel, why is there such a high level 1.10). In 2008, the average maternal mortality rate of maternal mortality in some countries? The reasons in the region was 27 per 100,000 live births. may range from poverty at the household level (given However, there is enormous variation in mater- that the poorest countries in the region experience nal mortality rates across the region. The difference the highest mortality rates) to the specific character- in the maternal mortality rates of the countries with istics of the health system in each country (given the the highest and the lowest rates is 75 per 100,000 high share of births attended by skilled health per- live births, which is three times the value of the sonnel). In the region, Tajikistan has the lowest share maternal mortality rate in the OECD in 2008. The of births attended by skilled health personnel; yet, only 12 percent of births occur unattended. Figure 1.11 Maternal Mortality Rates, Infant and Under-5 Mortality 1990–2008 Infant and under-5 mortality rates in Europe and 60 Central Asia compare well with rates in East Asia, Maternal mortality rate 50 Latin America, and the OECD. However, the varia- 40 tion across countries within the region is high. The 30 20 Central Asian countries, in particular, show mortal- 10 0 1990 1995 2000 2005 2008 15 The maternal mortality rate is the annual number OECD ECA of deaths among women due to pregnancy or child- birth-related causes per 100,000 live births. We use Source: World Development indicators Database, World Bank, Wash- the model estimates. For more details, consult World ington, DC, http://data.worldbank.org/data-catalog/world-develop- Development Indicators Database, World Bank, Wash- ment-indicators/. ington, DC, http://data.worldbank.org/data-catalog/ Note: eCA = europe and Central Asia. world-development-indicators/ (accessed 2011). 18 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.12 Maternal Mortality Rates by Country, 2000 and 2008 140 Maternal Mortality Rate 120 100 80 60 40 20 0 Kyrgyz Republic Turkmenistan Tajikistan Georgia Kazakhstan Russian Federation Azerbaijan Moldova Albania Uzbekistan Armenia Romania Ukraine Turkey Latvia Slovenia Belarus Montenegro Croatia Bulgaria Hungary Lithuania Estonia Bosnia Macedonia, FYR Czech Serbia Poland Slovak Republic 2000 2008 OECD 2009 LAC 2008 EAP Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. LAC = Latin America and the Caribbean. ity rates that are above the average rates in East Asia 1.14 and 1.15). Infant and under-5 mortality rates and Latin America. Nonetheless, in recent decades, are closely correlated, which is perhaps indicative the improvements in these countries has been signif- of the underlying quality of the health care services icant, although generally below the regional average. available for children. While the infant and under-5 The average infant mortality rate in Europe and mortality rates in the Czech Republic and Slovenia Central Asia is higher than the average rate in the are considerably below the averages in the OECD, high-income OECD countries, but converging. Dur- the rates in the countries of Central Asia and the ing the 1990s, the average infant mortality rate in south Caucasus are significantly higher than the Europe and Central Asia converged toward the regional average. The under-5 mortality rate was high-income OECD average. The rate fell from 65 per 1,000 live births in Tajikistan, but only 3 in 33 per 1,000 live births in 1990 to 23 in 2000.16 Slovenia. Similarly, the respective infant mortality During the last decade, the rates in the region have rates in these two countries were 54 and below 3 continued to improve. In 2009, the average infant per 1,000 live births. The significant variation in and under-5 mortality rates in the region were the infant and under-5 mortality rates is closely only 16 and 19 per 1,000 live births, respectively correlated with gross national income per capita, (figure 1.13). with the exception of Azerbaijan, Kazakhstan, and Although at a slower pace, the average under-5 Turkmenistan, where mortality rates are well above mortality rate in the region has also been converging what one might predict based on income. Turkey is toward the rate in the OECD. The average under-5 another example. There, income per capita would mortality rate in the region fell from 40 per 1,000 suggest the infant mortality rates would be some- live births in 1990 to 28 in 2000 and 17 in 2009. what lower. It may be, in this case, that changes in Despite the important progress observed in the re- health care outcomes are not keeping up with the gion, the average under-5 mortality rate was above high rates of economic growth. the average in the OECD, which was 9 per 1,000 live births in 2009 (figure 1.13). In the region, there is large variation across coun- 16 The average infant mortality rate in high-income tries in infant and under-5 mortality rates (figures OECD countries in 2000 was 6 per 1,000 live births. Gender Issues in Human Capital 19 Figure 1.13 Infant and Under-5 Mortality Rates, 1990–2009 a. Infant and under-5 mortality rates b. Infant mortality rates, 2009 per 1,000 live births per 1,000 live births 45 60 40 35 50 30 25 40 20 15 30 10 5 20 0 10 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 0 0 10,000 20,000 30,000 ECA - High income ECA - High income IMR OECD - IMR Child MR OECD - Child MR GNI per capita, PPP (current intl $) Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eCA = europe and Central Asia. Figure 1.14 Under-5 Mortality Rates, 2000 and 2009 100 Mortality Rate, under 5 (per 1000) 90 80 70 60 50 40 30 20 10 0 Kyrgyz Republic Uzbekistan Tajikistan Croatia Latvia Bosnia and Herzegovina Czech Republic Ukraine Slovenia Azerbaijan Albania Turkmenistan Belarus Moldova Bulgaria Macedonia, FYR Georgia Poland Romania Armenia Russian Federation Montenegro Estonia Turkey Serbia Hungary Lithuania Kazakhstan Slovak Republic 2000 2009 OECD 2009 LAC 2009 EAP 2009 Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. LAC = Latin America and the Caribbean. Belarus, Estonia, and Turkey have experienced der-5 mortality rates in the countries with the high- the biggest drops in infant and under-5 mortality rates est rates in the region has been mostly below the over the last decade, but many other countries in the regional average, indicating a lack of convergence region have seen big improvements as well. Infant between the south Caucuses, Central Asia, and the and under-5 mortality rates dropped, on average, rest of the region. In Azerbaijan, the infant mortal- by one-third in the region during 2000 and 2009, ity rate fell from 78 per 1,000 live births in 1990 while, in Belarus, Estonia, and Turkey, rates fell by to 56 in 2000 and 41 in 2009. In Tajikistan, the over 50 percent. The decline in the infant and un- country with the poorest performance, the under-5 20 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.15 Infant Mortality Rates 80 70 60 50 40 30 20 10 0 Tajikistan Turkmenistan Uzbekistan Azerbaijan Kyrgyz Republic Kazakhstan Georgia Armenia Albania Moldova Turkey Romania Ukraine Bulgaria Macedonia, FYR Russian Federation Latvia Montenegro Bosnia and Herzegovina Serbia Hungary Lithuania Poland Croatia Estonia Belarus Czech Republic Slovenia Slovak Republic 2000 2009 OECD 2009 LAC 2009 EAP 2009 Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. LAC = Latin America and the Caribbean. mortality rate fell from 117 per 1,000 live births in ment level and below the average in the OECD. 1990 to 93 in 2000 and 65 in 2009. Among the Central Asian countries, Tajikistan had the highest fertility rate (3.4) in 2009, and Turk- menistan had the lowest (2.4). Fertility Rates In the region over the last two decades, total fer- The substantial variation in the fertility rate across tility rates have been diverse across countries and over the region is consistent with the diversity observed in time. During the 1990s, the total fertility rate fell in the demographic transition. The total fertility rate in all countries in the region except Croatia. In some 19 countries in the region is around 1.8 births per countries, such as Turkmenistan, the decline was woman (figure 1.16). The Central Asian countries, considerable (from 4.5 in 1990 to 2.8 in 2000), which have younger populations, have high fertility while, in other countries, such as Montenegro, rates, while countries with populations classified as it was insignificant (from 1.87 in 1990 to 1.82 in aging or aged have fertility rates below the replace- 2000). However, during the last decade, the picture Figure 1.16 The Evolution of the Total Fertility Rate, 1990–2009 3.5 3.0 Total Fertility Rate 2.5 2.0 1.5 1.0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 EAP OECD LAC ECA Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. eCA = europe and Central Asia. LAC = Latin America and the Caribbean. Gender Issues in Human Capital 21 Figure 1.17 The Variation in Total Fertility Rates across Countries, 2009 4.0 3.5 3.0 Total Fertility Rate 2.5 2.0 1.5 1.0 0.5 0.0 Tajikistan Kyrgyz Republic Kazakhstan Uzbekistan Turkmenistan Azerbaijan Turkey Armenia Montenegro Estonia Bulgaria Georgia Albania Lithuania Russian Federation Slovenia Czech Republic Moldova Croatia Ukraine Macedonia, FYR Belarus Poland Romania Hungary Latvia Bosnia and Herzegovina Slovak Republic 2009 OECD LAC EAP Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. LAC = Latin America and the Caribbean. was completely different. In half the countries in and Ukraine are among the top five countries in the the region, there was an increase in the total fertility world in terms of the number of children put up for rate, while, in the other half, there was a decline.17 intercountry adoption.18 Many of the children are The magnitude of the decline in the total fertility the result of unwanted pregnancies, thus suggest- rate during the 1990s does not seem to have been ing that contraceptive methods are not sufficiently correlated with the changes in fertility in the 2000s, widespread and that there may be a social stigma on indicating that many factors, including the econom- single motherhood, since most of the mothers giv- ic situation and societal changes, are likely to be in ing their children up for adoption are young and un- play. For example, in Azerbaijan and Kazakhstan, married. (The abortion rates in Russia and Ukraine the total fertility rate decreased by 0.67 and 0.72, are among the highest in the region; see below.) respectively, during the 1990s, but has increased by Income per capita does not seem to be related to 0.3 and 0.8, respectively, since then. Meanwhile, the changes in the total fertility rate. There are many fac- total fertility rate fell by 0.8 and 0.9 in Tajikistan tors that explain the total fertility rate. One of them during each decade, respectively (figure 1.17). is income per capita. Thus, in many countries, an Thus, the declining trend in the average total fer- economic boom translates into a boom in fertility, tility rate in the region during the 1990s has changed to a steady trend in the new millennium. During the 1990s, there was a drastic decline in the average to- 17 Albania, Bosnia and Herzegovina, Georgia, FYR Mace- tal fertility rate in Europe and Central Asia, which donia, Moldova, Montenegro, Poland, Slovak Repub- resulted in a convergence toward the rates in the lic, Tajikistan, Turkey, Turkmenistan, and Uzbekistan OECD. The average total fertility rate in Europe showed a decrease in the total fertility rate in both and Central Asia fell from 2.4 in 1990 to 1.8 in decades, while Azerbaijan, Belarus, Bulgaria, the Czech 2000, the average in the OECD. In 2009, the av- Republic, Estonia, Hungary, Kazakhstan, the Kyrgyz erage rate in Europe and Central Asia was still the Republic, Latvia, Lithuania, Romania, Russia, Slovenia, and Ukraine showed a decline in the total fertility rate same (figures 1.16 and 1.18). during the 1990s and a rise during the next decade. Selected countries in the region are key sources of 18 The other three countries are China, Guatemala, and the children involved in intercountry adoptions. Russia the Republic of Korea. 22 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.18 Changes in the Total Fertility Rate across Countries, 1990–2009 1.0 0.5 0.0 –0.5 –1.0 –1.5 –2.0 –2.5 Albania Bosnia and Herzegovina Georgia Macedonia, FYR Moldova Montenegro Poland Slovak Republic Tajikistan Turkey Turkmenistan Uzbekistan Azerbaijan Belarus Bulgaria Czech Republic Estonia Hungary Kazakhstan Kyrgyz Republic Latvia Lithuania Romania Russian Federation Slovenia Ukraine 1990–1997 1998–2009 Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. which may explain the increase in fertility in coun- Abortion rates are decreasing in the region as tries such as Tajikistan. However, the data do not women obtain access to modern methods of contra- indicate any association in the region between the ception. The average abortion rate in the region total fertility rate and GDP per capita or growth in declined considerably between 1999 and 2008, GDP per capita. from 25 to 15 abortions per 1,000 women 15 to The low fertility rate is a concern among pol- 49 years of age. However, the average is still higher icy makers because of the implications for the demo- than the rates in the OECD. The abortion rate was graphic transition and because the fertility rate does only 6 in Germany and Switzerland, for example not seem to be responding in the short run, at least, to (figure 1.19). policy changes. Some countries, such as Croatia and Despite the declining trend in the region, the Romania, are concerned about the implications abortion rate remains high in many countries; it has of the low fertility rate and have adopted mecha- even increased in Georgia. The abortion rate in Es- nisms to encourage families to have children. We tonia, Romania, and Russia is more than two times note two of these policies: first, these countries are higher than the average in Europe and Central Asia. generous in terms of maternity leave coverage, but In Georgia, the abortion rate increased during the not as generous in parental leave; second, child al- last decade. Additionally, in the region, there is a lowances are provided instead of childcare services. positive association between the use of modern methods of contraception and abortion rates. In three countries (Estonia, Romania, and Russia), the Abortion and Contraceptive Use use of contraception is above 70 percent among The variables commonly used to explain fertil- women between 15 and 49 years of age. ity are marriage rates, abortion rates, contracep- The age at first birth does not seem to be a deter- tive use, the duration of breastfeeding, and age at minant of fertility in most countries in the region. In first marriage. Marriage rates, although relevant, countries with higher fertility rates, women tend to are becoming less important in explaining fertility have their first born at a younger age. However, the because of the increasing number of births out of fact that the fertility rate is low in these countries wedlock. Thus, although 70 percent of children are (slightly above 2 children per woman) and that the born in wedlock, the share of extramarital births average age at first birth is relatively high (above 27 has increased by 50 percent since the late 1990s. years of age) suggests that there is probably no caus- Gender Issues in Human Capital 23 Figure 1.19 Abortion Rates across Countries, 1999 and 2008 per 1,000 live births 60 50 40 30 20 10 0 Poland Croatia Azerbaijan Slovak Republic Tajikistan Georgia Czech Republic Armenia Slovenia Serbia Kyrgyz Republic Lithuania Moldova Albania Hungary Latvia Kazakhstan Bulgaria Ukraine Estonia Romania Belarus Russian Federation Europe and Central Asia Spain Germany Switzerland Finland Italy Iceland Norway Sweden United States 1999 2008 Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: eCA = europe and Central Asia. al relation between age at first birth and fertility. countries at the two extremes—Georgia and Slove- This conclusion is supported by two observations. nia—is on the order of 10 times (figure 1.21). How- First, the average age at first birth in the region is ever, though adolescent fertility should always be slightly lower than the average age in OECD coun- a concern, it is low in the region relative to other tries, 27.5 versus 28.4 years, respectively. Second, a developing regions. simple correlation between the average age at first Adolescent fertility rates are not correlated birth and the total fertility rate shows no significant with total fertility rates at the country level. This is association between these two variables. positive because it implies that adolescents are not The rate of adolescent pregnancy in the region driving the fertility rates observed in the region. is low and is decreasing over time. In the region, the Nonetheless, for health reasons and for the sake rate of adolescent pregnancy—the number of live of future labor market opportunities, it is not ben- births to women between 15 and 19 years of age eficial for adolescents to bear or take care of chil- per 1,000 women in the age-group—is below the dren. In our case, the lack of correlation between average in the OECD, although it is still above the the two rates may imply that adolescent mothers average in high-income OECD countries. The re- either do not have information about the benefits gion’s average rate of adolescent pregnancy was 23 of delaying birth or do not have access to educa- in 2008, while, in the OECD, the rate was 27, and, tion and labor market opportunities. Whatever the in the high-income OECD countries, it was 17. The explanation, this should be a matter of concern for average adolescent pregnancy rate in the region fell policy makers not least because it indicates there is by about 30 percent during the last decade given room for efficiency gains in labor resources. that the rate was 33 in 1998 (figure 1.20). There is significant variation across the region in the adolescent fertility rate. The adolescent fertility Missing Women rate is high in the countries of the south Caucasus The term missing women refers to the observa- and in Bulgaria, Moldova, Romania, Ukraine, and tion that the proportion of girls and women in the Turkey. The difference in the rate between the population of a country is too low. The indicator is 24 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 1.20 Adolescent Fertility Rate, Selected Regions, 1998–2009 40 Adolescent fertility rate 30 (15–19) 20 10 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 ECA OECD High Income OECD EAP Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. eCA = europe and Central Asia. Figure 1.21 Adolescent Fertility Rates across Countries, 2009 45 40 35 30 25 20 15 10 5 0 Georgia Bulgaria Turkey Armenia Azerbaijan Moldova Kyrgyz Republic Romania Kazakhstan Ukraine Tajikistan Russian Federation Serbia Lithuania Macedonia, FYR Estonia Belarus Slovak Republic Hungary Turkmenistan Latvia Montenegro Albania Croatia Poland Uzbekistan Czech Republic Slovenia Bosnia EAP High Income OECD Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. computed using the mortality risk of women and of sex imbalance that are equally high. In China, men at every age in a country relative to the mortal- 118 baby boys are born for every 100 live births of ity risk at the same ages in a group of high-income girls, while, in Armenia, Azerbaijan, and Georgia, reference countries. the corresponding numbers are 116, 115, and 111, Armenia, Azerbaijan, and Georgia have the respectively (figure 1.22). highest sex imbalance at birth in the world after Chi- GDP growth is not driving the sex imbalance at na. Although most of the attention regarding miss- birth, which means that three other factors may be at ing women is focused on China and India, the three play. A simple correlation of GDP per capita (or countries in Europe and Central Asia show levels GDP growth) and sex imbalance at birth indicates Gender Issues in Human Capital 25 Figure 1.22 Sex Ratio at Birth, 2008 120 115 110 105 100 95 China Armenia Azerbaijan Korea, Rep. Montenegro Serbia Macedonia, FYR India Albania Bosnia and Herzegovina Bulgaria Croatia Czech Republic Estonia Hungary Kazakhstan Kyrgyz Republic Latvia Lithuania Poland Moldova Romania Russian Federation Slovak Republic Ukraine Slovenia Tajikistan Turkey Turkmenistan Uzbekistan Source: uN (2009a). no association.19 There are three other factors that are located in underserved communities may also influence the sex imbalance at birth: first, the de- be of poor quality and possess inadequate human cline in fertility may contribute to the emergence and financial resources. In addition, there may be of the choice of sex in the child; second, ultra- unaddressed cultural constraints explaining why sound technology has became widely available, al- girls are kept home, while boys are sent to school. lowing for prenatal sex discrimination; and, third, To address these challenges, governments need the preference for sons among some families re- to make a concerted effort to promote access to mains strong, at least in selected developing coun- quality education for poor and minority children, tries. The combination of all these factors under reduce school segregation, train teachers and assis- the different models of choice in fertility can be tants to deal with the special challenges, and, per- detrimental.20 haps, introduce incentives to encourage families to send their daughters through schemes such as the conditional cash transfers used by several coun- Disparities within Countries tries in Latin America and even the western Bal- It is important to acknowledge that gender differ- kans (for example, FYR Macedonia). Finally, it is ences are usually exacerbated within each country so that poor women are less able than wealthier women to defend themselves against discrimina- 19 More details about this association may be found in tion and the lack of access. In Europe and Central chapter 3 of World Development Report 2012: Gender Asia, Roma women offer an example (box 1.2). Equality and Development (World Bank 2011a). 20 Various models that include a preference for sons may be used to explain fertility. For example, couples may decide, in an initial stage, on the number of children The Implications for Policy Design they desire and then use ultrasound technology to real- ize the preference for sons. Thus, ultrasound technol- To close the gender gap in secondary schools, special ogy is making other models of fertility obsolete. An emphasis needs to be placed on children in poor or example of a model that might be becoming obsolete is the stopping model, whereby families have children minority groups. A significant share of this gender until they produce a son. Under such model assump- gap can be explained by the low participation of tions, the mother may decide on how many children to children in poor and minority groups, especially have. Moreover, there are studies that show that the sex in Central and Eastern Europe. Many schools that of the second born is related to the sex of the first born. 26 Opportunities for Men and Women: Emerging Europe and Central Asia productivity of new entrant workers, and promot- Box 1.2: Health Access among Roma Women ing research and development in environmentally At a population of 12 million (according to some estimates), friendly industries and processes. Roma constitute the largest minority in Europe. The existing The problem of the missing women in the region evidence suggests that Roma have generally poor health out- comes and face significant barriers to adequate health care, calls for more attention to provide evidence-based pol- including discrimination on the basis of their ethnicity. icies. It is important to understand if this problem Roma women have less access to health services com- pared with non-Roma women. The lack of access to health ser- is common to all households or particular to the vices among Roma women is more significant in birth delivery bottom of the income distribution. For example, than in prenatal care: 95 percent of Roma women in FYR Mace- if the gender imbalance at birth is common to all donia and 89 percent in Serbia received prenatal care in 2006 compared with 99 percent of non-Roma women. In birth at- households, one policy used in many countries tendance, the difference between groups is similar: on average, such as China and India is to ban prenatal screen- 82 percent of Roma women in FYR Macedonia and 93 percent in Serbia give birth with the assistance of skilled personnel, ing.21 However, policies should generate incentives while the average among non-Roma women in these countries to prevent prenatal screening because banning pre- is 99 percent. natal screening is debatable on the grounds of free Abortion rates are higher among Roma women. In survey studies conducted in Serbia, half of all Roma women reported choice, but also because there are plenty of ways having had at least one abortion in their lifetimes, and 7 percent to bypass the law. Thus, policies should be de- of all Roma women reported undergoing more than 10 abor- tions. These rates are extremely high compared with the na- signed to discourage prenatal screening by raising tional average, which is about 10 percent. Moreover, anecdotal the associated costs, for example, health insurance evidence indicates cases of unsafe abortion (WHO 2010). might not cover the test for prenatal screening or high tariffs might be placed on medical screening devices by private companies or individuals. If this problem is found only at the bottom of the also important to address the underlying concerns income distribution, additional instruments such of parents about sending girls to school, which may as conditional cash transfers can be used. In the sometimes be as simple as providing safe transpor- state of Haryana, India, there is such a program tation for girl students. whereby families receive an immediate cash com- The demographic transition that is under way pensation if they give birth to a daughter and a has raised government interest in increasing fertility, long-term savings bond redeemable on the daugh- although a more realistic alternative may be to focus ter’s 18 birthday provided she’s unmarried, with resources on making the next generation more pro- additional bonuses for education. In addition, the ductive. Traditionally, a focus on fertility has been problem of missing women accumulates through considered a policy to mitigate the demographic the life cycle if girls and women have differential transition to elderly populations. Most demog- access to health and nutrition, leading to a differ- raphers used to recommend replacement-level ent set of policies. fertility rates (that is, an average of two surviving To address the maternal mortality resulting from children per adult woman). However, today, the induced abortions, greater emphasis on modern con- world total fertility rate is below this level, and this policy recommendation seems unrealistic. Govern- ment policies geared toward reversing declining 21 Austria, the Republic of Korea (Asia), New Zea- fertility have not been widely successful even in land (Oceania), Switzerland, and Vietnam fully pro- the more advanced economies. Thus, policies are hibit prenatal screening. Australia, Belgium, Bosnia and Herzegovina, Bulgaria, Canada, China, Croatia, moving toward ways of increasing the value gener- Cyprus, Denmark, Estonia, Finland, France, Germany, ated by newer population cohorts so that children Greece, Hungary, Iceland, India, Israel, Italy, Latvia, will be able to produce more for a larger popula- Lithuania, the Netherlands, Norway, Portugal, Russia, tion in a context of environmental sustainability. San Marino, Singapore, Turkey, and the United King- These types of policies thus advocate raising the dom prohibit prenatal screening for social reasons human capital of younger cohorts, increasing the (Darnovsky 2009). Gender Issues in Human Capital 27 traception methods is necessary. The government will tions in the areas of alcohol and tobacco abuse, need to consider launching information campaigns changes in diet, and the promotion of physical that demystify the unintended effects of contracep- activity need to be developed. The secondary pre- tives, training practitioners to offer a wider range of vention programs should address hypertension, contraceptives to clients, understanding the needs cholesterol, and diabetes control. In addition, it is and preferences of clients, increasing the availabil- important to pursue improved road safety, as well ity of modern contraceptives free of charge or for as emergency medical services. These need to be a nominal price, and diminishing the coverage sup- complemented at the federal level by (1) adopt- plied by public health systems for abortion costs. ing appropriate legislation, policies, and strategies Evidence suggests there is partial knowledge about to support the program interventions, (2) build- modern contraceptives among both clients and ing institutional capacity, and (3) assuring federal health care providers. Clients perceive oral contra- oversight and accountability. ception as unreliable and are more likely to inter- The aging of the population may require sig- rupt the use of this method, which will result in nificant changes to social pension programs in the unwanted pregnancies. Health practitioners in the region. Although, at present, a significant share of region recommend condoms, intrauterine devices, the elderly receive benefits through pension pro- and oral contraceptives and do not recommend grams, this may no longer be the case in the future other methods (such as injectables, Norplant, and given the increased informality in the labor mar- sterilization). In addition, although all contracep- ket. However, many governments in the region do tive methods are legal, availability is limited to the provide social pensions, that is, cash transfer pro- more traditional methods. Finally, contraceptive grams that help the elderly who have no recourse methods are not always available through public to any other form of income. An expansion of health services, nor are they necessarily supplied these programs may be necessary in the future, free of charge. Moreover, the quality of available though these programs will be competing with contraception is poor and should be improved and other public expenditures. Depending upon the even fostered through research and development coverage of the existing pension systems, govern- on new methods. Increasing the use of modern ments may need to rethink their public pension contraceptive methods is key in the region because systems to determine the optimal (and afford- of the high abortion rates. The purpose would be able) way to ward off old age poverty. According not only to reduce maternal morbidity and ma- to one proposal, all elderly would receive a small ternal mortality, but also to focus on the fact that social pension that would be affordable and pro- modern contraception is a more cost-effective way vide some small income to all elderly. However, of family planning than abortion, particularly given this would need to be developed in coordination the limited physical availability of hospitals in some with revisions to the public pension system, as countries. well as the social assistance system. Many of the male mortality factors can be rem- The long-term care requirements of an aging edied at low cost. The evidence appears to indicate population call for innovative responses to compen- that the excessively high rates of deaths among sate women family members who provide the major- prime age males are largely a result of noncommu- ity of this care. Throughout Europe, but also in the nicable diseases (of the circulatory system) and, to region, the majority of long-term care is provided a secondary degree, traffic accidents. To address through informal home care arrangements rather this, a comprehensive national program operat- than by public or private state institutions. In the ing at all levels of government (federal, regional, future, reforms in the financing and delivery of and municipal) is necessary. At the regional level long-term care will need to be addressed, especial- (especially, in the larger countries such as Russia), ly in the case of informal caregivers. Because the new programs need to be developed. First, prima- burden of the care of the elderly will only increase ry prevention programs that undertake interven- in the future, there is a strong possibility that the 28 Opportunities for Men and Women: Emerging Europe and Central Asia current practice of unpaid informal caregivers formal caregivers may need the support of profes- may not be sustainable. Some countries, such as sionals because some tasks associated with these Austria, Germany and selected Nordic countries, activities may be well beyond their capability. In provide pension credits to informal caregivers the Netherlands, informal and formal long-term given that these are women who have forfeited care is provided, although this requires strong in- the opportunity of employment. In addition, in- stitutional capacity. Gender Issues in Human Capital 29 Women in the Labor Market 2 Introduction While the second decade of the transition to a market economy brought economic growth to Europe and Central Asia, the growth varied across countries. Until the recent global crisis, growth was robust, yielding better living standards and lower levels of pov- erty. Many of the reforms adopted since the fall of the Berlin Wall had begun to mature and yield results. Ten new member states of the European Union (EU) from the region showed strong indications of convergence with the high-income economies of the EU.22 However, despite all of the favorable economic and social changes, the labor market was an exception; it was less responsive to economic growth; and labor force expansion and employment growth were minimal during this period (World Bank 2005b). Developments in labor markets have significant implications for living standards. Because differences in labor market outcomes among men and women are widespread and persist globally, it is useful to distinguish labor market analysis by gender. The Eu- rope and Central Asia region is no exception. The countries with substantial economic growth were also the countries with relatively higher female labor force participation and lower wage gaps between men and women (see figure 2.1). Yet, despite this positive association among important indicators of welfare, the fact remains that women have lower labor market presence and earn lower wages than men. However, men and women in the labor market are different not only with re- spect to their labor market status, but also with respect to the reasons why they hold certain positions in the labor market. This means there are major implications for social policy. Thus, the greater work burden of women within the household (for 22 The EU10 consists of Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic, and Slovenia; see annex A for a complete list of the countries in Europe and Central Asia considered in our analysis. 31 Figure 2.1 Correlation among Per Capita GDP Growth, Labor Force Participation, and the Wage Gap a. Female labor force participation b. Gender wage gap 60 60 50 50 40 y = 0.4911x + 41.887 40 y = 0.0012x + 29.935 30 R² = 0.1875 30 R² = 0.1215 20 20 10 10 0 0 0 5 10 15 20 25 30 0 5000 10000 15000 Source: For the annual gDP per capita growth rate and the female labor force participation rate: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators/. Note: Chart a: the fitted line = 0.4911(gDP per capita growth) + 41.887; r² = 0.1875. Chart b: the fitted line (female wages as a percentage of male wages) = −0.0012(gDP per capita) + 29.932; r² = 0.1204 (for 2008). gDP = gross domestic product. example, childcare and household production) Description of Labor Markets implies that parental leave policies and flexible work arrangements have a much stronger im- Labor markets in Europe and Central Asia have pact on female labor market participation than three main characteristics.24 First, the diversity on male labor market participation. Discrimina- of the region means there is large variation across tion and segregation have the potential to explain some of these outcomes in the labor market. So- 23 See Altonji and Blank (1999) for a survey on discrimi- cial policies that facilitate women’s greater labor nation and segregation. See Holzer and Neumark force participation must therefore target the spe- (2000) for a survey on affirmative action policies. cific problem, and they must also necessarily be 24 Labor force participation and the economically active population include all residents who are employed or gender specific.23 unemployed. The employed are all persons above a This chapter has two main objectives. First, specified age who, during a specified brief period, either it aims to provide an overview of how men and one week or one day, were in the following categories: women have performed in labor markets in the (a) paid employment, that is, (i) at work: persons who, region so as to identify the gender gaps that are during the reference period, performed some work for larger and more persistent. Given the diversity of wage or salary, in cash or in kind; or (ii) with a job, but not at work: persons who, having already worked in the region, gender differences vary significantly their present jobs, were temporarily not at work dur- across countries and subregions. The second ob- ing the reference period, but had a formal attachment jective is to summarize the factors behind gender to their jobs; (b) self-employment, that is, (i) at work: gaps by reviewing analytical work available on the persons who, during the reference period, performed region. As a byproduct, knowledge gaps will be some work for profit or family gain, in cash or in kind; identified. or (ii) with an enterprise, but not at work: persons with an enterprise, which may be a business enterprise, a The chapter is organized into two main sec- farm, or a service undertaking, who were temporarily tions. First, we present an aggregate description of not at work during the reference period for any spe- labor markets in the region over the last decade. cific reason. The unemployed are all persons above a We provide a description of the current status of specific age who, during the reference period, were (a) men and women and identify those countries in without work, that is, were not in paid employment or which gender gaps are more persistent. Second, we self-employment; (b) currently available for work, that is, were available for paid employment or self-employ- describe the factors that contribute to these gen- ment during the reference period; and (c) seeking work, der gaps. that is, had taken specific steps in a specified reference period to seek paid employment or self-employment. 32 Opportunities for Men and Women: Emerging Europe and Central Asia countries. Second, these countries exhibit one sons in the region, or 75 percent of the male popu- of the highest levels of occupational segregation, lation between 15 and 64 years of age in 2008. meaning that men and women pursue separate oc- Over half of women labor force participants in cupational streams. Third, with a few exceptions, the region (57 percent) live in the populous and middle- most of the countries still show high female labor income countries of the Commonwealth of Indepen- force participation rates that are comparable with dent States (CIS). The subregion with the second the rates in developed countries; meanwhile, male largest number of women in the labor force is the participation rates are well below the world aver- EU10, with 21 percent of the total. The four coun- age. Hereafter, we present a more detailed descrip- tries with the largest share of the region’s women la- tion of labor force participation rates, employment bor force participants are Poland, the Russian Fed- and unemployment rates, occupational segrega- eration, Turkey, and Ukraine. These four countries tion, and labor market transitions. also have the largest share of the region’s male labor force; Turkey is second to Russia in terms of share (16 percent of the region’s male labor force). The Participation Rates low-income countries of the CIS—the Kyrgyz Re- Women make up only slightly less than half the labor public and Tajikistan (in 2009)—and the western force in the region. The latest data for the region show Balkans have a relatively small share of the region’s that about 97 million women participated in the la- total female and male labor force participants, ac- bor market in 2009. Though 60 percent of women counting for around 2 percent. aged 15 to 64 are in the labor force and have been The region’s female labor force participation rates since 1999, there was an increase of about 6.0 mil- are near the global average, but the male participation lion women in the labor force between 1999 and rates are well below the global average. The female la- 2008 because of population growth (figure 2.2). bor force participation rate among 15- to 64-year- Women constitute 45 percent of the aggregate olds in Europe and Central Asia, at 59 percent, is labor force in the region, and this proportion has slightly above the global average of 57 percent and also remained constant over the past decade. Men’s the average in Latin America and the Caribbean of labor force participation stands at 117 million per- 55 percent, but well below the rate in East Asia of Figure 2.2 Labor Force Participation Rates a. By gender, 2009 b. Change, 1999–2009 90 6 80 5 70 4 60 3 2 50 1 40 0 30 –1 20 –2 10 –3 0 –4 ECA EAP LAC MENA SAS SSA High ECA High LAC MENA SSA EAP SAS income income OECD OECD Female Male Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eAP = east Asia and the Pacific. eCA = europe and Central Asia. LAC = Latin America and the Caribbean. MeNA = Middle east and North Africa. SAS = South Asia. SSA = Sub-Saharan Africa. Labor force participation rates are for the population 15 to 64 years of age. Women in the Labor Market 33 70 percent. The region’s male participation rate, 74 a small increase of 1 percentage point in women’s percent, is significantly lower than the correspond- labor force participation, while East Asia and the ing rate in any other region, and it is also lower than Pacific—the region with the highest female par- the global average. Yet, compared with the high-in- ticipation rate—saw a decline from 67 to 64 per- come countries of the Organisation for Economic cent. Male labor force participation rates decreased Co-operation and Development (OECD), both fe- across the board: the global average decreased by 2 male and male participation rates in the region are percentage points. significantly lower, by 6 percentage points. This also In some countries in Europe and Central Asia, highlights that there is no standard labor force par- the labor force participation rates of men and women ticipation rate for men and women. Concern only have changed dramatically over the last decade. As arises if labor force participation rates do not ade- noted above, the average female labor force partici- quately capture the population interested in work- pation rate remained unchanged in the region dur- ing because people faced with a long-term inability ing 1999–2009. There were changes at the sub- to find jobs have dropped out of the search for work. regional level. Over the period, the female labor The country-level labor force participation rates force participation rates increased by 6 percent- of men and women vary greatly across the region. age points in the low-income CIS countries, ris- Women’s participation rates for ages 15 to 64 years ing from 54 to 60 percent, while the rate fell by 6 ranged from a low of 26 percent in Turkey to a high percentage points in Turkey. The western Balkans of 73 percent among working-age women in Ka- and the middle-income CIS countries saw a slight zakhstan in 2009. If we take the (unweighted) aver- increase, of 2 percentage points, while the EU10 age of all the female labor force participation rates saw no change over the decade. The average male by country, the average for the region is 61 percent. participation rate increased by 1 percentage point Male participation rates—though, on average, sig- during this period, while the changes by subregion nificantly higher than female participation rates— were similar to the changes in the female partici- also vary greatly across the region, from a low of 57 pation rates, though of different magnitudes. The percent in Moldova to a high of 83 percent in the countries that were outliers in terms of the largest Kyrgyz Republic. The gender gap in participation, decreases in male and female labor force participa- which averages 14 percentage points (indicating tion rates are Moldova and Romania, with declines that women’s participation rates are much lower ranging from 11 to 10 percentage points. At the than men’s), ranges from 48 percentage points in other extreme, Tajikistan experienced an increase Turkey to 4 percentage points in Moldova. The of 15 and 13 percentage points, respectively, in fe- size of the gender gap, on average, is in line with the male and male labor force participation rates. female labor force participation rates as indicated Over the last decade, the gender gap in labor by the strong negative correlation (0.82) between force participation rates in the region has remained the female labor force participation rate and the unchanged, though this has not always reflected high- gender employment gap. er women’s participation. The gender gap measures Over the last decade, the region’s averages have the difference in the male and female labor force remained largely stagnant, similar to global trends in participation rates. The average (unweighted) dif- male and female labor force participation. Global ference remained unchanged at 14 percentage averages indicate that women’s participation rates points over 1999–2009. By subregion, the gender remained stable, while male rates fell by 1 per- gap is the lowest among the middle-income CIS centage point in 1999–2009. The variation in the countries, at 10, and most of the rest of the region changes across regions is high (figure 2.2, chart b). is in the range of 12 to 22, except Turkey, where For example, in Latin America, a region with many the gender gap is 48. The gender gap in Western middle-income countries, there was a surge in fe- Europe fell, from 15 to 11; this has been driven by male labor force participation rates by 5 percentage large declines in the gender gap in Cyprus, Luxem- points. High-income OECD countries experienced bourg, the Netherlands, and Spain. 34 Opportunities for Men and Women: Emerging Europe and Central Asia With the exception of Turkey, the labor force in Europe and Central Asia is in line with the aver- participation rates of men and women show a slight age in Western Europe and in high-income OECD convergence in most subregions of Europe and Central countries. However, the trend in the gap between Asia over time. Figure 2.3 reports the labor force par- men and women over time is not converging, as ticipation rates among men and women in 2009 for observed in Western Europe. The gap was stable each subregion and the change in the rates over the across all subregions in Europe and Central Asia decade 1999–2009. An important difference with from 1999 to 2009, while, in Western Europe, there respect to Western Europe is apparent: in Western was a convergence of about 4 percentage points. Europe, the gender gap in participation decreased Since this occurred during a decade of high growth over the decade, continuing a process of conver- in Europe and Central Asia, the trend is troubling. gence that started in the 1960s. The convergence is Young women in Europe and Central Asia par- driven by an increase in female participation rates ticipate in the labor market less than young women in and stable male participation rates. CIS countries developed countries. The gender gap in participation have shown slight convergence: the gender gap de- varies with age. Women are less likely to work than creased from 11.0 to 9.6 percent. However, the mod- men during childbearing years. In the countries of est convergence in the CIS subregion is not driven Europe and Central Asia, women have children at by greater female participation, but by less male par- a relatively young age compared with women in de- ticipation. Some subregions, such as the low-income veloped countries. The gender gap in participation CIS countries, showed a slight increase in the gender is greater in Europe and Central Asia than in the gap over the decade, but this was largely caused by developed countries of Europe and North America changes occurring in the Kyrgyz Republic. among women between 20 and 35 years of age (fig- The gender gap in labor force participation is ure 2.4). Although the gender gap in participation small compared with the gap in other regions, but the in the region decreases as women become older, trend in the gap is troubling. A lower participation this is partly caused by a decrease in male partici- rate among women relative to men is common in pation rates rather than a significant increase in fe- almost all labor markets. Thus, the level of the gap male participation. Figure 2.3 Labor Force Participation Rates by Subregion a. By gender, 2009 b. Change, 1999–2009 percent percentage points 90 10 80 8 70 6 60 4 50 2 40 0 30 –2 20 –4 10 –6 0 –8 EU10 Western CIS LIC CIS MIC Turkey EU10 Western CIS LIC CIS MIC Turkey Balkans Balkans Female Male Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: CiS LiC = low-income CiS countries (Kyrgyz republic and Tajikistan) in 2009. CiS MiC = middle-income CiS countries (Armenia, Azer- baijan, Belarus, georgia, Kazakhstan, Moldova, the russian Federation, Turkmenistan, ukraine, and uzbekistan) in 2009. The gender gap is the difference between the male and female labor force participation rates. Women in the Labor Market 35 Figure 2.4 The Gender Gap in Labor Force Participation by Age, 2009 25.0 20.0 Percentage points 15.0 10.0 5.0 0.0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70+ ECA Selected Developed Countries Source: World Bank staff calculations based on data of the uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: eCA = europe and Central Asia. Although women 60 to 64 years of age are more to the gap observed in the labor force. The average likely to work now than such women were 20 years employment rate among women was 45 percent of ago, the increase in the participation rate among this all working-age women, which, though comparable group has been low compared with the corresponding with employment rates in Western Europe, is well group in the OECD countries and compared with men below the Lisbon objectives of 60 percent.25 of the same age. Women retire at a younger age rela- Employment growth was relatively weak in the tive to men in the region and women in the OECD region even during the robust growth period of 1999– countries. Figure 2.5 shows the labor force par- 2008. The number of employed persons in 2008 ticipation rates for men and women 60 years old was 198 million, or 51 percent of the working-age and over. The growth in the women’s rate has been population aged 15 years or older. Employment much greater in the OECD countries. However, grew, on average, by 1 percent per year in 1999– women 70 or older are more likely to participate in 2008, a period when average economic growth was the labor market in Europe and Central Asia than high. The region’s economy grew by 6 percent a in the OECD countries. This may be an indication year during the period, compared with the world of the vulnerability of this population group. average of 3 percent. However, in the region, em- ployment creation was not particularly responsive to growth: for every 1 percent of growth in gross Employment domestic product (GDP), employment increased Women constitute slightly less than half the employed in by 0.10 percentage points in the region, compared the region. In 2009, an estimated 91 million women with the world average of 0.52 percentage points. 15 years and older in the region participated in the The situation in 2009 was significantly worse in labor force; the largest numbers were in Russia and Europe and Central Asia than in the rest of the Ukraine, where women accounted for 50 million world because of the global financial crisis. among the employed. Women comprised 46 percent of the total number of persons employed. The aver- age gender gap in employment was 14 percentage 25 In Lisbon in March 2000, the heads of EU govern- points; the largest gap was in Turkey (39 percentage ments subscribed to the goal of a general increase in points), and the narrowest in Belarus (0 percentage the employment rate. For women, they agreed upon a points). The gender gap in employment is similar specific target of 60 percent by 2010. 36 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 2.5 Labor Force Participation among Men and Women, 1980–2009 FLFP 35 30 25 In percentage 20 15 10 5 0 1980 1990 1995 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 ECA Women 60–64 ECA Women 65–69 ECA Women 70+ OECD – Women 60–64 OECD – Women 65–69 OECD – Women 70+ MLFP 60 50 40 In percentage 30 20 10 0 1980 1990 1995 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 ECA Men 60–64 ECA Men 65–69 ECA Men 70+ OECD – Men 60–64 OECD – Men 65–69 OECD – Men 70+ Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: eCA = europe and Central Asia. Women’s employment growth has been stronger tion are likely to be factors unique to each country. than men’s, though by only a small amount. During Another important aspect of the changes in em- 1999–2008, women’s total employment increased ployment among men and women is that they are by 11 percent compared with 9 percent among men. strongly negatively correlated (with the exception The countries with the highest growth rates in fe- of Turkey): female employment has been declining, male employment are also among the poorest: Ta- while male employment has been rising. Moreover, jikistan and Uzbekistan, by an average 4.8 and 3.4 the gender gap has remained relatively stable over percent a year, respectively. These countries have the decade, except in Moldova and Tajikistan. also seen the highest growth in male employment. The employment rate is lower among women Five countries experienced negative female employ- with more children and women with children under ment growth rates: Georgia, Lithuania, Moldova, 3 years of age. Women with children are less likely Romania, and Turkey. These same countries, less to work in the region relative to OECD countries, Turkey, but plus Croatia, also experienced negative and this difference increases with the number of growth rates in male employment. The diversity children (figure 2.6). While the employment rates of the subregions to which these countries belong among women with no children are 71 and 78 per- indicates that the drivers of employment contrac- cent, on average, in Europe and Central Asia and in Women in the Labor Market 37 Figure 2.6 Female Employment Rate by Number of Children and the Age of the Youngest Child, 2008 a. By number of children b. By age of youngest child 100 100 78 82 75 76 80 71 75 72 80 Percentage Percentage 69 64 69 67 67 62 56 57 55 60 60 42 40 40 31 20 20 0 0 0 1 2 3 or more No under 3 3–5 6–16 17 and children above ECA OECD Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: The averages for no children vary for the two groups because the countries included in each case are different. in chart a, europe and Central Asia (eCA) includes Bosnia and Herzegovina, Bulgaria, Croatia, the Czech republic, estonia, Hungary, Lithuania, Moldova, Poland, ro- mania, Serbia, and Slovenia; OeCD includes Austria, Belgium, Canada, Denmark, Finland, France, germany, greece, iceland, ireland, israel, italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the united Kingdom, and the united States. in chart b, europe and Central Asia includes Bosnia and Herzegovina, Croatia, the Czech republic, estonia, Hungary, Lithuania, Poland, romania, and Serbia; OeCD includes Austria, Belgium, Bosnia and Herzegovina, Canada, Croatia, Cyprus, Czech republic, Denmark, estonia, Finland, France, germany, greece, Hungary, ireland, israel, italy, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Portugal, romania, Serbia, Spain, Sweden, Switzerland, the united Kingdom, and the united States. The rates refer to women 25 to 49 years of age. the OECD, respectively (a 7 percentage point gap), outliers such as the former Yugoslav Republic of the respective employment rates among women Macedonia, where unemployment affects almost with 3 or more children are 42 and 57 percent, or a one-third of the labor force. The EU10 saw a mini- 15 percentage point gap between the regions. The mal decline in unemployment rates. Meanwhile, group affected the most is women whose young- though there was a positive trend in unemployment est children are under 3 years of age. In Europe and rates in Europe and Central Asia, this indicator Central Asia, the employment rate is only 31 per- may underestimate the number of people who are cent in this group, while, in OECD countries, it is without work, but who would like to work; this is 64 percent (a 33 percentage point gap). because unemployment rates sometimes fail to cap- ture discouraged workers who have given up look- ing for jobs after long and unsuccessful job searches. The Unemployment Rate The gender gap in unemployment rates in the re- Unemployment fell significantly in the region over the gion is small. In 2009, the average female and male decade. The average unemployment rate in the re- unemployment rates were 12.7 and 13.0 percent, gion decreased from 13 to 11 percent during 1999– respectively.27 The average gender gap in the re- 2009.26 However, this trend hides variations across gion was less than 1 percent. The subregion with subregions. For example, the unemployment rates in the largest gender gap was the western Balkans, the middle-income CIS countries declined over the where the gap was especially severe in Kosovo, at decade, reaching around 7.5 percent in 2009 from 13.0 percent in 1999. However, the unemployment 26 This average is based on 18 countries. There are no rate remained constant at around 11 percentage over data for Central Asia or the low-income CIS countries. the decade. Meanwhile, Turkey’s unemployment 27 This is based on the latest available data for 21 coun- rate rose from 7 to 14 percent. The western Balkans tries; it excludes Albania, Armenia, Belarus, Kazakhstan, was a high unemployment subregion and includes Montenegro, Tajikistan, Turkmenistan, and Uzbekistan. 38 Opportunities for Men and Women: Emerging Europe and Central Asia 16 percentage points. The second largest gap was may reflect the fact that persons with a primary edu- in Lithuania, at 7 percentage points. The relatively cation are more likely to be poor and unable to af- low unemployment gap in the region is comparable ford long job searches, while tertiary graduates are with the gap in Western Europe and unlike the gaps likely to be in high demand. There are two striking in many other developing regions, such as Latin differences among countries within the region. The America and the Middle East and North Africa. middle-income CIS countries show a pattern that The gender gap in unemployment in the EU10 in- diverges from the pattern in the rest of the region: creased, especially in 2009. Because of a lack of data, more well educated workers constitute a much larg- it is difficult to determine whether unemployment er proportion of the unemployed than less well edu- rates among men and women are converging. The cated workers (33 percent in 2007). Second, in the only subregion with data over the entire decade is EU10, though persons with tertiary education con- the EU10. The average unemployment rate in the stitute a small share of the unemployed, this share EU10 consistently fell over the decade, and the rose rapidly from 6 to 10 percent in 1999–2008. gender gap narrowed until 2008. In 2009, the gap The gender gap in unemployment rates by educa- widened abruptly as male unemployment doubled. tional attainment is relatively small, except in Turkey. In the late 1990s, the female unemployment rate The gender gaps in primary, secondary, and tertiary was lower than the male unemployment rate, but, unemployment rates were −5, −1, and −4 percent- after 2002, the female unemployment rate grew to age points, respectively (figure 2.7). (The negative about a half percentage point higher. Compared sign indicates that female unemployment rates ex- with the Western Europe average, this gender gap ceed male unemployment rates.) There is a gender remains small in both absolute and relative terms. gap across subregions, but it is small. For example, Though youth unemployment is high in the region the largest gender gap occurs in the EU10 coun- and the gender gap in the rates is almost 2 percent, tries among persons with only primary education youth unemployment remains relatively modest by (−6 percentage points). Meanwhile, Turkey’s gen- international standards. In general, the unemploy- der gap is −15 percentage points among secondary ment rates among men and women 15–24 years and tertiary school graduates, compared with −27 old are higher than the overall unemployment rate among primary school graduates, indicating that, in in a country. Young labor force participants may Turkey, it is more difficult for women to find em- have high expectations of finding the best jobs in ployment than it is for men. terms of pay and occupation and, consequently, Unemployment rates are slightly higher among must undertake longer job searches until they find young women than among young men, particularly in jobs or change their expectations about the nature countries with high youth unemployment. Although of acceptable jobs. Europe and Central Asia is no there is no appreciable difference in the averages for different. The average youth unemployment rate in the region in the unemployment rates among young 2009 was 25 percent. Though time series data are men and women, this is the result of large variations unavailable for the decade, youth unemployment in unemployment rates across countries and gen- has fallen slightly over the last few years. The fe- der. In figure 2.8, panel a, we see that, with the clear male and male youth unemployment rates in 2009 exception of one country (Azerbaijan), the unem- were 26 and 25 percent, respectively. Countries ployment rates in most countries are higher among with relatively high unemployment rates also show young women than among young men. Figure 2.8 high unemployment rates among young workers. also shows that the gender difference in unemploy- Unemployment rates according to educational at- ment rates among young men and women increases tainment varies greatly across the region. The unem- with the rate of youth unemployment (the points to ployment rate among secondary school graduates in the right are more likely to be farther away from the the region is 57 percent compared with 28 and 17 45 degree line). For example, in Armenia, the unem- percent for individuals with primary school and ter- ployment rate among young women is 69 percent, tiary school degrees, respectively. These differences while, among young men, it is 47 percent. Women in the Labor Market 39 Figure 2.7 Gender Gaps in Unemployment a. Unemployment rates, 2008 b. Gender gap by education, 2007 share of labor force by gender, % percentage points 30 30 25 20 20 10 15 0 10 5 –10 0 –20 EU10 Kyrgyz CIS MIC Turkey Western EU CIS LIC CIS MIC Western Turkey Republic Balkans Balkans Female Male Primary Secondary Tertiary Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: Data are not available for all countries. CiS LiC = low-income CiS countries (Kyrgyz republic and Tajikistan) in 2009. CiS MiC = middle- income CiS countries. The average for CiS MiC includes Azerbaijan, Belarus, georgia, russian Federation, and ukraine. The gender gap is the difference between male and female unemployment rates. Figure 2.8 Youth Unemployment and Long-Term Unemployment, by Gender, 2008 a. Youth unemployment rates b. Long-term unemployment rates percent percent 80 100 70 80 Female Long-term 60 Unemployment unemployment Female youth 50 60 40 30 40 20 20 10 0 0 0 20 40 60 80 0 50 100 Male youth unemployment Male Long-term Unemployment Source: World Bank staff calculations based on data of the World Development indicators Database, World Bank, Washington, DC, http://data. worldbank.org/data-catalog/world-development-indicators/. Note: The young unemployed are all the persons aged 15–24 years who, during the reference period, were (a) without work, that is, were not in paid employment or self-employment; (b) currently available for work, that is, were available for paid employment or self-employment during the reference period; and (c) seeking work, that is, had taken specific steps in a specified reference period to seek paid employment or self- employment. The long-term unemployed are the persons who have been unemployed for 12 months or more. The long-term unemployment rate is the share of the long-term unemployed in the total unemployed population by gender. Long-term unemployment rates are relatively 28 Long-term unemployment refers to the number of similar between men and women.28 There is al- people who have experienced periods of unemploy- most no gender difference in the long-term un- ment extending for a year or longer; it is expressed as a employment rates in the region, as indicated in percentage of the total unemployed. 40 Opportunities for Men and Women: Emerging Europe and Central Asia figure 2.8, panel b, where most of the points are In Europe and Central Asia, self-employment is on the 45 degree line (no gender differential). more widespread in countries with large rural popula- However, if women are more likely than men to tions. In countries in which agriculture is an impor- leave the labor force if they are unemployed, the tant contributor to the economy, a larger share of absence of gender gaps in long-term unemploy- the employed are self-employed (figure 2.9, panel ment rates may not accurately reflect the differ- a). The correlation was quite strong even in 2009, ences in the success of job searches by men and when it was 0.47.29 This relationship between the women. This is supported by the evidence pre- share of agriculture value added and self-employ- sented elsewhere below in the section on labor ment rates increased over time until 2005, then de- market transitions, where we show that women creased, among the 21 countries in the region on are more likely than men to leave the unemploy- which data are available. Thus, it is possible that, ment category by exiting the labor force. over time, self-employment became more closely associated with agriculture rather than with the development of small entrepreneurship in higher- Self-employment productivity sectors. Self-employment is a diverse category that changes over In the region, one in four employed persons is self- time. Self-employment is a useful concept because employed.30 Average self-employment rates are high in of its links to entrepreneurial ability and informal the middle-income CIS countries, but relatively low activity in the labor market. Self-employment in- in the EU10. In 2009, the subregion with the high- volves a wide range of careers, including, for exam- ple, physicians, plumbers, electricians, and farmers. Additional data would be necessary to determine 29 To a lesser extent, there is also a positive relationship the extent of self-employment in high- and low- between self-employment and the share of a country’s productivity jobs or in particular sectors. However, population living in rural areas. 30 This assertion is based on 21 countries. The coun- self-employment is frequently associated with un- tries not in the sample are Albania, Armenia, Belarus, deremployment and the informal sector, especially Montenegro, Serbia, Tajikistan, Turkmenistan, and in low- and middle-income countries. Uzbekistan. Figure 2.9 Self-Employment a. Economies with large rural populations b. By gender percent percent of employed, by gender 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 0 20 40 60 80 EU10 Kyrgyz CIS MIC Western Turkey Republic Balkans Female Male Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: Data are available for a subset of the countries in europe and Central Asia. The countries excluded are Albania, Armenia, Belarus, Kazakh- stan, Montenegro, Tajikistan, Turkmenistan, and uzbekistan. MiC = middle-income countries. Women in the Labor Market 41 est self-employment rate was the middle-income CIS Turkey is an exception. It shows a high, but de- countries, at 31 percent, whereas the EU10 showed creasing level of male and female self-employment. the lowest rate, which, at 16 percent, is comparable Among the countries on which data are available, with the rate in Western Europe. The western Bal- Georgia and Ukraine showed the greatest change in kans falls in the middle. The country with the highest self-employment over the last decade: an increase rate of self-employment in recent years is Georgia, of about 7 and 10 percentage points, respectively. at 64 percent, followed by three countries, Armenia, Azerbaijan, and Turkey, at 39, 58, and 40 percent, respectively. Russia showed the lowest share of self- Sectors of Employment employed in total employed, at 7 percent. In Europe and Central Asia, the service sector em- More men than women are self-employed in Eu- ploys more people than the agricultural sector and rope and Central Asia. On average in the region, the industrial sector combined. The service sector an estimated 26 percent of employed women are employs, on average, 57 percent of all employed, self-employed, compared with 28 percent of men while agriculture and industry employ 16 and 27 (figure 2.9, panel b). Thus, the average difference percent, respectively.33 At the aggregate level, other between male and female rates of self-employment than in the low-income CIS subregion, service sec- is negligible. Though all subregions exhibit this pat- tor jobs dominate in employment in all subregions. tern, the gap in self-employment rates among men Yet, there is significant variation across countries and women is narrowest in the middle-income CIS and subregions, especially in the employment rates countries, where the average difference is 2 percent- in agriculture. For example, over 50 percent of the age points. However, this difference hides extremes. total employed are active in the agricultural sector For example, the two countries in which female self- in middle-income countries such as Albania and employment rates significantly exceed male self- Georgia, while 10 percent of the employed in Rus- employment rates are Azerbaijan and Turkey, with sia are active in this sector. differences of 11 to 12 percentage points, respec- The service sector employs the majority of women tively. In the EU10 subregion on average, female in most countries in the region. With the exception self-employment rates are moderate, at 11 percent, of four countries (Armenia, Georgia, Romania, which is almost half the average rates among men and Turkey), over half of women workers are em- (of 17 percent), while, in Western Europe, female ployed in the service sector. In the EU10 and the self-employment rates are low, at 9 percent, which western Balkans, the average is about two-thirds is almost half the average rates among men. of all employed women. This structure of fe- In Europe and Central Asia, male and female male employment conforms to the pattern seen self-employment rates have been stable over the last in OECD economies, that is, women tend to be decade. On average, self-employment has remained concentrated in the service sector (figure 2.10). In constant among men and women over the past de- the countries in which the highest proportion of cade.31 However, there has been variation across female employment is not in services, women are subregions, as well as significant variation across countries. For example, the EU10 has seen a slight 31 The average self-employment rate in the 17 countries decline in self-employment rates among men and for which longitudinal data are available was 16.5 per- women that is similar to developments in Western cent in 1999 and 17.1 percent in 2007. Europe, though with a slightly bigger magnitude. 32 There is a complete time series for only Moldova, Rus- Changes in aggregate self-employment rates over sia, and Ukraine over the period. Note that, within time do not appear to be strongly correlated to the middle-income CIS subregion, the countries show relatively lower proportions of self-employment. changes in GDP; thus, countries with higher growth 33 The averages are based on 25 countries in the region; rates have not seen a decrease in self-employment. because of a lack of recent data, they exclude Belarus, However, the subregions with high proportions of Bosnia and Herzegovina, Turkmenistan, and Uzbeki- self-employed have seen an increase over time.32 stan. 42 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 2.10 Sectoral Employment, by Gender a. Sectoral employment b. Gender gap in sectoral employment percent of employment by gender percentage points 70 30 60 20 50 10 40 0 30 –10 20 –20 10 –30 0 EU10 Kyrgyz CIS MIC Turkey Western Agriculture Industry Services Republic Balkans Female Male Agriculture Industry Services Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: The gender gap is the difference between male employment in a particular sector as a share of total male employment and female em- ployment in a particular sector as a share of total female employment. A negative gender gap means that more women than men are employed in the sector as a share of total employment among women. CiS MiC = middle-income CiS countries (Armenia, Azerbaijan, Belarus, georgia, Kazakhstan, Moldova, the russian Federation, Turkmenistan, and ukraine) in 2008. disproportionately present in agriculture (rather relative contribution to GDP. Women have moved than in industry). Though male employment is less into the service sector in slightly higher proportion concentrated in services relative to female employ- than men in all the countries of Europe and Central ment, the share is still high, at 46 percent versus 35 Asia, a trend not found in Western Europe. If the percent in industry and 16 percent in agriculture in trend continues, this sectoral allocation should be the most recent years. beneficial for the reduction of gender gaps in the Though agriculture and industry are not the labor markets of the region (see elsewhere below). largest employers of women on average, women still The customary gender segregation by industry is constitute a significant share of the employees in these also found in the region. Table 2.1 shows the distri- sectors. In the 22 countries on which data are avail- bution of employed men and women across indus- able, women employees make up half of all workers tries. In Europe and Central Asia, men and women in the service sector. However, they also comprise are likely to work in distinct industries. While men about 17 percent of workers in agriculture and 17 are more likely to be employed in manufacturing, percent of workers in industry. The variation across construction, and transport, women are more likely countries in the averages is relatively modest. How- to be found in communal services, wholesale and ever, Turkey remains an outlier, with about 38 and retail, and restaurants. For example, while 34 per- 15 percent of women in agriculture and industry, cent of employed men work in manufacturing, only respectively. 22 percent of employed women do so; in contrast, Over time, the share of women working in the ser- 40 percent of employed women and only 20 per- vice sector is growing. If services represent a sector of cent of employed men work in communal services. traditional employment among women in all OECD The region is characterized by high occupation- economies, the trend over time is more specific to al segregation in labor markets. Table 2.1 presents Europe and Central Asia. The service sector has the distribution of employed men and women been expanding over the decade, experiencing not across occupations. Occupational segregation only growth in employment, but also growth in the in the labor market is significant in Europe and Women in the Labor Market 43 Central Asia relative to other regions.34 Women the region is doing as well as, if not better than devel- are concentrated in the category of profession- oped countries, on average. As in developed coun- als and technicians and in services, while men tries, there are fewer women in management posi- are mainly concentrated among administrative tions in Europe and Central Asia. The country with personnel and machine operators.35 Almost 40 the most equitable access to management positions percent of women, but only 22 percent of men in terms of gender is Serbia, where the gap between work as professionals or technicians, whereas 18 men and women in management is only 1 percent- percent of men and 6 percent of women work as age point (figure 2.11). Yet, few men and women in machine operators. (Box 2.1 offers the example of Serbia hold management positions: only 4 percent the teaching profession in primary and secondary of employed men, for example. At the opposite ex- education.) treme, a larger gap is observed in Estonia and Tur- Although there is some variation across the re- key, where employed men hold 16 and 11 percent gion in women’s employment in management positions, of management positions, respectively, while only 8 and 3 percent of employed women hold such posi- tions, respectively. Table 2.1 | The Sectoral Structure of Occupational segregation begins before entry Employment in the labor market. Occupational segregation at Occupation and sector Men Women the workplace is a consequence of the choice of Occupation the field of study when one is young. A recent Professionals and technicians 22 38 study shows that, although the correlation be- tween the field of study and occupation does Directors and upper management 6 4 not appear close if one uses broad occupational Administrative personnel and intermediary 31 20 level definitions given that most graduates work as Service workers 9 17 professionals or technicians, the field of study is Skilled agriculture 2 1 a strong determinant of occupation within the Machine operator 18 6 broad categories. For example, in the Czech Re- public, while 50 percent of men who have stud- Armed forces 0.0 0 ied science pursue occupations as professionals Elementary occupations 12 14 or technicians in physics, mathematics, or engi- Total 100 100 neering, this is true of only 20 percent of women Economic sector who have studied science (Flabbi 2011). Agriculture, hunting, forestry, and fishing 8 5 Mining and quarrying — — Manufacturing 34 21 Electricity, gas, and water supply .. .. 34 For more details, see Ñopo, Daza, and Ramos (2011), Construction 12 1 which is a background paper for World Development Wholesale and retail, trade, and hotels and 12 20 Report 2012: Gender Equality and Development (World restaurants Bank 2011a). The authors analyze the segregation in Transport and storage 10 4 labor markets in Europe and Central Asia and other Finance and business services 2 4 regions. 35 Professionals include occupations in which the main Communal services 20 41 tasks require a high level of professional knowledge Other services not well specified 2 3 and experience in the physical and life sciences, the Total 100 100 social sciences, or the humanities. Technicians include technicians and associate technicians, that is, occu- Source: Ñopo, Daza, and ramos (2011). pations in which the main tasks require technical Note: Weighted average for the region, including Albania, Bosnia and Herze- govina, Bulgaria, Croatia, the Czech republic, estonia, Hungary, the Kyrgyz knowledge and experience in one or more areas of the republic, Latvia, Lithuania, Moldova, Montenegro, Poland, romania, the rus- physical and life sciences, the social sciences, or the sian Federation, the Slovak republic, Tajikistan, and Turkey… = negligible. humanities. 44 Opportunities for Men and Women: Emerging Europe and Central Asia Box 2.1: Most Teachers are Women Women continue to work in traditional female occupations such as teaching. Most teachers in primary and secondary schools are women. In Europe and Central Asia, 91 percent of primary teachers and 72 percent of secondary teachers are women. In a few countries, such as Armenia, all primary teachers are women (table a). In Belarus, the Czech Republic, Kazakhstan, the Kyrgyz Republic, Russia, Slovenia, and Ukraine, 98 percent or more of primary teachers are women. In most of these countries, women also predominate in instruction in second- ary education, accounting for more than 80 percent of secondary teachers. Table a | Share of women teachers, 2008 Country Primary Secondary Tertiary Albania 70 61 48 Armenia 100 84 — Azerbaijan 87 66 52 Belarus 99 80 57 Bulgaria 93 79 47 Croatia 91 68 42 Czech Republic 98 66 48 Estonia 94 78 — Georgia 86 82 53 Hungary 96 71 38 Kazakhstan 98 86 64 Kyrgyz Republic 98 74 59 Latvia 93 81 57 Lithuania 97 82 56 Macedonia, FYR 72 54 44 Moldova 97 76 56 Montenegro 70 61 40 Poland 84 69 43 Romania 86 67 43 Russian Federation 98 81 61 Serbia 84 64 45 Slovak Republic 89 74 44 Slovenia 98 72 37 Tajikistan 68 49 33 Turkey 50 41 40 Ukraine 99 79 — Uzbekistan 85 63 38 Sources: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators/; uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: — = not available. Segregation also exists by type of employer; to be employed in the public sector. For example, women tend to work in the public sector. Although in Serbia, while 32 percent of employed women we do not have the average for the region, several work in the public sector, only 24 percent of men country studies find that women are more likely work in this sector. In contrast to women, men Women in the Labor Market 45 Figure 2.11 The Gender Gap in Management percent 8 7 Gap in percentage of managerial positions 6 5 4 3 2 1 0 Turkey Montenegro Latvia Bulgaria Armenia Russian Federation Moldova Ukraine Kazakhstan Macedonia, FYR Georgia Belarus Kyrgyz Republic Serbia ECA Selected Developed Countries Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: The developed countries included are Austria, Canada, Denmark, Finland, France, germany, greece, iceland, ireland, israel, italy, Luxem- bourg, Malta, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the united Kingdom, and the united States. eCA = europe and Central Asia. are more likely to work in the private sector (58 In the region, women spend less time than men percent of employed men versus 52 percent of in labor market activities. As in the developed employed women). Similar patterns of concen- countries, women in the region spend, on av- tration are found in Bulgaria, Russia, and Serbia erage, almost two hours less per day than men (see Dimova, Gang, and Landon-Lane 2006; in activities related to the labor market (figure Oglobin 2005; Reva and Sulla 2011). 2.13, panel a). The variation across the coun- tries in the region on which data are available is small. The largest difference is in Turkey, where Time use Analysis men spend 3.5 more hours than women in labor Women are more likely than men to be part-time work- market activities, and the smallest difference is ers; however, part-time employment is less prevalent in in Bulgaria, where men spend only 1 more hour Europe and Central Asia than in the OECD countries. than women per day performing labor market On average in Europe and Central Asia, 12 percent activities. of employed women and 7 percent of employed men Women in the region spend much more time work part time (figure 2.12). The availability of part- than men on household activities. In the region, time work in general and, in particular, for women is relative to men, women spend three more hours considerably more limited in the region than in the per day on household work (figure 2.13, panel b). OECD countries (excluding the United States). In The average time women in the region and wom- the OECD countries, 36 percent of employed wom- en in the OECD countries spend on household en and 11 percent of employed men work part time. chores is similar. The gender differential is largest The lack of access to part-time employment, espe- in Turkey, where women spend 4.5 hours more cially among women and relative to OECD levels, than men on household chores. It is smallest in may have implications for decisions on labor force Latvia, where women spend only two hours more participation and for fertility rates among women. than men on such chores. 46 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 2.12 Part-Time Workers, by Gender percent 80 70 60 50 40 30 20 10 0 Azerbaijan Bosnia and Herzegovina Bulgaria Croatia Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovak Republic Slovenia Turkey Albania Armenia Kazakhstan Moldova Montenegro Russian Federation Serbia Macedonia, FYR Female Male Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Figure 2.13 The Difference in Time Use between Men and Women a. Labor market work and free time b. Household chores men, less women, hours women, less men, hours 3.5 5 4.5 Men – Women (in hours) 3 Women – Men (in hours) 4 2.5 3.5 2 3 2.5 1.5 2 1 1.5 1 0.5 0.5 0 0 Armenia Bulgaria Estonia Hungary Kazakhstan Kyrgyz Republic Latvia Lithuania Macedonia, FYR Poland Romania Turkey Armenia Bulgaria Estonia Hungary Kazakhstan Kyrgyzstan Latvia Lithuania Macedonia Poland Romania Turkey Work Free Time ECA Work HH Chores ECA HH Chores OECD HH Chores ECA Free Time OECD Work OECD Free Time Source: World Bank staff calculations based on data of the uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: The countries used to compute the regional average are Armenia, Bulgaria, estonia, Hungary, Kazakhstan, the Kyrgyz republic, Latvia, Lithuania, FYr Macedonia, Poland, romania, and Turkey. The OeCD countries include Austria, Belgium, Canada, Denmark, Finland, France, germany, italy, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the united Kingdom, and the united States. eCA = europe and Central Asia. HH - household. Women in the Labor Market 47 As a result, women have less free time than men. Labor Market Transitions Of the average three-hour difference in the amount Labor market transitions represent opportunities of time women and men in the region spend on and costs among individuals.38 They can lead to household chores per day, men spend two hours jobs that are more well remunerated, but also to on labor market activities and one hour resting; we lower wages after periods of inactivity or to unem- therefore see that women have less free time than men. This is quite consistent across the region. The most salient exception is Turkey, where men and 36 Noteworthy among the findings revealed through the women have almost the same amount of free time time use data is the relatively small amount of time spent on labor market activities across all countries. because, while women spend more time on house- Men report that they spend an average of five hours a hold chores, men spend an almost equal amount of day on labor market activities. This rather small num- time working on other activities, for example labor ber of hours can be partially explained by the fact that market activities.36 the unemployed and men and women who are out of The patterns in the gender division of work in the the labor force are included in the sample. If we con- household in the region are traditional; thus, women sider only the employed, the average among men rises to six hours a day. Although low, this number can be spend more time than men on caring for other house- partially explained by part-time jobs. hold members. Our calculations indicate that the 37 This finding needs to be viewed with caution because it traditional division of household chores between is based on only two countries: Armenia and Romania. men and women persists in the region (figure 2.14). In Romania, women report spending 2.5 hours a day Women spend three-quarters of their time in cook- taking care of other adults in the household. Women ing and the care of children and adults in the house- in Armenia report spending no time on this activity. 38 This section is mostly based on the work of Dimova, hold. Indeed, while a quarter of the time women Gang, and Landon-Lane (2006) for Bulgaria and Blunch spend on domestic chores is taken up by care for and Sulla (2011) for Serbia. A caveat should be noted: other adults in the household, men do not spend both studies analyze the situation during times of crisis any time on this task.37 and employment restructuring. Figure 2.14 The Distribution of Household Chores among Men and Women 40 % of total time spent in HH chores 35 30 25 20 15 10 5 0 Food Cleaning Laundry, Gardening, Shopping Child care Adult care Other Preparation, and Other Ironing, Construction and Domestic Dish Washing Upkeep Handicrafts and and Repairs Services Activities prod. textiles Women Men Source: World Bank staff calculations based on data of the uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: HH = household. 48 Opportunities for Men and Women: Emerging Europe and Central Asia ployment. The likelihood of experiencing a labor few years, employed women have been only 1 per- market transition varies depending on the sector cent less likely than men to become unemployed. of employment, the occupation, and gender. Panel Although the difference in the probability of mov- survey data are necessary for the analysis of labor ing from employment to unemployment between market transitions, but such data are rare. However, men and women is significant, the size of the dif- two recent studies on Bulgaria and Serbia are use- ference is negligible. There may be many reasons ful. Though the magnitude of the data is different, behind the difference. Thus, there is evidence that the study findings are consistent with results on the women are usually the first to be laid off during United Kingdom and the United States, on which a crisis. Yet, women may also be protected from abundant panel data exist. being laid off if, because of occupational segrega- Women are moderately less likely than men to tion, they work in industries that are less affected find jobs if they are unemployed, and they are more by crisis.41 likely to exit the labor force.39 During the global fi- Whether employed women are more or less nancial crisis and its aftermath in Serbia, unem- likely than employed men to leave the labor force ployed women were 4.2 percent less likely than depends mostly on age and educational attainment. men to find jobs (Blunch and Sulla 2011). There On average, employed women are only 0.3 per- are several possible reasons for this outcome. First, centage points more likely than employed men to women may spend less time than men searching leave the labor force. However, there are gender for work. Second, men may have access to more differences depending on age and educational at- efficient information networks on available jobs. tainment. For example, while workers of retire- Third, women may be searching for specific jobs ment age (65 and above) are less likely to drop (in selected sectors or occupations; see elsewhere out of the labor force than young workers (15–24 above) and may therefore require more time in un- years of age), women of retirement age are 43 per- employment waiting for vacancies in these jobs to centage points more likely than men in the age- become available. Fourth, women may suffer from group to retire. discrimination at various stages of the job search Selected evidence indicates that women are more and hiring process. Despite these adverse scenar- likely to stay in public sector jobs, while men are more ios, women end up taking jobs. Whether or not likely to stay in private sector jobs. A study on labor they have to lower their reservation wage to obtain market transitions in Bulgaria around 1996 finds jobs and, if they do, by how much have not been that the probability of leaving jobs in the public determined. sector is 67 percent among women and 62 percent Women are more likely to become discouraged among men (Dimova, Gang, and Landon-Lane workers. Workers are discouraged if they decide to 2006). The gender differences in these probabili- leave the labor force because they have been unable to find jobs under terms they consider acceptable.40 Blunch and Sulla (2011) estimate that unemployed women are 2.8 percent more likely than unem- 39 This assertion is based on analysis on Bulgaria and Ser- bia, but also on the United Kingdom and the United ployed men to be discouraged and exit the labor States. force. This may be indicative of the higher cost of 40 A worker may be discouraged because of the exten- the job search and the lower probability of success sive period in unemployment while waiting for work, in the job search among women compared with because of the terms of the wage offer, or because of men. A better understanding of the constraints on other reasons. See Sabarwal, Sinha, and Buvinic (2011) women would contribute to the design of policies for a detailed description of the added worker effect and the discouraged worker effect at the various stages to promote employment among women and im- of the business cycle. prove the functioning of the labor market. 41 This was the situation during the 1997 crisis in Indo- Employed women are as likely as unemployed nesia. See Hallward-Driemeier, Rijkers, and Waxman men to become unemployed. In Serbia during the last (2011). Women in the Labor Market 49 ties become more striking if we take into consid- force. (Box 2.2 examines the case of women in the eration the destination: men generally leave the informal sector.) public sector for jobs in the private sector, while Educated women in wealthier households are women generally leave the public sector to exit the more likely to be employed in the private sector. labor force.42 The probability that men employed in Women with higher educational attainment and the public sector will find jobs in the private sector living in households with higher incomes are more is 12 percent, while the corresponding share is only likely to be employed in the private sector and are 6 percent among women. Men working in the pub- less likely to leave the private sector for work in lic sector are also more likely than women in the the public sector, or because of unemployment public sector to move to unemployment (8 versus or simply because they are exiting the labor force. 4 percent, respectively). Likewise, women with higher educational at- Men working in the private sector are more likely tainment and higher household incomes are less than women working in the private sector to stay in the likely to remain out of the labor force. In Bulgaria, sector. The study in Bulgaria finds that 47 percent marital status and the presence of children in the of men employed in the private sector versus 36 household are additional explanatory factors in percent of women employed in the sector will be determining whether a woman is working or not working in the sector during the subsequent year. and whether she is working in the private or pub- This difference is partly explained by the greater lic sector. probability among women employed in the private sector to become unemployed or leave the labor Understanding Changes in the Labor Market Shifts in employment among men and women may be driven by changes in labor demand or in labor Box 2.2: The Informal Labor Market in Europe and supply, which may also differ across skilled and un- Central Asia skilled workers in the labor force. An analysis of a Do women have less access than men to formal sector jobs? labor market would help determine if a decline in Are they more likely to be trapped in the informal sector? labor force participation is the result of a change Informal employment may be defined in many ways. It usually refers to employment among workers who are not in the demand for labor or a change in the sup- (a) carried on the accounting books of an employer, (b) formally ply of labor. Usually, this type of analysis is car- registered, (c) included in the social security system, and/or ried out by differentiating between skilled and (d) receiving the basic benefits and protections provided in la- bor laws. unskilled labor. Shifts in wages and employment A recent study in Albania, Georgia, Hungary, Poland, Rus- may be indicative of whether a change in demand sia, and Ukraine finds that women are as likely as men to work in the informal sector. However, men typically earn more or a change in supply is dominant and help clar- than women in the formal sector and in the informal sector. ify the direction of the change.43 We may infer In the formal sector, women tend to have better skills, but obtain lower returns to education relative to women in the informal sector. The gender wage gap is generally driven by the difference between the wages of men and women in the 42 The authors of the study do not find evidence of a gen- formal sector, although, in Albania, most of the gender wage der effect in the exit of individuals from the public sec- gap is explained by the gap in the informal sector. Women in tor into unemployment or out of the labor force. the informal sector are also more likely to remain employed 43 A rise in employment and wages indicates an increase in in the informal sector relative to women in the formal sector and relative to men. demand (a positive shift). A rise in employment, accom- Women in the informal sector are also more likely than panied by a fall in wages indicates an increase in labor women in the formal sector to move back and forth between supply (a positive shift in supply). A fall in employ- informal employment and inactivity. This may be an indication ment and a rise in wages indicate a decrease in the sup- of women‘s self-selection into the informal labor market. How- ply of labor (a negative shift of supply). Finally, a fall in ever, even after we control for selection, we find that these results are confirmed. employment and in wages indicates a decrease in labor demand (a negative shift in demand). Obviously, there Source: Bardasi,Paci, and Pignatti (2007). may be a combination of shifts in the two curves, and the final result depends on the shift that is dominant. 50 Opportunities for Men and Women: Emerging Europe and Central Asia from aggregate indicators that the labor markets Box 2.3: Migration and Remittances in CIS Countries in Europe and Central Asia are diverse. In the last decade, average real wages and employment Migration and remittances have been increasing in Europe and increased in Bulgaria, Estonia, Hungary, and Lat- Central Asia in the last decade. Migration corridors exist from low-income CIS countries (Armenia, the Kyrgyz Republic, Mol- via. During the same period, an increase in labor dova, Tajikistan, and Uzbekistan in 2007) to middle-income CIS force participation and a fall in average real wages countries (Kazakhstan, Russia, and Ukraine), Western Europe (particularly Germany), Israel, and the United States. In addi- were observed in Armenia, Belarus, and Slovenia. tion, it is estimated that there are large flows of undocumented Meanwhile, an increase in average real wages and migrants. a decrease in labor force participation occurred The importance of remittances in some of these countries is not negligible compared with other developing countries. in Georgia, Lithuania, and Moldova. In all these In 2009, officially recorded remittance receipts as a share of cases, the changes in average real wages and em- GDP were 12 percent in Moldova and 35 percent in Tajikistan. Moreover, remittance growth rates have accelerated in some ployment moved in the same direction for both countries over the last decade. For example, remittances in the men and women. Kyrgyz Republic grew from negligible amounts to 21 percent The change in demand and the change in the during the period. A high proportion of migrants do not send money home demand for skilled labor relative to unskilled labor (from 17 percent in Bulgaria to 53 percent in Tajikistan). Most may partly be a consequence of shifts in the struc- migrants send small amounts of money: 60 percent of all mi- grants send less than US$200 a year. However, remittances are ture of economies because of the trade liberaliza- significant for the receiving households. Wealthier households tion that occurred in many of these countries. If tend to receive larger remittances. Remittances are greater in exporting sectors are intensive in unskilled labor absolute value among recipient households in the top of the income distribution. However, the importance of remittances and employ women, while importing sectors are in household expenditures is greater among households at the intensive in unskilled labor and employ men, trade bottom of the income distribution. openness will alter the relative prices of the two Sources: Quillin et al. (2007); Mansoor and Quillin (2007). labor inputs and may therefore affect the relative wages of men and women. The decrease in labor supply in some countries may be partly explained by an increase in education and in (out-)migration. If men and women decide during the 1990s, and the decline persisted dur- to acquire more human capital by pursuing ter- ing the next decade. In Serbia, female labor force tiary education, but also postgraduate education, participation fell from about 70 percent during so- they would delay their entrance into the labor cialist times to 58 percent in recent years (Babović market and, as a consequence, reduce the supply 2008, cited in Blunch 2010). of labor. Additionally, the decline in labor supply Evidence from analysis of the probability of may be caused by the large waves of migration transition in labor force status suggests that women observed in a few of these countries (box 2.3). self-select into or out of certain occupations. The se- Armenia, the Kyrgyz Republic, Moldova, Tajiki- lection of women in the labor force is positive, stan, and Uzbekistan show higher levels of out- which means that women with higher potential to migration, while Kazakhstan, Russia, and Ukraine produce value added tend to work, while less well receive immigrants. educated women tend to choose to remain out of The process of women and men following more the labor force. Moreover, among employed wom- traditional gender roles in the labor market (retradi- en, women who are more well educated tend to tionalization) that started with the transition and that work in the private sector. This indicates that fam- has resulted in a large decrease in female labor force ily responsibilities and the lack of support in the participation is continuing in certain countries. This decision by women to combine jobs and house- process has been clearly documented in the region hold chores represent important constraints and by Paci (2002), Paci and Reilly (2004), and a few that only women with higher earnings (and with other researchers. For example, in Bosnia and Her- husbands who also have higher earnings) can over- zegovina, female labor force participation declined come these constraints. Women in the Labor Market 51 The Gender Wage Gap Moldova, Turkey, the western Balkans, with the exception of Albania) have a gender wage gap of An examination of the gender wage gap can lead to about 20 percent, putting them in a similar or bet- a better understanding of the constraints that men ter position than the United States. Other coun- and women face in the labor market, whether these tries have gaps that are among the highest in the constraints arise from past or present choices (such world (Tajikistan, with a gap of about 65 percent), as the choice of occupation or the choice of the level while still others (Albania and Russia) have gaps of educational attainment), market frictions, or dis- in an intermediate range, about 35 percent. These crimination. First, we describe the raw gender wage magnitudes and rankings do not change dramati- gap in the region. Second, we describe the factors cally if the gap is computed conditioned on human contributing to the wage gap using decomposition capital characteristics. The averages by subregions methods that allow us to assess the effects of indi- are informative only in the case of the EU10, the vidual characteristics on the gender wage gap. Third, western Balkans, and Turkey because there are too we analyze changes in the wage gap over time. few data points on the other subregions. The aver- ages indicate a moderate regional gap, with values ranging from 17 percent for the EU10 in the raw The Raw Gender Wage Gap differential to 23 percent in the western Balkans Is there a gender wage gap despite the relatively for the Mincerian-based differential. significant equality in employment in the labor The gender wage gap is not a result of differ- market in the region? Wage gaps among workers ences in human capital. If we compare men and arise for many reasons, including location, type of women with the same level of human capital, work, level of education, and experience. Wage women are paid even less relative to men than gaps may also arise because there are too many or in the general case. However, the information too few women in the labor market or because the we have on human capital is partial: it only economy is expanding too quickly or too slowly. captures the level of educational attainment, Wage gaps may be partially a result of discrimi- but not the type or quality of education. Thus, nation, whereby employers perceive workers of a two persons may both be university graduates, particular gender or ethnicity as less capable and, though one may have a degree in history, and hence, not worthy of equitable wages. To deter- mine whether a wage gap exists simply because of the gender of workers, we analyze country-specific 44 There is a caveat: the data are limited. Our analysis data using regression analysis.44 relies on data on 12 of the 30 countries in the region. There is a significant gender wage gap in the These 12 countries are located across subregions. The region.45 The data indicate that women’s hourly EU10 are represented by 4 of the 10 countries in the wages are 22 percent less than the hourly wages of estimation sample; the western Balkans and Turkey men, on average, in the region.46 Moreover, because are fully represented; and the low- and middle-income women work fewer hours, the monthly wage gap is CIS countries are represented by only Moldova, Rus- sia, and Tajikistan. larger, indicating that women earn 29 percent less 45 This is based on 24 countries in the region. than men per month. The available data show that 46 This refers to the unweighted average of the gender women earn systematically less than men even if wage gap in the countries of the region. The gender one controls for differences in human capital. The pay gap in hourly wage rates refers to the gender gap in only exception is Bosnia and Herzegovina, where average hourly earnings. This indicator aims to capture the raw gap is negative; however, this gap disappears the difference between men’s and women’s overall positions in the labor market. It measures the differ- once we control for human capital characteristics.47 ence between men’s and women’s wage rates indepen- The extent of the gender wage gap varies greatly dent of the number of hours worked or the type of within the region. The differences across countries activity or occupation. are large (figure 2.15). Some countries (the EU10, 47 For example, see Ñopo, Daza, and Ramos (2011). 52 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 2.15 The Gender Wage Gap 60 50 40 In percentage 30 20 10 0 Georgia Azerbaijana Armenia Russian Federationa Albania Estonia Kyrgyz Republic Croatiaa Ukrainea Czech Republic Belarus Moldova Lithuania Slovak Republic Kazakhstan Tajikistan Hungaryb Macedonia, FYR Bulgariab Latvia Serbia Poland Romania Slovenia Sources: Miluka and grown (2010); Apostolova (2010); Blunch (2010); uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. a Monthly. b 2006. the other a degree in information technology, between female participation and the gender which are remunerated differentially by the wage gap only within the EU10. market. Thus, it may be that the type of human capital rather than simply the quantity of hu- man capital may play a role in explaining the The decomposition of the Gender Wage Gap gap. This subsection is based on a series of academic At the regional level, the participation rates of studies carried out on the region or worldwide, women do not impact the gender wage gap. Partici- but with abundant representation of the region. pation rate differences may potentially have a Most of the studies decompose the gender wage significant and large impact on the gender wage gap using different techniques, making compari- gap for two reasons. First, if few women are ac- sons across countries a challenge. Fortunately, in tive in the labor market, then discrimination and prejudice are more likely to survive, there- by generating a gender gap.48 Second, if female 48 The two major theories of discrimination (statisti- participation is low, then women participating cal discrimination and taste discrimination) have in the labor market may be a select sample of this implication. In the statistical discrimination the population. If the selection is positive (that model (Phelps 1972), the presence of fewer women in the labor market may lead to more discrimination is, participating women are relatively more pro- because employers have more difficulty in assess- ductive than nonparticipating women), the ob- ing the productivity of women. In the taste dis- served gap will underestimate the gender wage crimination case (Becker 1971), fewer women, plus differential; if the selection is negative, the gap some labor market frictions, imply that prejudiced will be overestimated.49 Because of these dif- employers can survive and continue to operate in the ferent sources, the impact of participation on market. 49 Paci and Reilly (2004) provide evidence of sample the gender wage gap is ambiguous and should selection in a handful of countries in the region. They be empirically verified. In Europe and Central find limited sample selection in the countries under Asia, there is a significant positive correlation consideration (see their table 5.3). Women in the Labor Market 53 one of these studies, Paci and Reilly (2004), we find a description of the implications in terms of comparability across different decomposition 50 More information on decomposition methods can be methods.50 In this subsection, we review the main found in Fortin, Lemieux, and Firpo (2011). The core conclusions that arise from the examination of of the information here is derived from Blunch (2010), gender wage gap decompositions, and, in the next Ñopo, Daza, and Ramos (2011), and Paci and Reilly subsection, we discuss the changes in the gender (2004). Other recent studies on the region are Apos- tolova (2010), Miluka and Grown (2010), Ognjenović wage gap during the two decades since the onset (2010), and Eriksson, Pytliková, and Warzynski (2010). of the transition. Blunch (2010) uses the United Nations Development The disparity across countries in the region makes Programme’s Social Exclusion Survey for six countries: any generalization difficult. Europe and Central Asia Kazakhstan, FYR Macedonia, Moldova, Serbia, Tajiki- is a region of wide contrasts. It is therefore diffi- stan, and Ukraine. Ñopo, Daza, and Ramos (2011) and cult to draw conclusions from the decomposition. most of the other studies use Living Standards Mea- surement Study surveys. For this reason, we discuss here the contributions 51 In the case of Europe and Central Asia, the choice of characteristics and returns to the gender wage of decomposition method does not seem to alter the gap. The contribution of each component depends conclusions. Paci and Reilly (2004) have conducted a on the data used, the selection of the sample, the meta-analysis of decompositions of the gender wage definition of the variables, and the decomposition gap performed on countries in the region during the method.51 1990s. They find that estimates are generally not sensi- tive to the decomposition method used. However, the fit of Mincer equations to the countries is not as good as the estimations performed in developed countries. The gender wage gap explained: differences in As a result, the estimates tend to overvalue the unex- characteristics plained part of the gender wage gap. This means that If we control for educational attainment, we find that data collection must be improved to foster more accu- the gender wage gap in the region increases. The gen- rate policy recommendations. Most of the comparability problems associated with der wage gap increases if we control for the charac- data quality arise from the measurement of two vari- teristics of men and women.52 As we see in chap- ables: earnings and experience. The estimated results ter 1, the differences between men and women in found in various studies are sensitive to the choice of terms of education are small in the region, but, in these two key variables. Studies that use hourly rate of any case, they favor women, who are more likely pay—as opposed to monthly wages—in the decompo- to complete tertiary education. The raw gender sition tend to obtain lower gender wage gaps because, in the region, men tend to work more hours per week gap is 18 percent; after we control for education, than women. Second, using age as a proxy for experi- the gap rises to 27 percent. This means that, at the ence instead of actual or potential experience also same educational level, women earn an average of decreases the estimated gender wage gap. US$27 less per US$100 earned by men. Moreover, 52 Ñopo, Daza, and Ramos (2011) use a semiparametric only US$9 of the US$100 earned by men can be decomposition technique adopted from the program explained by educational attainment. evaluation literature. 53 Full-time work and part-time work are defined by a The gender wage gap is partly explained by self- set of dummy variables, including working less than selection among women into lower paying jobs and oc- 20 hours a week, working between 20 and 40 hours cupations. The gender wage gap remains constant a week, and working more than 40 hours a week. The if we control for occupation, industry, and job type of employment is defined as a set of dummy vari- characteristics, such as number of hours worked, ables for employee, employer, or self-employed. A rise whether the job is in the formal or informal sector, in the number of hours of work per week considerably increases the gender wage gap. This is an indication of and whether the worker is an employee, an employ- the existence of a wage penalty for women and part- er, or self-employed.53 In particular, Europe and time workers. In contrast, work in the formal sector Central Asia is the region with the highest level of and certain types of employment decrease the gender segregation, and this contributes to explain the gen- wage gap. 54 Opportunities for Men and Women: Emerging Europe and Central Asia der wage gap. At the same time, these variables ex- bor market. This may be pure discrimination or plain only about 9 percent of the gender wage gap. statistical discrimination. Additionally, analyses This means that only a third of the gender wage gap have found that, in a situation of great variability in is explained by worker and job characteristics, in- wages across a group, workers with greater ability cluding occupation. signal their greater productivity by acquiring more The field of study is a minor determinant of the education.55 This may be the prevalent situation in gender wage gap. The field of study is significant, Europe and Central Asia. but does not play a major role in explaining the gen- der wage gap. Job characteristics, including occupa- tion, play a more important role. While job char- Change in the Gender Wage Gap over Time acteristics explain almost 10 percent of the gender The unexplained gender wage gap decreased in the region wage gap, field of study explains only 1 percent of during the 1990s. An interesting result of the analysis the gap. The association between field of study and of Paci and Reilly (2004) arises after the incorpora- occupation is strong, however. tion of a trend in the meta-analysis of the decomposi- The determinants of the gender wage gap vary tions of the gender wage gap during the 1990s.56 This in some countries in the region. A study by Blunch trend was found to be negative, indicating that the (2010) that relies on different data and a differ- unexplained gender wage gap decreased by about 1 ent decomposition method finds alternative re- percentage point per year during that decade. sults for Kazakhstan, FYR Macedonia, Serbia, and However, in many countries in the region, the Ukraine. Whether occupational segregation is a unexplained gender wage gap increased during the constraint or a product of self-selection should 1990s. Assessing the change in the unexplained be the subject of more research, but the evidence gender wage gap in each country (or in each study), on labor market transitions discussed above indi- we see that, in 11 of the 26 countries examined in cates that women tend to stay in the public sector, Paci and Reilly (2004), there was an increase in the partly because of nonpecuniary benefits such as gender wage gap.57 Moreover, in many cases, the maternity leave, flexible hours, or job attachment magnitude of the increase was significant. For ex- (Dimova, Gang, and Landon-Lane 2006; Jurajda ample, in less than 10 years, the gap increased from 2003). 26 to 45 percent in Armenia, from 29 to 47 per- cent in Azerbaijan, and from 22 to 57 percent in Tajikistan. The unexplained gender wage gap More analysis and more data are needed to as- The conclusion we may draw from all the studies sess whether the trends in the evolution of the gender is that most of the gender wage gap in the region wage gap observed in the 1990s continued during the remains unexplained. This result contrasts with last decade. Whether these patterns persisted in the usual finding on Western Europe or the United the 2000s is difficult to assess given that there is States, where education, experience, occupation, and industry explain about 90 percent of the gap.54 The significant unexplained portion of the gender 54 For more details about the United States, see O’Neill wage gap points to discrimination: women simply (2003), Blau, Ferber, and Winkler (2002), and Blau do not have access to equal pay. and Kahn (1997, 2004). Our two main explanations for the gender wage 55 Lang and Manove (2006) propose a model that com- gap are discrimination in the labor market and the bines statistical discrimination and educational sorting need to give a cost signal to employers if women show that explains why blacks obtain more education rela- greater variability in productivity. The archetypal tive to whites at similar cognitive ability. 56 Technically, a trend is a variable the value of which interpretation of the increment in the estimated increases by 1 each year to reflect a linear evolution of gender wage gap if we control for educational at- the independent variable over time. tainment is that there is discrimination in the la- 57 See table 3.2 in Paci and Reilly (2004) for more details. Women in the Labor Market 55 no study, to our knowledge, that compares find- The generous maternity leave benefits in certain ings based on the use of the same data, but vari- countries in Europe and Central Asia seem to rep- able definitions of the gender wage gap. Analysis resent a response to low total fertility rates. All the of the evolution of the gender wage gap over time countries with generous maternity leave benefits can contribute to understanding how the labor have total fertility rates below the replacement market reacts to changes in the composition of level. Bosnia and Herzegovina has the lowest to- the labor force as the characteristics of entrant co- tal fertility rate and the most generous maternity horts change. It can also help in understanding the leave policy. However, two other countries—Al- consequences of changes in institutions and wage bania and Montenegro—with generous maternity structures. leave laws have considerably higher total fertility rates than Bosnia and Herzegovina (1.6 and 1.9, re- spectively, versus 1.2 children per woman in Bos- The Implications for Policy Design nia and Herzegovina; all three provide 365 days of paid maternity leave). In this section, we discuss how maternity leave, Parental leave benefits are unusual in the re- childcare policies, and pension benefits relate to gion. Parental leave benefits are almost nonexistent gender wage gaps in employment and pay. in Europe and Central Asia. Only two countries offer parental leave benefits: Azerbaijan, which offers 14 unpaid days of parental leave, and Lat- Maternity and Parental Leave via, which offers 10 paid days of parental leave.59 Maternity leave is quite generous in Europe and Cen- This indicates that the region is still conservative tral Asia. Maternity leave was reformed in most regarding the scope of parental leave benefits, and, countries through the revision of labor codes dur- despite the low fertility rates observed in many ing the transition to a market economy. Most of countries, none of them have adopted more mod- these laws are more generous than the old laws or ern systems, such as the systems of Norway and the laws prevalent in developed countries. Table Sweden. 2.2 shows the maternity leave benefits available in Increasing parental leave benefits is a way to in- Europe and Central Asia. Most countries offer, on centivize higher fertility rates. The question whether average, six months of paid maternity leave.58 This parental leave benefits, especially maternity leave, is quite a generous policy compared with poli- are sufficient to incentivize higher fertility rates is cies in countries such as the United States, where unresolved. However, the available evidence seems maternity leave is only 12 weeks, but stingy com- to indicate that there is a positive association be- pared with policies in countries with low fertility tween the two.60 In Sweden, the change in family rates such as Sweden, where the maternity leave is 15 months. There is some variation across countries in the re- 58 In some countries, the basis of the paid leave varies gion in maternity leave benefits. Maternity leave ben- with the length of the maternity leave and ranges from the full wage of the last working year to 85 percent of efits vary considerably across the region. In Bosnia the last annual wage. and Herzegovina and in FYR Macedonia, the ben- 59 Bulgaria offers a year of paid leave, as well as shorter efits are generous and comparable with those in partially paid leave, for single male parents. the Nordic countries; they are 365 and 270 days, 60 This assertion is based on the estimates of Ruhm respectively. In Romania and Ukraine, the mater- (1998), who examines data on selected Western Euro- nity leave is 112 days, similar to the benefit in the pean countries (Denmark, Finland, France, Germany, Greece, Ireland, Italy, Norway, and Sweden). More United States. Most of the countries in the region extended entitlements (nine months) raise the pre- offer 130 days of paid or partially paid maternity dicted ratios of female employment to population by leave. approximately 4 percent, but with a decrease in hourly wages of around 3 percent. 56 Opportunities for Men and Women: Emerging Europe and Central Asia Table 2.2 | Maternity Leave Legislation Mother Father Number of days paid leave additional leave before Number Child care leave Illegal job Partially equally Child Country birth after birth all of days until the termination paid leave divided Paid Unpaid Care Albania 42 365 yes Azerbaijan 70 56 126 14 yes a 14 Bosnia and 42 365 1 yr Herzegovina Bulgaria 45 135 2 yrs old b Croatia 45 365 yes Czech Republic 42 196 yes Estonia yes Hungary 28 168 3 yrs old yes Kazakhstan 70 56 18 months old Kyrgyz Republic 70 56 8 months old yes Latvia 56 56 10 Lithuania 70 56 90 Montenegro 45 365 Poland 112 Moldova 70 56 3 yrs old Romania Russian 70 70 14 3 yrs old yes Federation Slovakia 42 196 3 yrs old yes 3 yrs old Macedonia, FYR 45 270 yes Turkey 56 56 180 Ukraine 112 Source: gender Law Library (database), World Bank, Washington, DC, http://wbl.worldbank.org/WBLLibrary/elibrary.aspx?libid=17. Note: The information has been retrieved from the gender Law Library using the Labor Code of each country. a Single parents until the child is 3 years old. b Single parents only: 1 year. policies that began in the 1980s contributed to an Increasing maternity leave benefits may reduce increase in the total fertility rate (Hoem 1990). female labor force participation. The other relevant In the United States, it has been found that wom- question we must ask in analyzing maternity leave en with high rates of desired fertility do not sort benefits is whether the benefits will discourage themselves into jobs with more maternity benefits. female labor force participation. As in the case of However, women already working in jobs with fertility rates, there is no unanimous answer to this higher maternity benefits are more likely to have long-standing question. However, in the case of the more children (Averett and Whittington 2001). United States (where the work culture is consider- The question remains unanswered whether these ably different relative to other parts of the world), benefits offset the high costs of generous maternity it has been shown that, among women who had benefit systems. jobs before giving birth, those with maternity leave Women in the Labor Market 57 benefits return sooner to their jobs (immediately long-term care are expected to double in almost after the 12 weeks of leave) than those without the all Eastern European and former Soviet countries benefits. In general, maternity leave coverage is not and will eventually account for between 0.5 and related to female labor force participation, at least 1.0 percent of GDP. The projections of a World not in the United States, but to the length of the Bank study show that, if institutionalized care is leave benefit.61 extended to cover 20 percent of the elderly with Interruptions in labor force participation because disabilities, expenditures on long-term care alone of childbearing reduce wages. The interruptions in will consume between 2 and 4 percent of GDP labor force participation related to childbirth and (Alam, Anós Casero, and Khan 2008). Countries child-rearing have a definite negative impact on in the region must carefully deliberate on the policy women’s wages. It has been estimated that, in the choices in the provision of such services. Informal United States, a 12-month interruption in full-time care will be an increasingly important part of such jobs because of childbearing reduces the lifetime policies. Women as informal providers of care will incomes of women by about 10 percent.62 There be greatly affected. are multiple mechanisms through which this oc- The opportunity costs of any increases in care curs. First, women do not accumulate any work responsibilities will be particularly high, especially experience while they are out of the labor force, among younger women. A recent World Bank re- placing them in an unfavorable position if they port warns of the dangers of increasing the burden have to compete with men for jobs. Second, dur- on the declining number of informal caregivers, ing the time women are out of the labor force, their that is, women currently not in employment. The labor market skills can depreciate. Moreover, it is capacity and willingness of informal caregivers to more difficult to find a job if one begins the search while out of the labor market rather than while unemployed (Omori 1997; Fallick, Haltiwanger, 61 For more details, see Berger, Hill, and Waldfogel (2005). and McEntarfer 2010; Fernández-Kranz and Rodrí- 62 For more details, see Francesconi (2002), who builds a structural model of fertility and work decisions and guez-Planas 2011). provides illustrative simulations of the benefits of each Women perceive interruptions in labor force decision. He compares a profile of no full-time job participation because of childbearing as detrimental interruptions to each of two alternative choices: a year- to their careers. In seven countries of the region, long interruption after childbirth versus a year of part- 25 percent of women who had each had at least time work after childbirth. His simulations show that one child since 1987 stated that their job inter- there is no statistically significant difference between part-time work and interruption choices. Working part ruptions because of childbearing had negative time is slightly better than being out of the labor force, consequences on their careers.63 Only 14 percent but the difference is negligible. The other interesting of the women in these countries who had had at result of the exercise is the finding on the cost of the least one child before 1987 had the same view, im- interruption of full-time work, which is about 25 per- plying that the experience about which they were cent of income utility relative to a full-time working concerned occurred after the end of the socialist career in the short run, followed by a rebound effect once a woman returns to full-time employment. How- period. In Poland and Ukraine, respectively, 50 ever, the total loss in income over the life-cycle is about and 40 percent more women reported negative 10 percent. This is for the most conservative simula- consequences compared with the average across tion involving the birth of only one child and only one the seven countries.64 year-long interruption. The losses are obviously higher with more interruptions (that is, more children). 63 The seven countries are the Czech Republic, Estonia, Hungary, Poland, Slovenia, the Slovak Republic, and Childcare and Elderly Care Ukraine. Population aging will disproportionately impact wom- 64 For more details on women’s perceptions of the con- en as the supply of health services declines and the need sequences of career interruption because of childcare, for the care of the elderly increases. Expenditures on see Zhelyazkova and Valentova (2009). 58 Opportunities for Men and Women: Emerging Europe and Central Asia continue providing care is also a major concern: ent over the past 12 months and is available un- “There is a real danger of unpaid informal caregiv- til the child is 2 years old (box 2.4). This type of ers becoming overloaded and feeling compelled policy does not specify which parent should stay to move their elderly family members to an insti- at home to receive the benefit, though it is more tution� (Chawla, Betcherman, and Banerji 2007, likely that mothers will. This raises the question of 34) Such caregivers will need to be provided with whether this policy will have the same impact as support. a maternal leave policy. Answering this question Looking after other persons, raising children, and would require significant national dialogue, but performing housework are the main reasons women also thorough consideration of the entire range of give for staying out of the labor force or working part family policies, such as the maternity leave, so as time. A recent Eurostat survey finds that women to be consistent, gain efficiency, and foster accu- are “held back from working full-time by the work rate targeting. that they carry out raising children, looking after The enrollment rate in childcare facilities among other persons, or housework; this was the single under-3-year-olds is lower, on average, in Europe and largest reason (36.0 percent) for women working Central Asia than the observed average rate in se- less than 30 hours per week in 2007, in contrast to the proportion (5.0 percent) of men for whom this was the reason that they worked less than 30 hours per week� (European Commission 2010, 27) (fig- Box 2.4: Family Policies in Romania ure 2.16). The state child allowance is a monthly cash transfer to all chil- Innovation in policy design is needed to avoid dren age 0 to 18 (or more if the child is still in school), with dif- unwanted substitution effects between raising fertil- ferentiated benefit levels for all children 0 to 2 years old (RON 200), children 2 to 18 years old (RON 42), or children with dis- ity rates and discouraging female labor force partici- abilities who are 0 to 3 years old (RON 200) or 3 to 18 years old pation. In Romania, the child allowance benefit is (RON 84). The value of the benefit has been raised five times since it was implemented. a monthly cash transfer for a parent who stays at The child-raising benefit is a monthly cash transfer for a home to take care of a child. It is equivalent to 85 parent who stays home to care for a child. It is equivalent to percent of the average income earned by the par- 85 percent of the average income earned by the parent over the previous 12 months. The value ranges from a minimum of RON 600 to a maximum of RON 4,000. It is granted until the child turns 2 years old or 3 if the child is disabled. On Janu- ary 1, 2011, the parameters of the program were modified. A parent who opts to participate in the program for two years Figure 2.16 Reasons Men and Women receives a monthly allowance equivalent to 75 percent of the Work Less Than 30 Hours a average income earned by the parent over the previous 12 months. The value of the benefit ranges from a minimum of Week, European Union, 2007 RON 600 to a maximum of RON 1,200. A parent who opts to percent of responses participate in the program for one year receives a monthly allowance equivalent to 75 percent of the average income Housework, looking after earned by the parent over the past 12 months. The value of children or other persons the benefit ranges from a minimum of RON 600 to a maximum Do not want to work more of RON 3,400. Parents who opt for the one-year program and Cannot �nd a job return to work before the end of the program are eligible for for more hours a back-to-work bonus of RON 500 per month for the second Number of hours in all jobs year upon their return to work. are considered as full time These programs have resulted in a modest increase in Personal illness or disability fertility rates at the expense of a reduction in labor force par- Undergoing education ticipation among parents of childbearing age. These programs or training may be revised in light of further analysis to achieve the same Other reasons goals in a more efficient and simpler way. The overlap in the programs increases costs, allows room for error and fraud, and 0 10 20 30 40 could be eliminated or reduced through simplification and co- ordination with other family and social programs. Women Men Source: World Bank (2011c). Source: european Commission (2010). Women in the Labor Market 59 lected Western European countries. On average in In the region, mature women who retire early Europe and Central Asia, only 13 percent of chil- could become caregivers for children and the elderly. dren under the age of 3 are sent to childcare. In Countries in the region could take advantage of the Western Europe, the rate is 34 percent (figures fact that women are retiring relatively young and in 2.17 and 2.18). While childcare enrollment in Eu- good health and exhibit high life expectancy. These rope and Central Asia fell considerably during the women could provide both childcare and elderly 1990s and has not increased in the last decade, the care. In many countries in the world, grandmoth- trend is quite the opposite in Western European ers are providing childcare. For example, since the countries. introduction of the grandparent childcare benefit, The low enrollment rate in childcare is a conse- the Australian government has been paying grand- quence of the lack of childcare services. The enroll- parents who take care of their grandchildren for up ment rate in childcare among under-3-year-olds to 50 hours per week. The United Kingdom is con- varies greatly across the region. This large variation sidering a similar policy; the city council of Not- may be caused by a lack of labor market opportuni- tinghamshire conducted an experiment in 2004 in- ties among women that would encourage women volving payment to grandparents on a weekly basis to send their young children to childcare and par- to take care of their grandchildren. Although this ticipate in the labor market or by social norms or kind of policy is not consistent over time because, preferences, but also by a lack of childcare services. as women become more attached to the labor mar- This last seems to be a strong determinant. For the ket, their opportunity cost in providing care in- small sample of countries in the region on which creases, it may represent a fruitful resource for the we have information on childcare availability, near future given the characteristics of the 55–65 childcare enrollment rates increase in tandem with age cohort among women.65 the places available in childcare facilities. Providing Awareness should be raised to demystify care as a more (subsidized) childcare could serve not only to female activity and to involve men more regularly. As promote female labor force participation, but also we see elsewhere above, most of the constraints— to increase fertility rates. many times, self-imposed—in the labor market are related to the fact that women have to carry out work on the labor market, as well as household chores. The elimination of some of these constraints arises through the development of new technologies Figure 2.17 Childcare Enrollment Rates that reduce the time required to perform household among Under-3-Year-Olds chores. To improve the opportunities for women per 100 in the age-group to gain access to the same jobs as men, husbands 40 should become more involved in household chores, 35 particularly in childcare and elderly care. Percentage of under 30 3 children enrolled 25 20 Pensions 15 10 The aging of the population has raised concerns 5 about the sustainability of pension systems in view 0 of the growing number of beneficiaries and the de- 1980 1990 1995 2000 2001 2002 2003 2004 2005 2006 2007 2008 clining number of younger contributors. However, EU and World Bank projections show that appro- ECA Western Europe Source: uNeCe Statistical Database, united Nations economic Com- mission for europe, geneva, http://w3.unece.org/pxweb/. 65 For more details, see Posadas and Vidal-Fernández Note: eCA = europe and Central Asia. (2011). 60 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 2.18 Share of 0- to 3-Year-Olds Enrolled in Childcare, 2008 percent 50 45 40 35 30 25 20 15 10 5 0 Poland Tajikistan Kyrgyz Republic Romania Azerbaijan Hungary Czech Republic Armenia Macedonia, FYR Montenegro Bulgaria Moldova Ukraine Croatia Latvia Russian Federation Lithuania Estonia Slovenia Childcare enrollment ECA Western Europe Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: eCA = europe and Central Asia. priate and timely policy reforms can significantly the pension system deriving from working women; reduce the impact of aging populations on pension on the other hand, it will reduce the period over systems. which these benefits need to be spread, which is In most countries in Europe and Central Asia to- particularly important given that women have lon- day, women are able to retire at a younger age relative ger life expectancy. In addition to enhancing the to men. The average age of retirement in the region sustainability of pension systems, longer tenures is 55 for women and 57 for men. This is consid- will boost women’s pension benefits, as well as the erably lower than the observed average in OECD opportunities of women to reach senior decision- countries, where the age for both women and men making positions. is 60. In Turkey, the minimum retirement age var- To avoid unintended secondary effects on fertil- ies depending on whether the person was registered ity rates, while encouraging female labor force partici- with the system in 1999 or not. For new entrants pation, some countries in the region have considered since 1999, the retirement age for women has been linking the retirement age to fertility rates. Pronatalist set at 58 and for men at 60 (except in one of the policies in many countries in Europe and Central three current schemes). The minimum retirement Asia require women to work fewer years than men. age for people who were already in the system in In the Slovak Republic, for example, the retirement 1999 has been increased to 52 for women and 56 age for women depends on the number of children for men (Chawla, Betcherman, and Banerji 2007). the women have raised: the greater the number of To avoid a severe fiscal crisis related to the lack children, the lower the retirement age. However, of sustainability in a pension system, labor force par- because the fertility decision is based on a complex ticipation should increase, particularly among women. mix of factors, including the cost of bringing up To raise labor force participation, an untapped area children, the opportunities for women’s participa- of reform is the equalization of the retirement age tion in higher education and employment, house- among women and men. Raising the age of retire- hold economic status, marital status (including ment for women will have a doubly positive effect. divorce and cohabitation), and the degree of com- On the one hand, it will increase the revenues of patibility of work with childcare, the balance of Women in the Labor Market 61 evidence suggests that the effect of transfer-based to improve labor productivity may potentially off- pronatalist policies is negligible (Chawla, Betcher- set the effects of reductions in the size of the labor man, and Banerji 2007). force. For example, growth decomposition exercises Because growth in labor productivity has been the show that, in most of the countries in the region, main engine of growth in the region, it is essential that the growth in labor productivity in recent years has countries implement policies to continue benefiting from been the single greatest contributor to increases in this trend. The cornerstone of the reform strategy per capita incomes (Chawla, Betcherman, and Ba- going forward is increased productivity. Measures nerji 2007). 62 Opportunities for Men and Women: Emerging Europe and Central Asia Women in Entrepreneurship 3 Introduction The expansion and recognition of the formal standing of entrepreneurship were among the major changes brought about by the transition from the planned economy to the lib- eralized market economy in Central and Eastern Europe and Central Asia in the 1990s. In the rapidly changing social and economic environment of the transition countries, entrepreneurship held the potential to contribute to both economic development and social inclusion. Moreover, entrepreneurship also provided another means of income generation among men and women in economies characterized by limited growth in employment even during high-growth years. Entrepreneurship also has a broader economic impact and is important from the perspective of job creation, private sector development, and wealth creation. Women’s participation in entrepreneurship can enhance the expansion of these economic goods and simultaneously lead to less inequality in the two largest subgroups in the popula- tion: men and women. However, in addition, as discussed elsewhere in this chapter, businesses owned and managed by women are different from businesses owned and managed by men; thus, the gender specialization that appears to occur in the labor mar- ket is reflected in entrepreneurship as well. One important issue is whether the smaller number of women entrepreneurs and the differences in the businesses of women reflect preferences or are the result of constraints placed on women that unfairly target the ability of women to grow businesses and accumulate wealth. The results reported in this chapter indicate that, though there is significant rep- resentation of women in leadership positions in firms as owners or managers, their rep- resentation is well below the level commensurate with either women’s presence in the labor market or women’s educational qualifications. The chapter explores the issues and relative challenges faced by women entrepreneurs. In brief, we find as follows: 63 ◆ Women constitute about one in three entre- explores the relative constraints faced by women preneurs in the region. Though the relative entrepreneurs. The fourth section discusses the proportion is comparable with the interna- performance of these entrepreneurs. The final sec- tional average, there are fewer women and tion concludes by providing the main findings and men entrepreneurs as a share of the work- recommending policy reforms. force in Europe and Central Asia relative to other regions. ◆ The types of firms operated by women and The Involvement of Women in men entrepreneurs are different. The firms Entrepreneurship of women are smaller and are concentrated in different subsectors in both services and Entrepreneurship and business creation can be sig- manufacturing. nificant contributors to private sector enlargement ◆ In aggregate, firms owned by women perform and to the dynamism of an economy. However, it is less well than firms owned by men. This trend difficult to estimate the size of the private sector in could be changed if the scale of the business- any economy in Europe and Central Asia. The same es of women were expanded. This might be is true of women’s contributions to and participation achieved through the expansion of credit, as in entrepreneurship. Indicators that capture informa- well as the expansion of access to business tion on the value added by the private sector or the development services, especially better links share of private sector employment in total employ- with supply chains. ment are broadly lacking in the region. Nonetheless, limited data exist on the amount of new business The remainder of this chapter is structured creation in the formal sector, and this gives a sense of as follows. The next section presents available data private sector activity (see figure 3.1). on women who are business owners, managers, Though we cannot quantify men’s and wom- and founders. The second section analyzes gender- en’s contributions to the private sector, we can based differences in the characteristics of entrepre- compare and contrast the performance of men and neurs and their enterprises. The following section women as entrepreneurs to give some indication of Figure 3.1 Women and Men Employers a. Density of new business, 2009 b. Women and men employers, 2008 new registrations per 1,000 people 15–64 workforce by gender, % 9.0 12.0 8.0 10.0 10.0 7.0 6.0 8.0 5.0 5.9 5.9 6.0 4.0 3.4 3.0 4.0 2.7 2.3 2.0 1.1 1.7 2.0 1.0 0 0 Tajikistan Poland Lithuania Russian Bulgaria Estonia ECA LAC MNA High income Federation OECD 2009 ECA Female Male Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: eCA = europe and Central Asia average. LAC = Latin America and the Caribbean. MNA = Middle east and North Africa. OeCD = Organisa- tion for economic Co-operation and Development. 64 Opportunities for Men and Women: Emerging Europe and Central Asia the gender gap in this important aspect of econom- Self-employment rates and employer rates are ic activity. This section describes women’s involve- negatively correlated, but only among women. If self- ment in business as entrepreneurs in Europe and employment among women declines, the propor- Central Asia. There is no single definition of entre- tion of women employers increases. This may be preneurship, which includes varied economic ac- a reflection of the difficulty self-employed women tivities, such as ownership, strategic management, face in growing their own businesses. This may be and the founding and establishment of businesses. because, relative to men, women encounter great- Moreover, to take advantage of the few available er obstacles than men, or it may be that the inci- gender-disaggregated data sets or regional and na- dence of self-employment among women reflects tional studies, our analysis is opportunistic and a market necessity, while the status of employer adopts the definition that best allows us to exploit represents a career choice made by women who, the information at hand to gain insights. among other reasons, desire to expand their wealth (see box 3.1). Self-employment rates among men and women are strongly negatively correlated with Women Employers income per capita as measured by the log of gross There are relatively fewer employers in Europe and Cen- domestic product (GDP) per capita at purchasing tral Asia than globally.66 The concept of entrepre- power parity, whereas the incidence of employers neur includes the concept of employer. Persons who among men and women increases somewhat as in- employ others constitute only a small share of the come per capita increases. total work force in the economies of the region, but The better business environment has not led to a also in other parts of the world. The average share of greater prevalence of women or men employers in the employers among the employed was 3.1 percent in region. The Doing Business indicators provide one Europe and Central Asia in 2008. The share varied measure of the ease of doing business in a coun- from about 6 percent in Turkey to about 1 percent try, though only in the formal sector (World Bank in Armenia, Georgia, and Moldova. If one compares and IFC 2009). Europe and Central Asia perform this with the high-income countries of the Organisa- relatively well compared with other regions of the tion for Economic Co-operation and Development world. The average rank of Europe and Central (OECD) and Latin America and the Caribbean Asia in the ease of doing business was 71 in 2009, (both 4.3 percent), one finds that there are relatively fewer employers in Europe and Central Asia. In Europe and Central Asia, women are more 66 The concept of employers used here reflects the defi- likely to be employers if men are also more likely to be nition of the International Classification by Status in employers. The share of male employers in the to- Employment, of 1993 (http://laborsta.ilo.org/applv8/ tal male workforce is 3.4 percent, compared with data/icsee.html), as follows: “Employers are those 1.1 percent among women; both of these shares are workers who, working on their own account or with low by international standards (figure 3.1). Though one or a few partners, hold the type of job defined as there is a large difference in employer rates by gen- a ‘self-employment job’ [that is, jobs where the remu- neration is directly dependent upon the profits (or the der, they are strongly correlated (0.8); countries potential for profits) derived from the goods and ser- with higher shares of male employers are also like- vices produced], and, in this capacity, on a continuous ly to have higher shares of women entrepreneurs basis… have engaged one or more persons to work for among all employed women. The exceptions are them in their business as ‘employee(s)’.� Armenia, Azerbaijan, and Turkey, where the ratio 67 The employer gender gap is the difference in the among male and female employers is large, and the employer rates between men and women. However, given the variations in employer rates in the region, the concentration of male employers is particularly ratio of male to female employer rates may be more high. The ratios are closest in Kazakhstan and the appropriate as a measure. This employer gender ratio Russian Federation, though the overall proportion shows that the same countries are at the extremes of of entrepreneurs is low in these two countries.67 the distribution. Women in Entrepreneurship 65 estimates of the informal sector appear to indicate Box 3.1: Entrepreneurship: Opportunity or Necessity? that, in terms of the contribution to GDP, a sub- Some men and women are motivated to become entrepreneurs stantial portion (though, perhaps, not the majority) to pursue business opportunities, while others are pushed into of the enterprises in Europe and Central Asia are, entrepreneurship by necessity. Those entrepreneurs who es- indeed, active in the formal sector (World Bank, tablish and grow their enterprises because they enter into busi- ness willingly may have different attachments to their busi- forthcoming). Consequently, any analysis of formal nesses and different risk profiles relative to those who pursue firms (for example, registered firms) provides use- entrepreneurship because they see few other attractive oppor- tunities in the market. ful insights into overall activity in the private sector In 2007, the Global Entrepreneurship Monitor involved an of a country. effort to collect data on entrepreneurship in 41 countries. Nine Entrepreneurship and ownership are closely of the countries were in Europe and Central Asia (Croatia, Hun- gary, Kazakhstan, Latvia, Romania, Russia, Serbia, Slovenia, associated and are often considered identical in the and Turkey). Though, globally, opportunity was the dominant case of companies owned by individual owners or motivation among men and women entrepreneurs, this was less true in Europe and Central Asia. On average, about half groups of owners. In most countries in Europe and the women entrepreneurs in any individual country pursued Central Asia, the majority of firms are not publicly entrepreneurship because of opportunity. However, there was traded because most firms are either owned by in- significant variation. At one extreme, in Slovenia, the wealthiest country in Europe and Central Asia, almost four women entre- dividuals or by small groups of individuals, many preneurs in every five saw entrepreneurship as an opportunity. of whom are related. This subsection discusses the However, in Croatia, Serbia, and Turkey, only 25–45 percent of women who self-identified as entrepreneurs reported that they ownership of firms disaggregated by men and wom- had established their businesses because of opportunity. en in the region and in specific countries. In 2009, women participated in the ownership of Source: Allen et al. (2008). more than one-third of all registered firms in the re- gion.70 At 36 percent of all firms, the share is slightly above the global average, though it is significantly below the share in East Asia and the Pacific, which compared with 83 and 95 for East Asia and the exhibits the highest rates of female participation Pacific and for Latin America and the Caribbean, respectively. The overall rankings of the region for the friendliness of the business environment 68 The rate of new business registrations is the annual and for the prevalence of employers do not ap- rate at which new businesses are registered per 1,000 pear to be positively correlated. However, there persons 15–64 years of age. 69 The European Union’s current definition categorizes is a strong negative relationship between the Do- companies with fewer than 10 employees as microen- ing Business ranking and the rate of new business terprises, while those with fewer than 50 employees registrations.68 This apparent contradiction may are small, and those with fewer than 250 employees indicate either that not all employers are in the are medium. formal sector or that improvements in the busi- 70 The data are taken from the Enterprise Surveys (data- ness environment are more recent and have not base), World Bank, Washington, DC, http://www. enterprisesurveys.org/. The sample covers formal (reg- yet resulted in a significant shift in the stock of istered) firms with five or more employees in the man- business owners. ufacturing and service sectors, which corresponds to International Standard Industrial Classification codes 15–37, 45, 50–52, 55, 60–64, and 72 (see United Women Owners in the Formal Sector Nations, “Detailed Structure and Explanatory Notes: In most economies, small and medium enterprises ISIC Rev.3.1,� Statistics Division, United Nations Department of Economic and Social Affairs, http:// are more prevalent than larger ones.69 These enter- unstats.un.org/unsd/cr/registry/regcst.asp?Cl=17). prises are also frequently responsible for driving Service firms include firms in construction, retail, innovation and competition in the private sector. wholesale, hotels, restaurants, transport, storage, Though not all microenterprises and small enter- communications, and information technology. State- prises are necessarily active in the formal sector, owned firms are excluded. 66 Opportunities for Men and Women: Emerging Europe and Central Asia in ownership among all developing regions (figure with the concentration of women employers (see 3.2). The different rates of female entrepreneurship elsewhere above). across countries may reflect many factors, includ- Ownership within firms is highly concentrated ing the level of development of the private sector, in the region; thus, the ability of women and men the share of firms in sectors in which women are joint owners to exercise their ownership rights is po- concentrated, and woman-friendly credit policies. tentially limited. On average, an estimated 82 per- The countries in Europe and Central Asia with the cent of domestic firms in the region are owned highest rates of female participation in ownership by the largest shareholders, whereas, in other are Belarus, the Kyrgyz Republic, and Moldova regions, majority shareholders own from 67 per- (50–60 percent), while the countries with the low- cent (South Asia) to 78 percent (in OECD) of est female participation rates in ownership are Al- firms. In this case, unless majority shareholders bania, Azerbaijan, and Kosovo (11 percent). are woman, it is unlikely that, despite participa- Women’s participation in firm ownership var- tion in ownership on a broad scale, women exert ies greatly by type of enterprise. The type of firm control over the direction of companies or obtain matters significantly in indicating the degree of significant shares of the income streams from the control that owners have over the day-to-day op- firms. (However, the data are insufficient to iden- erations or even the strategic direction of firms. tify majority owners in shareholding companies Firms with multiple owners, such as public or partnerships.) shareholding companies, are likely to have di- luted ownership. The ownership arrangements of sole proprietorships are clearer in terms of de- Women in Top Management termining the role of the owners and the gender Though senior managers may not be entrepre- of the owners. In sole proprietorships in Europe neurs, they may have significant control over the and Central Asia, about one in three owners is direction of firms. Senior managers are respon- a woman, which appears to be largely consistent sible for key decisions affecting not only the daily Figure 3.2 Women’s Participation in Firm Ownership a. Regional comparison of women in enterprises b. Women’s ownership by type of �rm % of �rms with female owners % of �rms with female owners 60 60 53 51 52 50 50 48 40 39 40 38 36 36 34 30 30 30 20 18 20 16 10 10 0 0 ECA EAP LAC MNA SAS SSA Public Sh. Private Sh. Sole Prop. Partnership Ltd. Part Other Sources: Business environment and enterprise Performance Survey, european Bank for reconstruction and Development and World Bank, Lon- don, http://www.ebrd.com/pages/research/analysis/surveys/beeps.shtml; enterprise Surveys (database), World Bank, Washington, DC, http:// www.enterprisesurveys.org/. Note: Data are for 2009 or the latest available year. eAP = east Asia and the Pacific. eCA = europe and Central Asia. LAC = Latin America and the Caribbean. MNA = Middle east and North Africa. SAS = South Asia. SSA = Sub-Saharan Africa. Public Sh. = public shareholders. Private Sh. = private shareholders. Sole Prop. = sole proprietorship. Ltd Part = limited partnership. Women in Entrepreneurship 67 operations of businesses, but also long-term stra- and become critical decision makers within firms. tegic development and profitability. An analysis From the low representation of women in top of women in top management is likely to provide management positions, it appears that the large a more accurate picture of the degree to which pool of well-educated women, especially those women have a say in firms because, in many who have specialized in the social sciences, law, cases, firm ownership may be diluted, or women and business, are underutilized in Europe and (such as wives and daughters) may be owners, Central Asia. but effectively silent. Thus, managers may often The level of variation in women’s participation in have greater influence than nonmajority owners senior management positions in enterprises is high in of firms. This subsection discusses the share of the region, ranging from 0.3 to 31 percent of all firms. women in top management in formal sector firms The countries with the highest proportion of firms with at least five employees, as well as the sig- with top managers who are women are Latvia and nificant correlations between female ownership, Poland, at 31 percent. At the other extreme, only 5 management, and employment in the region and and 0.3 percent of firms in Azerbaijan and Kosovo, in specific countries. respectively, have women in top management posi- A woman is a top manager in about one in five tions. The subregion with the highest proportion of firms in the region. Women hold top management firms with top managers who are women is the 10 positions in an estimated 19 percent of firms in member states of the European Union (EU) that Europe and Central Asia (figure 3.3). Data on oth- are in the region (EU10) (22 percent), followed er regions are far less complete; not all countries by the middle-income countries of the Common- report information on top managers. The propor- wealth of Independent States, at 19 percent. There tion of firms with top managers who are women is also a relatively strong correlation (0.6) between was 27 percent in East Asia. As in other regions the proportion of firms with female participation in of the world, women in Europe and Central Asia ownership and the proportion of firms with wom- are much less likely than men to obtain seniority en in top management. Figure 3.3 Women’s Participation in Top Management a. By regions b. Selected countries % of �rms % of �rms 30 35 27 25 30 19 25 20 17 18 20 15 15 10 10 5 5 5 0 0 ECA EAP MNA SAS SSA AZE TJK TUR RUS GEO KGZ BGR EST UKR POL LVA Sources: Business environment and enterprise Performance Survey, european Bank for reconstruction and Development and World Bank, London, http://www.ebrd.com/pages/research/analysis/surveys/beeps.shtml; enterprise Surveys (database), enterprise Analysis unit, World Bank, Washington, DC, http://www.enterprisesurveys.org/. Note: Data are for 2009 or the latest available year. The average for east Asia (eAP) includes indonesia, Malaysia, Mongolia, the Philippines, and Vietnam. The average for europe and Central Asia (eCA) is based on data on 29 countries. The average for Middle east and North Africa (MNA) includes Lebanon, Syrian Arab republic, and republic of Yemen. The average for South Asia (SAS) is for Afghanistan, Bangladesh, and Nepal. The average for Sub-Saharan Africa (SSA) is for Angola, Benin, Botswana, Cape Verde, Democratic republic of Congo, republic of Congo, eritrea, gabon, Lesotho, Liberia, Mali, Malawi, and Togo. 68 Opportunities for Men and Women: Emerging Europe and Central Asia Women Entrepreneurs and the tively balanced participation across firms by size ap- Characteristics of Their Firms pears to indicate that women are not being excluded from either small or large firms. The characteristics of firms of women entrepre- However, among sole proprietorships and family- neurs and the firms of men entrepreneurs are dif- owned firms, woman-owned firms have fewer employ- ferent. This is true whether the women are owners ees. In the region, about 65 percent of all registered (sole or joint) or top managers. Firm characteristics enterprises owned by women employ fewer than vary by sector, number of employees, and domestic 10 permanent employees, in contrast with 56 per- or foreign ownership. cent among enterprises owned by men (figure 3.5). By their nature, the ownership in these types of firms is not diluted. Thus, women owners are much Women Owners in the Formal Sector more likely to play an important decision-making Small, medium, and large firms in the region show ap- role in these firms than in general. However, the proximately the same level of participation of women in predominance of women in smaller firms is diffi- ownership. On average in the region, women’s partic- cult to ascertain. This may be caused by many fac- ipation in ownership ranges from 36 to 39 percent tors, ranging from the particular sector of activity in firms by size (though with variation across coun- of women’s firms to the ability of women to gain tries) (figure 3.4); this is comparable with the results financing to expand their businesses and even in Latin America and the Caribbean. However, there women’s business preferences. However, this char- are large variations among female participation rates; acteristic of woman-owned firms is not unique; it is the greatest variation across firms occurs among also evident in developed countries.71 large firms, where the participation rates range from 4 percent in Montenegro to 64 percent in the former Yugoslav Republic of Macedonia. Within countries, 71 This finding is based on data for 2005 from the Busi- ness Environment and Enterprise Performance Survey, women’s participation rates by firm size are closely European Bank for Reconstruction and Development correlated, that is, if women show a high participa- and World Bank, London, http://www.ebrd.com/ tion in large firms, they are also likely to show a high pages/research/analysis/surveys/beeps.shtml; see also participation in medium and small firms. This rela- Sabarwal and Terrell (2008). Figure 3.4 Women’s Ownership of Firms by Firm Characteristics, 2008 a. Women’s ownership, management by �rm size b. By sector and �rm type % of all �rms % of all �rms 40 60 35 50 30 25 40 20 30 15 20 10 5 10 0 0 Small Medium Large Plastics Electronics IT Construction Retail Garments Owner Sole Prop. Top Manager All Female Source: Business environment and enterprise Performance Survey, european Bank for reconstruction and Development and World Bank, London, http://www.ebrd.com/pages/research/analysis/surveys/beeps.shtml. Note: iT = information technology. Sole Prop. = sole proprietorship. Women in Entrepreneurship 69 Figure 3.5 Sole Proprietorships and Firm Characteristics, 2008 a. By �rm size and owner’s gender b. By sector and �rm type all �rms, by gender, % % of full-time employees who are women 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 <10 10–50 51–100 101–500 >500 Manufacturing Services Other Female-owned �rms Male-owned �rms Female-owned �rms Male-owned �rms Source: Business environment and enterprise Performance Survey, european Bank for reconstruction and Development and World Bank, London, http://www.ebrd.com/pages/research/analysis/surveys/beeps.shtml. The participation of women and men in firm and greater female participation in the private sec- ownership in the region is concentrated in different sec- tor labor market may be correlated. tors. Women participate in the ownership of over half (55 percent) of the registered firms with 10 employees or more in the textile and garment in- Women in Top Management dustry, compared with only a quarter of the firms Women are more likely to achieve top management po- in basic metals, transport, and electronics.72 This sitions in small firms than in large and medium firms variation grows significantly among sole proprietor- in Europe and Central Asia and in most other regions. ships, among which some sectors show a negligible About one in five top managers of small firms in share of women owners (such as in basic metals or Europe and Central Asia is a woman. However, in plastics), while others show nearly equal represen- general, female participation in top management in tation (garments and retail). In some higher-pro- small firms is not a good predictor of female par- ductivity sectors, women have a strong presence. ticipation in medium or large firms in the region. For example, among sole proprietorships in infor- Nonetheless, there is great variation across the mation technology and electronics, respectively, 33 region. Thus, there are almost no women in top and 40 percent are owned by women. management positions in Kosovo, while, in Latvia, The employment rate of women is higher in sole almost one in three top managers is a women irre- proprietorships owned by women. Women propri- spective of firm size. Europe and Central Asia does etorships hire, on average, more women than men not perform particularly well in this indicator rela- in some sectors. However, sole proprietorships tive to other regions of the world. Women are in owned by men usually hire relatively fewer women top management positions in 33 and 22 percent of than men for full-time positions. This is true irre- spective of the sector (see figure 3.5). It is not clear whether this arises because of a preference on the 72 The full sample is weighted. The data are for 2008 and have been taken from the Business Environment and part of owners or some other characteristics of the Enterprise Performance Survey, European Bank for firms and employees that are not readily apparent. Reconstruction and Development and World Bank, However, irrespective of the underlying reasons, London, http://www.ebrd.com/pages/research/anal- this appears to suggest that ownership by women ysis/surveys/beeps.shtml. 70 Opportunities for Men and Women: Emerging Europe and Central Asia small and medium firms in East Asia and the Pacific Constraints on Women’s and in Latin America and the Caribbean, respec- Entrepreneurship tively, which is significantly higher than the share in Europe and Central Asia. In any case, it appears Men and women in the region face constraints in that, across the globe, women face obstacles in ob- starting and growing businesses partly because of taining senior management positions in large firms. the lingering effects of the transition process and, There are more women top managers in sectors in some instances, because of an incomplete tran- in which women owners are more prevalent. Women sition. However, across the region, several con- are more likely to be top managers in the garment straints, such as lack of access to credit, lack of net- industry, hotels and restaurants, and textiles than in works, and barriers in the business environment, any other sectors. The share of top managers who appear to affect men and women business own- are women is around 30–40 percent in these sectors. ers differently. In this section, we discuss the dif- The share of top managers who are women is only 6 ficulties women face in accessing finance, selected percent in the basic metals, machinery, and equip- regulatory obstacles, and women’s lack of access to ment industries. The significant variation in female networks. participation in top management mirrors to a greater degree the situation in female ownership discussed above. This correlation between the proportion of Access to finance women owners and women managers is strong, but Women in the region are less likely than men to obtain it is unclear why this occurs and whether it reflects bank financing, and they are more likely to pay higher the similar pattern seen where sole proprietorships interest rates if they do obtain bank financing. Access led by women are more likely to hire women relative to bank finance is thus a more important barrier to sole proprietorships owned by men. It may indi- for women business owners than for their male cate the existence in business of the occupational counterparts (Aidis et al. 2007; Sabarwal and Ter- segregation commonly seen in labor markets. rell 2008). In the formal sector in the region, wom- Women are more likely to obtain senior manage- an-managed firms with more than five employees ment positions in small firms that are in the service sec- have a 5.4 percent lower probability of securing a tor and in which at least one of the owners is a wom- bank loan relative to man-managed firms. Woman- an.73 Women are more frequently top managers in managed firms, on average, pay 0.6 percent more in small retail firms and in firms where at least one interest rates relative to their man-managed firms.74 of the owners is a woman. In countries on which In the next section, we discuss the effects of this on relevant data are available, women are, on average, business choices among women and on firm profit- three times more likely to be top managers in small ability. firms than in large firms (for example, Armenia, The lack of physical and financial assets penal- Georgia, and Ukraine). Similarly, women are more izes women in credit markets. Factors such as less well frequently top managers in retail firms. More than developed financial systems, lower collateral, and half the firms in which at least one of the owners is lack of credit histories appear to constrain women a woman are also run by women; the correspond- ing share in firms in which no women are owners is 4 percent (data on Armenia and Ukraine). Women 73 Here, firm size levels are 5–19 employees (small), are far less likely to hold top management positions 20–99 employees (medium), and 100+ employees in manufacturing than in services (data on Armenia (large). and Turkey), while, in the service sector (excluding 74 See Muravyev, Talavera, and Schäfer (2009), based on data of the Business Environment and Enterprise Per- retail), three times as many firms have top manag- formance Survey, European Bank for Reconstruction ers who are women. Women managers are also, on and Development and World Bank, London, http:// average, younger and less experienced than their www.ebrd.com/pages/research/analysis/surveys/ men counterparts (data on Uzbekistan). beeps.shtml. Women in Entrepreneurship 71 more than men. Evidence on the existence of dis- ilar reasons, MI-Bospo, a women-only microfinance crimination against women in interest rates and in institution (MFI) in Bosnia and Herzegovina, also access to finance is found more readily in the least requires the signature of husbands as a guarantee financially developed countries in the region (Mu- for loans to women who lack collateral. ravyev, Talavera, and Schäfer 2009). Because banks Women may be less inclined than men to pursue typically lend on the basis of hard assets, such as loans in some cases. Less access to loans may also be plant and equipment, of which service businesses explained as demand-side behavior caused, for ex- have few, women are less likely to obtain loans and ample, by fewer applications to banks by women are thus capital constrained and must operate at who may be more risk averse (Jianakoplos and Ber- a lower scale (Sabarwal and Terrell 2008; see box nasek 1998; Barber and Odean 2001; Dohmen et 3.2 for an effort to alleviate this problem in India). al. 2005). These differences may have important Women are also disadvantaged in collateral-based implications for business performance if higher risk lending because, in the process of privatization, pri- aversion leads women to restrict their investment in vate property has been disproportionately acquired business ventures. The 2005 Global Entrepreneur- by men, and, across most of Eastern Europe, new ship Monitor found that fear of business failure is property owners have been mostly men (Paci 2002; significantly greater among women than men in Lastarria-Cornhiel 2009).75 A study based on Living middle-income countries (Minniti et al. 2006). In Standards and Measurement Study analysis in Ta- addition, Kepler and Shane (2007) find that nascent jikistan finds that, for long-term loans, women are male entrepreneurs examine more ideas and gather charged a 16 percent interest rate, compared with 4 more information while pursuing start-ups than percent for men, because women are assumed to be do female nascent entrepreneurs. They report that less creditworthy given their significantly lower rate firms are adjusting their input to reflect the capital of ownership of land, assets, and livestock and their constraint and that the capital constraint is partially overall lower employment ratios and wages (World responsible for the smaller scale of women’s busi- Bank 2009). In Tajikistan, banks also require the nesses (see also Sabarwal and Terrell 2008). signatures of husbands for loans to women. For sim- MFIs provide women an opportunity to build a credit history, though this resource is underexploited. Overindebtedness and the proliferation of nonper- forming loans in the microfinance sector following Box 3.2: Encouraging Property Ownership through the financial crisis have underscored the need for Stamp Duty Reduction more credit bureaus in the region. Women com- India’s 11th Plan encourages ownership rights for women by prise a large number of microfinance clients and offering incentives for ownership of property in women’s name. have excellent repayment rates; however, despite Women home buyers benefit from tax exemptions, lower stamp duties and easier availability of home loans. A lower stamp duty this, they fail to build credit histories or the repu- rate helps in saving on the overall costs while purchasing prop- tation collateral that could help them graduate to erty, thus acting as a significant boost for prospective women buyers. bank loans or larger loans given that MFIs in many Such is the increase of prospective women buyers that countries in Europe and Central Asia, similar to developers are also considering incentives aimed at women. State and local governments in Uttar Pradesh, Delhi, Orissa, and Punjab have launched some initiatives in this regard. For example, in 2002, the state of Delhi cut stamp duty rates from 75 In Bosnia and Herzegovina, despite laws on equality, 8 to 6 percent for women owners. In case of joint ownership by men and women, the duty is 7 percent. Using the opportunity only 6.5 percent of women respondents to a survey that India’s favorable macroenvironment provided, MannDeshi reported owning land in their own names (STAR Net- Bank—a women’s cooperative bank in the state of Maharash- work of World Learning, cited in IFC and MI-Bospo tra—has advocated for stamp duty reduction for joint property 2008). In Montenegro, only 3 percent of property is registration for women borrowers. This bank also honors and registered in women’s names, while the corresponding rewards husbands that undertake such joint registrations. figure in rural parts of Kosovo is 10 percent. Moreover, Source: Narain (2009). 40 percent of women noted that the lack of property makes it difficult to gain access to financing. 72 Opportunities for Men and Women: Emerging Europe and Central Asia MFIs globally, do not participate in credit bureaus turnover among enterprises, as well as significant and credit registries. Currently, in 10 countries competition, thus rendering any business growth in Europe and Central Asia, MFIs provide infor- in these sectors difficult (Welter and Kolb 2006). mation to credit bureaus and credit registries.76 Moreover, the sectors and the nature of the busi- Increasing the access of women to MFIs would nesses in which women are concentrated, such as benefit nearly 2 million women borrowers in the hotel and catering services, require more inspec- region.77 Though the World Bank collects data on tions (for example, sanitation, food quality, and the share of the adult population covered by public health), and the cost of these inspections may thus and private credit registries, there is no gender-dis- be disproportionately high for the smaller business- aggregated information. However, in five countries es that, in Europe and Central Asia, show significant (Croatia, the Czech Republic, Lithuania, Poland, female participation. Nonetheless, this also means and Serbia), over two-thirds of the population is that the business environment is generally deficient. covered, which indicates that there is likely to be Access to swift, affordable, and effective justice significant coverage among women. systems and dispute resolution mechanisms such as arbitration is important for men and women entrepre- neurs involved in small enterprises; these mechanisms Other Regulatory Obstacles are lacking in many countries in Europe and Central Across the region, bureaucratic hurdles to invest- ment, such as the registration of business activities, obtaining permits, dealing with tax inspections, 76 The 10 countries are Armenia, Bosnia and Herzegov- and demands for bribes, constrain women’s busi- ina, Bulgaria, Georgia, Kazakhstan, the Kyrgyz Repub- nesses more severely than men’s businesses. This is, lic, Mongolia, Montenegro, Russia, and Serbia. See however, also a reflection of the deficiencies in the Gender Law Library (database), World Bank, Wash- business environment and disadvantages in the size ington, DC, http://wbl.worldbank.org/WBLLibrary/ and the sectors of the businesses in which women elibrary.aspx?libid=17. 77 Gender Law Library (database), World Bank, Wash- more typically work. ington, DC, http://wbl.worldbank.org/WBLLibrary/ Women who head businesses report more often elibrary.aspx?libid=17, based on the MFIs reporting than men who head businesses that they face obstacles to Mix Market (database), Microfinance Information in finance, taxation, inspections, registration, permits, Exchange, Washington, DC, http://www.mixmarket. and so on.78 For example, the time needed to regis- org/data-center. ter a business is reportedly higher for women than 78 See the previous note. 79 In Tajikistan, relative to men, women who own incor- men in Georgia and Tajikistan.79 Similarly, relative porated businesses need an average of 10 more days to to men, more women in Tajikistan cite tax admin- register their companies. istration as an obstacle. In Ukraine and Uzbekistan, 80 Other studies show that the impact of investment cli- inspections are more of a problem for women than mate barriers in Europe and Central Asia differs con- men. In Belarus and Ukraine, women must make siderably by gender. Thus, Aidis et al. (2007) find that, unofficial payments more often.80 Do women en- in Ukraine, while both men and women entrepreneurs identify tax issues as a major constraint, regulations and trepreneurs face more severe constraints? The lim- legal barriers were identified by women more often. ited available evidence only captures the problems Women also reported more frequently that they were encountered by registered firms. Thus, information constrained by the lack of capital. In Lithuania, high on the obstacles women meet in trying to start or taxes were identified as a barrier by all entrepreneurs formally establish businesses is not at hand. regardless of gender. However, low purchasing power, While more research is required, the con- which is directly related to low consumer demand, was identified as the second most important constraint by straints blocking women entrepreneurs should also women entrepreneurs. The third most important bar- be examined in light of the sectors in which women rier cited by women entrepreneurs was lack of financ- start and try to grow their businesses. Thus, the ing for business investment, though this did not figure trade and service sectors are characterized by high among the top three barriers reported by men. Women in Entrepreneurship 73 Asia. Because women entrepreneurs in the region neurship. This section explores gender differences are concentrated in small businesses and face great- in entrepreneurial performance in terms of scale, er constraints in accessing formal legal structures profit, and total factor productivity. It is based on or courts, affordable alternative dispute resolution data of the Business Environment and Enterprise mechanisms may benefit them disproportionately. Performance Survey in 2005 and includes only the Though a third of the countries in the region have sample of firms in which women are the sole pro- created small claims courts, more initiatives, such prietors of individual firms or are the proprietors of as mediation and arbitration, as well as judicial cost family-owned firms.81 reduction, the simplification of claims procedures, Enterprises in Europe and Central Asia that are and improvements in access to justice, would likely owned and managed by women are small in terms of aid in combating late payments and resolving dis- sales revenue. However, women generate more prof- putes in commercial transactions, which can affect it per unit of sales revenue and show higher returns the survival of small businesses. to scale. Firm-level data on 26 countries show that, while the companies of men and women entrepre- neurs in the region are suboptimally small, women’s Access to Networks returns to scale are significantly larger than men’s, Business networks and associations are largely male implying that women would gain more if the scale dominated, and women are disadvantaged. Women’s were increased. lack of access to networks and business associations On average, women owned businesses underper- deprives them of information, resources, and voice. form in sales revenue by 6 percent relative to men. How- Networks and business associations play an im- ever, this difference is driven by variation in the size portant role in support, training, information man- of businesses (there is no such difference in larger agement, and access to mentors, role models, and businesses), and the gender efficiency gap in non–EU resources, as well as forums for advocacy. Women countries is larger than the corresponding gap in EU entrepreneurs in Europe and Central Asia are disad- countries. The difference in performance is driven by vantaged because formal and informal networks and the small scale of women’s businesses. The subopti- business development services are largely dominat- mal size of woman-owned firms arises because these ed by men. Such male-dominated networks led dur- firms are capital constrained and are concentrated ing Soviet times to greater gains in the early priva- in industries characterized by small firms. Internal tization process in transition economies. Even in financing and bank financing appear to be the two countries where governments have taken proactive most important sources of financing; it also appears steps to support women entrepreneurs (for exam- that, relative to men, women rely more on internal ple, in Slovenia), women have been obliged to access financing and less on banks (Sabarwal and Terrell resources through family and spouses in the absence 2008). More recent data of the Business Environ- of formal support structures (Drnovsek and Glas ment and Enterprise Performance Survey (2008–09) 2006). Moreover, the lack of access to networks de- confirm this trend over time. Thus, women’s enter- prives women of voice in ongoing reform processes prises (for example, in Azerbaijan, Belarus, Georgia, and in efforts to make the business environment and Tajikistan) exhibit a much lower performance work for them by mainstreaming their concerns. level in annual turnover or in annual profits relative to men’s enterprises. These findings are consistent Do Woman-Owned Firms Perform Comparably with Man-Owned Firms? 81 See Business Environment and Enterprise Perfor- mance Survey, European Bank for Reconstruction and If women face different conditions and institu- Development and World Bank, London, http://www. tional constraints relative to men, the effects can ebrd.com/pages/research/analysis/surveys/beeps. spill over into women’s performance in entrepre- shtml. 74 Opportunities for Men and Women: Emerging Europe and Central Asia with the conclusions of individual country studies: ence in firm profitability between men and women women entrepreneurs run businesses that are small- (Sabarwal and Terrell 2008). However, questions er in terms of sales and number of employees. remain: Why do women concentrate their business- Women entrepreneurs in the region generate the es in certain sectors (services rather than manufac- same amount of profit per unit of revenue as men en- turing)? Does sector concentration or the choice of trepreneurs. The performance of woman-owned activity reflect capital constraints at start-up? Is the firms in terms of total factor productivity is signifi- capital constraint self-imposed or does it reflect bias cantly more limited relative to man-owned firms in the financial system toward women entrepre- in the same industries. However, the average gap, neurs who lack collateral or a track record? (Boxes at −2.1 percent, is small and may not be economi- 3.3 and 3.4 offers examples of approaches to miti- cally significant. The overall finding that there are gate the financing problem at start-up.) gender differences in scale and profits is the same throughout the region; however, the gender gap in total factor productivity is significant only in the non–EU countries, not in the EU countries. Our Box 3.3: Equity Funds report finds that women and men operate busi- Lack of start-up finance is a constraint for women in both start- nesses at an inefficiently small scale in both of these ing and growing businesses. The following are two examples of equity funds that support women’s start-ups. sets of countries, but that there is evidence women Trapezia, an equity fund for women in the United Kingdom, in non–EU countries fare better than men in access is a private investment fund and is a venture capital and busi- to finance.82 We also find that significant gender ness angel for women entrepreneurs. It offers women inves- tors the opportunity to invest in women-focused businesses gaps exist only in the scale of firm operations and over a three- to five-year period. Its investment advisory panel not, as others have argued, in firm profitability or includes representatives from the Women in Business Unit of the Bank of Scotland and other seasoned entrepreneurs and efficiency. This supports the argument that, if the professionals (IFC 2006). existing capital constraints are lifted, women’s busi- The Women Private Equity Fund in South Africa was es- nesses are likely to expand at least as much as the tablished in 2003. The fund provides expansion capital to com- panies that are controlled or managed by women, employ a businesses of their male counterparts. Our report majority of women, or have a market focus on women. The strongly indicates that more research is needed to fund targets investments ranging from R5 million to R19 mil- lion. It also assists companies in strengthening their strategic explain the differences in firm performance be- focus. After three to five years of investment, the fund exits the tween EU and non–EU countries. Also, evidence investment in different ways, for example, through an initial suggests that policy in Europe and Central Asia public offering on the stock exchange, trade sales, international exits, or management buybacks (UN 2009b). needs to be geared toward easing the capital con- straints on women entrepreneurs and encouraging Source: Narain (2009). more women to start businesses in industries with greater growth potential. Is the smaller scale of women’s businesses in the re- gion caused by risk aversion among women? The avail- 82 There are studies showing that women entrepreneurs able evidence suggests that women entrepreneurs face higher barriers and pay higher interest rates than their men counterparts. For example, in Germany, in the region increase the scale of their businesses while both men and women have difficulty in access- if they have access to sufficient capital and that ing external finance, Constant (2006) points out women implement such scale increases at the same that women often have difficulties convincing finan- rate as men entrepreneurs in the region. This runs cial institutions and potential clients that they have against the perception that women’s businesses are strong management abilities and sound business ideas. smaller and less profitable because women are risk Among other European countries, a recent empirical study by Alesina, Lotti, and Mistrulli (2009) finds that averse or because they do not want to grow their self-employed women pay more than men in Italy for businesses (Jianakoplos and Bernasek 1998; Barber overdraft facilities and that this difference does not and Odean 2001; Dohmen et al. 2005). This report seem to be explained by any variable capturing differ- finds, however, that there is no significant differ- ential risk, directly or indirectly. Women in Entrepreneurship 75 to be smaller firms or microenterprises and often Box 3.4: Capital Seed Program in Chile focus on the shuttle trade. More support for wom- Capital Seed (Capital Semilla) is a program to distribute funds en would help women entrepreneurs in the region with the objective of fostering the creation of new firms and become more confident and positive and less risk strengthening existing microenterprises and small enterprises. averse and would also enable the region to lever- It is part of the Servicio de Cooperación Técnica (technical as- sistance service), a government office within the Ministry for age women’s full potential to assist in building sus- the Economy, Development, and Tourism. The program targets tainable and competitive economies. This section microentrepreneurs whose yearly sales are below UF 25,000, who pay taxes, and who have not carried on activities for more outlines polices and strategy directions to achieve than one year. these goals. The program provides start-up capital of between In addition to increasing women’s participation US$2,162 and US$4,325, of which an additional 30 percent comes from the entrepreneur in cash. The applicant goes in entrepreneurship (woman-owned businesses account through a competitive process that begins with an online self- for a mere 10th of all incorporated enterprises in the diagnosis of the entrepreneur’s skills and the economic activity of the firm or proposed firm. According to the result, the entre- region), there is also a need to support, grow, and diver- preneur can directly apply for funds by presenting a business sify women’s businesses. Our analysis has shown that plan or must submit to a conditional application in a business woman-owned firms, though often smaller, are as training program. If the applicant fails the diagnosis, the ap- plicant is eliminated from the process. The application process efficient as the firms owned by men. is open yearly during the month of March. However, in Europe and Central Asia, not The business plans are carefully evaluated by a jury from the Servicio de Cooperación Técnica according to criteria stat- only is the overall gender gap in entrepreneurship ed in the application forms. The jury grants the capital accord- significant, but a large number of women are also ing to the score assigned to the business plan and according involved in the shuttle trade or in microenterprises to regional criteria. During the time the entrepreneur uses the funds, the Servi- that face low entry barriers. Many women oper- cio de Cooperación Técnica provides technical support through ate their businesses from their homes or in street a consulting firm. In 2009, 650 projects were financed. In 2010, 879 entre- markets or can afford to engage in self-employment preneurs applied for the program, of which 178 were micro- only part time. In addition to increasing women’s firms, which were evenly distributed between men and women participation rates, it is thus equally important to managers, and 701 were entrepreneurs. More men (449) ap- plied than women (282). address the specific constraints women entrepre- neurs encounter in growing their businesses (see Source: http://www.sercotec.cl. box 3.5 for an example from Afghanistan). It is significant that, in the region, the share of women The Implications for Policy Design Box 3.5: Training Women in Nontraditional Sectors Women entrepreneurs play an important role in the private sector of the transition economies. Under an initiative led by Italian Cooperation, Pashtun women in low-income families in Kabul have been trained in nontra- However, they face significant constraints in start- ditional businesses (that had hitherto been a preserve of men) ing and growing businesses. There is an urgent need such as gem-cutting, the repair of mobile phones, and cater- for governments in Europe and Central Asia to rec- ing. Many of the trainees have graduated to work as caterers, lantern makers, mobile repairers, and gem-cutters. A group of ognize the potential of women entrepreneurs and trained gem-cutters has subsequently established Sultan Razia to address the constraints women entrepreneurs Gem Cutting Co., an enterprise in Kabul. A program evaluation carried out by the United States Agency for International Devel- face. Private sector and nonprofit organizations, opment of the company’s business model finds that it is a sound donors, and public-private partnerships should be basis for rapid expansion (Bowersox et al. 2007). Convincing the part of the solution. The current economic crisis women to undergo the training was not easy, however. Italian Cooperation first had to win over the shura (a consultative body) has highlighted the need to support women, who to allow women to train with them. As an additional guarantee, constitute at least half of the human capital and a a shura member was hired as an advocate. third of the private sector in the region and who are Source: Narain (2006). also more vulnerable because their businesses tend 76 Opportunities for Men and Women: Emerging Europe and Central Asia entrepreneurs who desire to grow their businesses yet function well (Welter et al. 2006). Informal is relatively high (Welter et al. 2006). networks help entrepreneurs in gaining access to Policy initiatives to support sectors in which resources and information and in voicing their woman-owned businesses are concentrated will ben- concerns. Women’s lack of access to networks efit women disproportionately. However, there is also therefore represents a disadvantage for them. This a need to support women’s participation in nontra- should be addressed through the development of ditional sectors. Linking women entrepreneurs to formal structures, such as business associations business development services and mentors and and business support mechanisms that are sensi- increasing women’s access to relevant technical tive to the needs of women entrepreneurs (box 3.6 education is an imperative for raising the competi- offers examples). Supporting women’s presence tiveness of women in the region. While initiatives in important business entities such as chambers exist in the region to support small and medium of commerce is also important for mainstreaming enterprises, such services are currently lacking for women’s concerns. A EuroChambres (2004) sur- microenterprises and small enterprises and for vey of 25 member countries recently found that a women traders (Aidis 2006; Rankin and Narain chamber of commerce in only one of these coun- 2009). Even in training programs for small and tries (Slovenia) had a woman president. medium enterprises, the region, with an average Improving the overall business environment will index score of 2.83 on a scale of 5.00, ranks be- help both men and women, but especially women, who low the high-income countries (4.36) and Latin face additional constraints. The countries in the re- America and the Caribbean (2.94).83 Furthermore, gion, especially countries in the early stages of mar- women’s concentration in small trade, small busi- ket development, are burdened with corruption, nesses, and microenterprises would need to be inefficient judicial systems, and disregard for rules addressed to enable women to take advantage of at the local level. These constraints affect particu- emerging opportunities by increasing their access larly women. In addition to rationalizing tax poli- to global supply chains. Women faced with fluctu- cies and tax administration and reducing the barri- ating domestic demand would particularly benefit ers to the establishment of formal sector businesses from such links. through improved, simplified business registration There is a need to offset the revival of patriar- processes, it will also be important to address is- chal traditions in the region by undertaking efforts sues such as building capacity, the sensitization of to support women’s role in society. The lack of self- confidence and the fear of failure among women entrepreneurs can be linked to the perception in the region of women’s role as housewives and of Box 3.6: Networking for Success: Examples of entrepreneurship as a male domain (Minniti, Al- Women’s Business Associations len, and Langowitz 2006; Allen et al. 2008). Coun- The National Association of Women Business Owners repre- terbalancing these negative social attitudes can be sents over 9,000 women-owned businesses organized into 80 achieved by creating a positive image of women en- chapters throughout the United States. It creates networking trepreneurship through media outreach, the sensi- opportunities for its members, sponsors nationwide confer- ences and public policy days, and provides tools for growing a tization of policy makers, and support for network- business. (For more information, go to http://www.nawbo.org/.) ing and mentoring among women entrepreneurs. The Jordan Forum for Business and Professional Women, which was established in 1976, provides a platform for devel- Women entrepreneurs in the region lack access opment, empowerment, and advocacy for women in business to networks; this represents a disadvantage among in Jordan. women in accessing resources, information, and ad- Source: iFC and Mi-Bospo (2008). vise and in lobbying for reform. Networks are es- pecially important for women entrepreneurs in a transition context because the formal institutional framework in many of these economies does not 83 Ranking of the Economist Economic Intelligence Unit. Women in Entrepreneurship 77 officials to the problems of women entrepreneurs, Box 3.8: A Small and Medium Enterprise Finance and corruption. Concerted efforts should also be Facility for Women made to ensure that any revised policies and regu- lations are widely disseminated and publicized so The European Bank for Reconstruction and Development signed a €50 million loan facility with Garanti Bank of Turkey as to reach women entrepreneurs, many of whom, to enable the Turkish bank to expand its portfolio of loans to for example, do not have access to the Internet microenterprises and small and medium enterprises in the for communication. Furthermore, because many economically less developed east and southeast regions of Turkey and for specific sectors and areas, namely, agriculture women in the region cite inspections as a major and women entrepreneurs. The objective of the program is to constraint, a risk-based approach in inspections is increase the access to finance among business women and the access of women to training and to raise the awareness of warranted, along with more clarity and transpar- this financial market. Since the start of the program in 2007, ency in inspection visits. Garanti Bank has lent US$157 million to 8,400 women, trained Women entrepreneurs in the region face signifi- approximately 1,600 participants, and received over 3,000 ap- plications for the bank-sponsored Turkey’s Woman Entrepre- cant challenges in accessing finance. Initiatives are neur of the Year contest. needed to target women entrepreneurs and address Source: iFC (2010). their needs beyond microfinance. While important, microfinance is clearly not sufficient to meet the needs of growth-oriented women entrepreneurs who need access to a variety of financial products, services, and service providers. (See box 3.7 for an services. A survey of Internet use by men and example of a successful program.) women entrepreneurs in Cyprus shows that not Recent government and donor initiatives in only do more men (74 percent) than women (38 the region are significant in this regard (annexes B percent) use computers in their businesses, but and C). Among donor-supported initiatives, recent that men also use the computers in more ways credit lines of the European Bank for Reconstruc- (such as in design, modeling, and financial state- tion and Development to commercial banks that ments) (Nearchou-Ellinas and Kountouris 2004). are aimed at women entrepreneurs are particularly Of the women respondents in Cyprus, 88 percent important (box 3.8). do not use the Internet at all. Similarly, a study Boosting women’s access to information and has found a statistically significant difference in communication technology will benefit women entre- the use of Internet, e-mail, and broadband by men preneurs in the region who have less access to such and women entrepreneurs in Tajikistan (World Bank 2009a). The Cyprus survey also finds that 96 percent of woman-owned enterprises are ac- tive only in their local communities. Deploying Box 3.7: Finance for the Missing Middle information and communication technology for e-commerce could help expand women’s business The Growth-Oriented Women’s Enterprises Programme, a part of the African Development Bank’s larger African Women in activities beyond the company premises. Business Business Initiative, is an attempt to address the constraints incubators are promising in this regard. Business faced by business women. The program is based on lessons incubator initiatives in Georgia, Mongolia, and from enterprise development research in Africa by the Inter- national Labour Organization and the African Development Slovenia, a part of InfoDev, a World Bank–sup- Bank in Africa. It is a comprehensive program that responds ported global development financing program, to women’s constraints such as lack of property rights and the lack of collateral, lack of adequate business financial records, have been actively increasing awareness and seek- limited capacity to prepare a business plan, and the high-risk ing to enhance the participation of women en- perception about women by banks. The program thus supports trepreneurs in incubator initiatives, including in women in growing their businesses through all three compo- nents: finance, training, and mentoring. Europe and Central Asia. Gender-disaggregated data and analysis are Source: iFC (2006). important in designing appropriate policy responses 78 Opportunities for Men and Women: Emerging Europe and Central Asia and in tracking what works and what does not work. Box 3.9: A Government-Led Initiative in India to Track Gender-disaggregated data are needed to monitor Gender-Disaggregated Data on Access to the growth of women’s enterprises and women’s Finance access to finance, as well as to determine whether policy interventions are working or not. A joint The Indian government has drawn up an ambitious 14-point action plan for public sector banks to increase women’s ac- United Nations Economic Commission for Eu- cess to bank finance. The government instructed the central rope–World Bank initiative in Europe and Central bank to maintain a database to track performance during the effort. Following the issuance of the government directive, the Asia aimed at increasing the capacity of national Reserve Bank of India, in 2000, asked public sector banks to statistical agencies to collect such data is significant disaggregate and report on the percentage of the credit going in this regard (UNECE and World Bank Institute to women within their total lending. The government’s action plan has set a target of increasing such loans from the 2001 2010). The same two organizations have also re- level of 2.4 percent to 5.0 percent of total lending. The data are cently launched training modules for policy makers reported annually in the Reserve Bank of India’s Trends and Progress Report. The aggregate net bank credit to women had on the collection of such data on access to finance increased to 6.3 percent by 2009, and 25 banks reached the and on entrepreneurship.84 (See box 3.9 for an ex- target. Though an assessment of the full impact of the policy ample from India.) requires further exploration, the data tracking has increased the general awareness about women’s low access levels. Source: Narain (2009). 84 See the Genderstats modules at http://go.worldbank. org/SLGFEKJT90. Women in Entrepreneurship 79 Summary of Findings 4 T his report reviews the performance of women and men during the first decade of the 21st century in three spheres: human capital endowments, labor markets, and entrepreneurship. Data are analyzed to determine whether women and men are performing well compared with each other, but also how they fare in a global context. The analysis is primarily quantitative and mines various data sets. This is a strength of the report because a quantitative analysis can add value by providing some measure of the degree of differences in the outcomes observed. Yet, it is a weakness as well because the outcomes are measured, but not always explained. Consequently, more work is needed in this area, especially qualitative analysis that is followed up by targeted individual surveys. Human Capital Endowments The countries of Europe and Central Asia have invested heavily in the education and health of their populations. These long-term investments have resulted in societies with well-educated and healthy citizens. They have also resulted in large payoffs among in- dividuals, giving men and women broadly comparable endowments. Thus, the stock of human capital in the region is substantial, and, despite the pressures resulting from the global economic crisis, social sector expenditures continue to be protected to a signifi- cant extent. Across the wide income spectrum in these countries, many key educational and health indicators vary by gender only moderately, though Turkey is an exception. Yet, despite this large stock of human capital, social developments in the region should give pause to decision makers and civil society alike. The population of the re- gion is aging rapidly. The growth imperative to raise productivity and employment is be- ing impeded by the fiscal straitjacket that many countries are facing and will continue to face for the foreseeable future. Though these are well-known challenges, there has been little discussion about whether gender issues are important in diagnosing the problems or framing the solutions. Though not all issues need to be gender differentiated, it is un- likely that demography, productivity, employment, and public expenditure are areas in which gender is marginal or irrelevant. This section discusses some of the most impor- 81 tant findings of the gender-disaggregated analysis of the region so that the average female and male gross human capital in Europe and Central Asia. enrollment rates are 60 and 45 percent, respectively. The region’s gender equality advantage in edu- Income appears to exacerbate the gap since more cation is eroding relative to the rest of the world, and well off countries exhibit a wider gender gap, on convergence with the high-income countries of the Or- average (figure 4.2). Yet, despite their strong pres- ganisation for Economic Co-operation and Develop- ence in tertiary education, women appear to be ment (OECD) has decelerated. The region has rightly poorly represented in science and engineering and prided itself for its low gender gap in primary and disproportionately represented in education, health secondary education. However, this advantage has care, welfare, and the humanities. This pattern is not eroded. Other regions have understood the im- unique to Europe and Central Asia, but is repeated portance of investing in education among girls and in other regions and in high-income countries. women and are succeeding in closing the gender gap The rapid demographic transition unfolding even at the secondary school level. Thus, though the in the region opens up opportunities for women, but region has not had to focus on closing a significant also increases the risks. In the majority of coun- gender gap, a small gap remains at the secondary tries, the population will contract rapidly because school level (figure 4.1). However, more important- of the low fertility rates. As more women obtain ly, though the income level of the region has been higher education and the demand for labor goes rising, enrollment rates among girls and boys have unabated, women have potentially even greater not been converging with the rates in high-income opportunities in the labor market. However, in all OECD countries, including in Western Europe. countries of Europe and Central Asia, the popula- The gender gap emerges significantly in tertiary tion is aging, and the majority of the elderly will be education and sets the stage for inequalities in the la- women: the average life expectancy of women ex- bor market and, possibly, in productivity. Substantially ceeds that of men by nine years. Without proper more women than men pursue higher education in pensions and savings, health and dependent care Figure 4.1 Secondary School Gross Enrollment Rates: The Region’s Gender Advantage is Disappearing a. Gender gap by region b. Rates by gender percent 120 105 100 100 80 95 60 90 40 85 20 80 0 75 ECA OEC EAP LAC MNA SAS SSA WLD 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 1991 2009 ECA Female ECA Male OECD Female OECD Male Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Note: The gender gap is the ratio of the gross secondary enrollment rate among girls and the corresponding rate among boys. eAP = east Asia and the Pacific. eCA = europe and Central Asia. LAC = Latin America and the Caribbean. MNA = Middle east and North Africa. OeCD = high-income countries of the Organisation for economic Co-operation and Development. SAS = South Asia. SSA = Sub-Saharan Africa. WLD = World. All regional data are for developing countries only. 82 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 4.2 Gender Gaps Emerge Significantly at the Tertiary Level a. Gross tertiary enrollment rates by income, 2009 b. Fields of study, 2008 percent % of total student body 120 30 26 100 25 80 20 15 15 60 15 14 10 40 10 9 7 20 5 4 0 0 Female Male Female Male Female Male Female Male 2.6 3.1 3.6 4.1 4.6 Science Social Educ, Other & Eng sciences, Health, Female Male business Welf and law Sources: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/; uNeCe gender Statistics (database), united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/ (data for 2008). Note: eng = engineering. educ = education. Other = agriculture, humanities, services, and unspecified fields. services, and housing, women will be vulnerable fare differently thereafter. The role of women in to old age poverty for the last decade of their lives. the household (for example, child and dependent In the south Caucuses, the low fertility rates are care and other household activities) has implica- exacerbated by a high gender imbalance at birth sec- tions for the professional lives of women in terms ond only to the imbalances in China and India. In of their special needs, such as maternity leave, Armenia, Azerbaijan, and Georgia, 11–16 percent flexible working hours, and leave benefits. Thus, more boys are born than girls, indicating prenatal women and men make distinct choices with regard sex discrimination. To a lesser degree, this gender to sectors of employment, occupational streams, imbalance exists in other countries of the region and the number of hours worked that inevitably as well, especially in a few of the countries of affect their earnings and wage rates. The presence the western Balkans. To determine the extent to or absence of public support for women’s fertility which this prenatal sex discrimination reflects en- and childcare needs can influence women’s labor trenched gender inequities in the economic sphere market behavior, as well as the decision to have (for example, in labor markets, migration oppor- children. In a region where reversing population tunities, and wealth accumulation through entre- decline and increasing employment are critical to preneurship) rather than household preferences sustaining economic growth, government action requires more analysis. In addition, there is a large may be pivotal. surplus of males in the 0–14 year age-group not Though more women are entering the labor only in the south Caucuses, but also in the western market in most developing regions and in high-income Balkans (especially, in the former Yugoslav Repub- OECD countries, women’s participation in the labor lic of Macedonia, Montenegro, and Serbia). market has remained stagnant, on average, in Europe and Central Asia over the past decade.85 In 2009, the Labor Markets 85 The labor force participation rates are computed as a Though men and women enter the labor market share of the population (by gender) between the ages with comparable human capital endowments, they of 15 and 64. Summary of Findings 83 Figure 4.3 Women’s and Men’s Labor Market Outcomes a. Labor force participation rates by region, 2009 b. Distribution of employment by occupation 15–64 age-group by gender, % Total employed by gender, % 100 40 38 35 80 31 30 60 25 22 20 19 20 40 20 17 18 15 20 10 9 6 0 5 ECA High LAC MNA SSA EAP SAS income 0 OECD Female Male Female Male Prof & Tech Admin Service Machine Other Sources: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/; uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: The data are for 2009 or the latest available year. See the note to figure 4.1 for an explanation of the terms. See chapter 2, table 2.1 for the specific country coverage in chart b. Prof & tech = professionals and technicians. Admin = administrative and intermediary level personnel. Service = service workers. Machine = machine operators. Other = directors and upper management, skilled agriculture, armed forces person- nel, and low-skill occupations. female and male labor force participation rates in taurants (20 percent). Communal services include the region were 58 and 74 percent, respectively, health care, education, and civil administration; compared with 65 and 80 percent, respectively, in many are government-provided services. Thus, high-income OECD countries (figure 4.3). Though the demographic transition and the decline in the in the low-income countries of the Commonwealth school-age population over time will adversely af- of Independent States (CIS), female and male labor fect women, who constitute almost 90 percent of force participation rates rose during 1999 and 2009 the teaching force. In addition to the concentra- and fell in Turkey, the rest of the region saw little tion in sectors, women are heavily concentrated change.86 Stagnation in labor force participation in three particular occupational categories in Eu- rates in the region could become a bottleneck for rope and Central Asia: 38 percent of all women future economic growth, especially in the econo- employed are professionals and technicians, and mies of the EU10 and selected CIS middle-income another 20 percent are administrative personnel countries that were particularly adversely affected (figure 4.3). by the crisis.87 Relative to men, women are more likely to work in the public sector and are concentrated heavily in 86 In 2008, the CIS consisted of Armenia, Azerbaijan, a few occupations and economic sectors. Europe and Belarus, Georgia, Kazakhstan, Kyrgyz Republic, Mol- Central Asia is one of the most highly segregated dova, the Russian Federation, Tajikistan, Turkmeni- regions occupationally. One in three employed stan (an associate member), Ukraine (an informal par- women works in the public sector, compared with ticipant), and Uzbekistan. 87 The EU10 refers to the 10 European Union (EU) 21 percent of employed men. About four in five countries of Central and Eastern Europe, that is, Bul- women are concentrated in three economic sec- garia, the Czech Republic, Estonia, Hungary, Latvia, tors: communal services (41 percent), manufac- Lithuania, Poland, Romania, the Slovak Republic, and turing (21 percent), and trade and hotels and res- Slovenia. 84 Opportunities for Men and Women: Emerging Europe and Central Asia Figure 4.4 Working Women’s Constraints a. Gender gap in time use, Europe and Central Asia and high-income OECD countries b. Employment rates, women with and without children hours per day percent 4.0 90 2.9 80 78 3.0 75 71 71 2.2 70 62 2.0 60 56 57 1.0 0.7 50 42 0.01 0.10 40 0.0 30 –1.0 –0.9 20 –2.0 –1.9 –1.7 10 –3.0 0 Gainful Domestic Free Time Others 0 1 2 3 Work Work Activities Number of children ECA OECD ECA OECD Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: The data are for 2009 or the latest available year. The gender gap for an activity is the number of hours spent by women, less the number of hours spent by men. The employment rate is the share of employed women (25–49 years of age) with the specified number of children (17 years or younger) in the total number of women in the same age-group with the specified number of children (17 years or younger). eCA = europe and Central Asia. Occupational segregation in the region may be the Though women are more highly educated than result of self-selection by women to balance work and men and though many are professionals or techni- family life. Women are clearly choosing different cians, women’s earnings are about 20 percent less types of jobs than men. There may be many reasons than men’s earnings. The gender gap in wages for this behavior, including the compatibility of the varies significantly across the region (figure 4.5). jobs they choose with raising children. On average, Moreover, women with the same educational working women in the region spend two hours less qualifications as their men counterparts earn 25 on paid work than men, but three hours more on percent less. However, only about a third of the domestic activities; the bulk of the latter is devoted gender wage gap can be explained by high occu- to food preparation and dish washing (37 percent); pational segregation. Women may be self-select- adult care (27 percent); cleaning, other upkeep, laun- ing into less well paying jobs or occupations that dry, and handicraft production (28 percent); and have favorable characteristics such as work hours childcare (11 percent). Though childcare appears to compatible with school hours, work flexibility, or take up relatively less time, other activities, such as favorable nonmonetary benefits.88 The large un- food preparation and laundry, are likely to increase explained difference between male and female because of the presence of children. If women have earnings may be the result of outright discrimina- children, their employment rates begin to fall dra- tion in the labor market, the need to give a costly matically in the region (figure 4.4). This sharp down- signal to employers if women show a greater vari- ward trend is not seen in the high-income OECD ability in productivity, or differences in hours countries, which may be because of the better in- frastructure for childcare, as well as more house- 88 Other reasons why women may be willing to take low- hold appliances such as dishwashers in the home. In paying jobs include shorter commutes, easy access by high-income OECD countries, one-third of under- public transportation, a safe work environment, less 3-year-olds are in childcare, compared with only 13 physically strenuous work (especially in the manufac- percent in Europe and Central Asia. turing sector), and greater job security. Summary of Findings 85 Figure 4.5 The Wage Gap a. Raw gender wage gap b. Enrollment rate, under-3-year-olds male wages in excess of female wages, % % of all under-3-year-olds 45 41 40 40 35 36 30 35 29 25 30 24 20 25 15 19 20 10 15 5 10 0 1980 1990 1995 2000 2001 2002 2003 2004 2005 2006 2007 2008 5 0 EU10 Central Asia Russia+ S. Caucuses W. Balkans ECA Western Europe Sources: For the wage gap: various (see chapter 2). For enrollment rate data: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: eCA = europe and Central Asia. worked and in actual experience. On the last fac- Entrepreneurship tor, women’s childbearing and relatively generous maternity leave benefits in the formal sector com- Entrepreneurship expands women’s economic op- bine to reduce women’s work experience; women portunities. It not only provides income, but it is a in Europe and Central Asia consider these labor means for women to accumulate assets and wealth. force interruptions to have costly long-term in- Entrepreneurship is important from a societal and come effects.89 economic perspective because it plays a key role in The mix of instruments to support women in private sector development, innovation, and em- balancing childbearing and childcare with paid ployment generation. Given that men and women work may be of critical importance to the demo- behave differently in the labor market, it is not sur- graphic and productivity challenges facing the re- prising that we find that men and women entrepre- gion. Governments in the countries of Europe neurs are different in terms of sectors of operation, and Central Asia support families in the func- constraints, employment generation, and profit- tions of childbearing and raising children. The ability. The necessity to enhance competitiveness most prevalent policies and programs include in the countries of Europe and Central Asia should generous maternity leave and child allowances. make the topic of entrepreneurship central to any However, the effectiveness of these programs strategy aiming at business development and ex- in maintaining women’s attachment to the labor pansion. market and in fertility have not been studied re- There are relatively few men employers and gionwide, nor has there been any assessment of even fewer women employers in Europe and Central the consequences of the lack of the widespread Asia compared with other regions. An estimated 2.4 availability of high-quality, affordable childcare percent of the active workforce in Europe and services. During the socialist period, childcare services were commonly available, but they quickly disappeared during the early years of the 89 The effect of a 12-month labor market interruption transition. in the United States resulted in an average 10 percent decline in the lifetime incomes of women who work full time. See Francesconi (2002). 86 Opportunities for Men and Women: Emerging Europe and Central Asia Central Asia consists of employers. The shares are in which they could play a pivotal role, that is, in higher in Latin America and the Caribbean (4.6 sole proprietorships. Moreover, top women man- percent) and in the high-income OECD coun- agers—who may have an even greater say in the tries (4.3 percent) (figure 4.6). Male employers direction of firms than some types of owners— account for 10 percent of all employed men in are comparatively few in the region. These man- the Middle East and North Africa, 6 percent in agers have an important role not only in day-to- Latin America and the high-income OECD coun- day firm operations, but also in determining the tries, but not in Europe and Central Asia or South strategic direction of firms. In the region, an esti- Asia, where the shares are only 3.4 and 1.2 per- mated 19 percent of firms have top women man- cent, respectively. This lower entrepreneurship agers, though the variation is substantial across in Europe and Central Asia is reflected among countries (figure 4.7). women as well. In Europe and Central Asia, only In Europe and Central Asia, woman-owned 1.1 percent of all employed women are employ- firms differ significantly from firms owned by men. ers compared with 2.7 percent in Latin America. Depending on the type of ownership, women’s This lack of engagement in business by women participation in the firm changes dramatically. At may be caused by many factors, including the one extreme, women and men are equal in terms of greater constraints to establishing and growing a ownership of partnership companies and publicly business or household decisions to reduce risk in traded companies. However, in private sharehold- an uncertain economic environment (particularly ing companies and sole proprietorships, about one among families with dependents). in three firms has female representation and female Women’s modest role in the private sector is input over the direction of the firms. Woman- seen across various dimensions of ownership and owned sole proprietorships are also smaller as mea- management. Though women appear to be repre- sured by sales and numbers of employees even after sented equally in some types of firms, especially one controls for country and industry. Women en- public shareholder companies and partnerships, trepreneurs are more heavily represented in certain there are proportionately fewer women in cases sectors, especially garments and textiles, but they Figure 4.6 Employers and Owners a. Comparison of employer rates b. Women’s ownership by type of �rm employers among total employed by gender, % % of �rms with female owners 12.0 60 10.0 51 52 10.0 50 48 8.0 40 36 38 34 5.9 5.9 6.0 30 4.0 3.4 20 2.3 2.7 2.0 1.7 1.5 10 1.1 0.5 0.0 0 ECA OEC LAC MNA SAS Public Sh. Private Sh. Sole Prop. Partnership Ltd. Part Other Female Male Sources: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- indicators/. Ownership data: Business environment and enterprise Performance Survey, european Bank for reconstruction and Development and World Bank, London, http://www.ebrd.com/pages/research/analysis/surveys/beeps.shtml. Note: eCA = europe and Central Asia. Ltd Part = limited partnership. Private Sh. = private shareholders. Public Sh. = public shareholders. Sole Prop. = sole proprietorship. Panel a: the data are for 2008 or the latest available year. Middle east is 2007, and South Asia is 2005. Summary of Findings 87 Figure 4.7 Firm Characteristics of Woman-Owned and Woman-Managed Firms a. Female top management, by selected countries b. Women-owned �rms, by sectors % of �rms % of �rms 35 60 56 30 50 43 45 25 40 32 33 20 30 24 26 15 20 10 10 5 0 0 Basic metals Electronics IT Wholesale Retail Chemical Garments & textiles Kosovo Azerbaijan Uzbekistan Ukraine Poland Latvia Source: Business environment and enterprise Performance Survey, european Bank for reconstruction and Development and World Bank, London, http://www.ebrd.com/pages/research/analysis/surveys/beeps.shtml. Note: Data are for 2009 or the latest available year. in chart a, the averages are taken from enterprise Surveys (database), enterprise Analysis unit, World Bank, Washington, DC, http://www.enterprisesurveys.org/. in chart b, the data refer to the selected sectors only. iT = information technology. are poorly represented in basic metals, transport, ties in opportunity. These gender differences may and electronics. worsen the already difficult economic future facing Women’s firms generally operate at an ineffi- the countries of Europe and Central Asia. There ciently small scale, resulting in a (small) gender gap is a need to accelerate growth through reforms in in profits. The average gender gap in profits in the health care, education, and pensions and through region is about 3 percent in favor of firms owned by improvements in productivity and competitive- men.90 This difference is driven by the smallness of ness against the backdrop of dramatic demographic women’s firms; the gap disappears after we correct changes. Some of the gender differences in health, for this characteristic. Women operate firms at an education, labor markets, and private sector busi- inefficiently small scale for two reasons. First, wom- ness are likely to have an impact on how the nec- en are concentrated in industries in which small essary fiscal, sectoral, and institutional reforms are firms dominate. Second, women appear to be capi- designed and implemented. tal constrained; this is apparent in capital-intensive It is important to address two types of gender industries. Firms that are more likely to have access issues. The first type pertains to issues that, if they to capital are larger, and women entrepreneurship are not addressed, threaten to cause the region to is underrepresented in these firms. move backward or to stagnate in the social sector. These include the following: Policy Implications and Critical ◆ The gender imbalance that has emerged in the Knowledge Gaps south Caucuses and in the western Balkans is a grave concern. A preference for males appears This report examines a narrow range of issues to have grown dramatically in recent years; if pertaining to men’s and women’s economic op- this issue is not addressed, there will be nega- portunities. However, even among the few topics tive consequences for population growth. that it surveys (human capital, labor markets, and entrepreneurship), women and men show differ- 90 These data have been corrected for country- and ent outcomes, which, at times, reflect inequali- industry-specific effects. 88 Opportunities for Men and Women: Emerging Europe and Central Asia ◆ The lack of progress in secondary enrollment and women have greater household responsi- rates and the small, but persistent gender gap bilities, but also jobs are being offered, espe- over the last decade are contrary to the expec- cially in the public sector, that are associated tations of countries seeking significant eco- with benefits that are well suited to meeting nomic growth and sharp declines in poverty. these challenges. Given the economic impor- tance of raising productivity and employment, The second type of gender issue that should research on the effects of improvements in be addressed in the region is important mostly be- the coverage of childcare services and non- cause of the implications for productivity and the monetary benefits that ease women’s role in health of labor markets. This type involves gender managing household responsibilities may be differentials. No single instrument is available that worthwhile at the country level. is appropriate for dealing with issues of this type, ◆ Women are underrepresented in private sector but a range of cross-sectoral policies can be effec- leadership. Data indicate that entrepreneur- tive. Though this report provides evidence on the ship is less common in Europe and Central causes of certain outcomes, more research in these Asia than in other parts of the world and that areas is warranted. This second type includes three women play only a modest role. To understand main issues, as follows: why relatively few women enter business pro- fessions, more research should be initiated on ◆ Women’s labor force participation rates in the the costs of business failure, the advantages of region lag behind the average in high-income job mobility, household risk mitigation strate- OECD countries. Labor market participation gies, and the (risk-adjusted) returns to entre- is a personal choice that is often associated preneurship among women relative to those with financial circumstances. Women may not available among men.91 Women entrepreneurs participate in the labor market because of their in formal sector firms appear, on average, to responsibilities in caring for dependents, their face more difficulty than men in obtaining limited opportunities for part-time work, the credit (because of firm size).92 Whether any early retirement options available to women, policies can be instituted to help the owners and the generous maternity leave in the formal of small firms should be explored. sector. Thus, reforms in long-term dependent care, retirement legislation, pension reform, Addressing gender differentials is a complex, and social protection could all contribute to but feasible task. Important areas of analysis must enhancing the labor market opportunities and still be addressed, especially to identify the causes long-term income potential of women. and the motivations behind certain outcomes in ◆ Occupational segregation that affects women the private and professional spheres. Governments may be the result of a choice women are mak- in the region play an important role in determining ing to achieve a better balance between work how women interact with the labor market, wom- and other aspects of their lives. Occupational en’s educational attainment, and even household segregation may already be occurring as early fertility decisions. Government policies in what may as tertiary education when many men and be considered a marginal area can ultimately have a women make their first career choices based significant impact on a country’s economic future. on education. Professions that allow them to receive relatively generous benefits, have greater flexibility in hours worked, and are 91 Do women pursue more stable jobs, while men take more risks? compatible with raising children may be more 92 The question that remains is whether the small size attractive to women. The choice of occupa- of women’s firms is also often due to lack of access to tion is made within an environment in which start-up capital or working capital. However, the data the availability of childcare services is limited necessary for further analysis are lacking. Summary of Findings 89 Annexes 91 92 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 45 indicators/. indicators/. Bosnia and Herzegovina Bosnia and Herzegovina Bulgaria Bulgaria Poland Poland Figure A.2 Figure A.1 Slovenia Slovenia Annex A: Data Romania Romania Slovakia Slovakia Croatia Croatia Belarus Belarus Latvia Latvia Czech Republic Czech Republic Lithuania Lithuania Hungary Hungary TFYR Macedonia TFYR Macedonia Ukraine Ukraine Georgia Georgia Estonia Estonia Russian Federation Russian Federation Republic of Moldova Republic of Moldova 2009 2009 Serbia Serbia Cyprus Cyprus Armenia Armenia Montenegro Montenegro Albania Albania 2050 2050 Azerbaijan Azerbaijan Opportunities for Men and Women: Emerging Europe and Central Asia Turkey Turkey Bahrain Bahrain Kazakhstan Kazakhstan Uzbekistan Uzbekistan Percentage of the Population above 80, 2009 and 2050 Percentage of the Population above 60, 2009 and 2050 Kyrgyzstan Kyrgyzstan Turkmenistan Turkmenistan Tajikistan Tajikistan Western Europe Western Europe Africa Africa Asia Asia Latin America Latin America Northern America Northern America World World Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- Source: World Development indicators Database, World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development- Table a.1 | Life Expectancy, by Country 1990–95 2000–05 2010–15 Change 1990–2010 Country, Women Women Women Women region Women Men − men Women Men − men Women Men − men Women Men − men Albania 74.9 68.9 6.0 79.0 72.6 6.4 80.4 74.2 6.2 5.50 5.30 0.20 Armenia 71.5 64.5 7.0 75.6 68.9 6.7 77.7 71.3 6.4 6.20 6.80 −0.60 Azerbaijan 69.7 60.7 9.0 71.2 65.6 5.6 73.5 69.1 4.4 3.80 8.40 −4.60 Belarus 74.9 64.2 10.7 74.6 62.5 12.1 76.0 64.5 11.5 1.10 0.30 0.80 Bosnia and 73.2 55.8 17.4 77.2 71.6 5.6 78.4 73.3 5.1 5.20 17.50 −12.30 Herzegovina Bulgaria 74.7 67.6 7.1 75.6 68.7 6.9 77.7 70.9 6.8 3.00 3.30 −0.30 Croatia 76.5 68.3 8.2 78.4 71.4 7.0 80.4 73.8 6.6 3.90 5.50 −1.60 Czech 76.4 69.3 7.1 78.7 72.1 6.6 80.3 74.3 6.0 3.90 5.00 −1.10 Republic Estonia 74.0 62.7 11.3 76.9 65.6 11.3 79.3 68.9 10.4 5.30 6.20 −0.90 Georgia 74.3 66.5 7.8 75.0 68.0 7.0 75.9 69.1 6.8 1.60 2.60 −1.00 Hungary 73.9 64.8 9.1 76.6 68.3 8.3 78.3 70.4 7.9 4.40 5.60 −1.20 Kazakhstan 70.3 60.5 9.8 70.4 59.1 11.3 72.1 60.1 12.0 1.80 –0.40 2.20 Kyrgyz 70.1 62.1 8.0 70.6 62.7 7.9 72.7 65.5 7.2 2.60 3.40 −0.80 Republic Latvia 73.9 61.9 12.0 76.2 65.3 10.9 78.1 68.7 9.4 4.20 6.80 −2.60 Lithuania 75.5 64.4 11.1 77.5 66.3 11.2 78.3 67.0 11.3 2.80 2.60 0.20 Macedonia, 74.0 69.4 4.6 75.9 71.1 4.8 77.3 72.6 4.7 3.30 3.20 0.10 FYR Moldova 70.9 63.6 7.3 71.2 63.6 7.6 73.1 65.6 7.5 2.20 2.00 0.20 Montenegro 79.1 72.8 6.3 77.0 71.9 5.1 77.4 73.1 4.3 –1.70 0.30 −2.00 Poland 75.9 67.0 8.9 78.8 70.4 8.4 80.4 72.3 8.1 4.50 5.30 −0.80 Romania 73.2 65.8 7.4 75.1 67.8 7.3 77.2 70.3 6.9 4.00 4.50 −0.50 Russian 72.5 60.5 12.0 71.8 58.5 13.3 74.1 61.9 12.2 1.60 1.40 0.20 Federation Serbia 74.5 69.2 5.3 75.6 70.9 4.7 77.1 72.5 4.6 2.60 3.30 −0.70 Slovak 76.2 67.8 8.4 77.8 69.8 8.0 79.3 71.8 7.5 3.10 4.00 −0.90 Republic Slovenia 77.6 69.6 8.0 80.3 72.6 7.7 82.6 75.4 7.2 5.00 5.80 −0.80 Tajikistan 65.9 58.6 7.3 68.6 60.9 7.7 70.4 65.1 5.3 4.50 6.50 −2.00 Turkey 68.5 64.0 4.5 73.3 68.5 4.8 75.2 70.3 4.9 6.70 6.30 0.40 Turkmenistan 66.6 58.9 7.7 68.2 60.4 7.8 70.0 62.4 7.6 3.40 3.50 −0.10 Ukraine 73.5 63.6 9.9 73.4 62.1 11.3 74.3 63.9 10.4 0.80 0.30 0.50 Uzbekistan 69.4 63.0 6.4 70.4 64.1 6.3 72.0 65.7 6.3 2.60 2.70 −0.10 Europe and 73.2 64.7 8.5 74.9 66.9 7.9 76.5 69.1 7.4 3.38 4.41 −1.04 Central Asia Commonwealth 70.5 62.0 8.6 71.6 63.1 8.6 73.4 65.5 8.0 2.92 3.52 −0.60 of Independent States Central and 73.1 64.5 8.6 74.7 66.7 8.0 76.4 68.9 7.5 3.30 4.38 −1.08 Eastern Europe Source: uN (2009a). Annexes 93 Table a.2 | Age-Specific Fertility Rates births per 1,000 women Country 15–19 20–24 25–29 30–34 35–39 40–44 45–49 Albania 14.2 104.8 136.7 80.7 29.6 7.5 1.2 Armenia 35.7 152.1 95.5 43.3 16.6 3.7 0.2 Azerbaijan 33.8 168.1 128.5 67.6 27.6 5.2 0.3 Belarus 21.3 89.8 79.9 46.4 15.9 2.5 0.1 Bosnia and 15.9 79.6 78.6 45.8 18.9 4.0 0 Herzegovina Bulgaria 42.2 77.5 84.4 55.4 12.5 5.6 2.8 Croatia 14.1 62.8 98.9 75.0 29.0 5.0 0 Czech Republic 10.6 46.3 102.0 87.8 29.9 4.7 0.2 Estonia 21.4 78.8 104.7 80.7 35.8 6.2 0.2 Georgia 44.7 106.7 88.7 48.7 20.4 6.8 0 Hungary 20.2 49.3 96.6 90.2 10.8 3.2 0 Kazakhstan 30.7 152.0 141.0 90.3 39.5 7.9 0.6 Kyrgyz Republic 32.3 182.9 146.0 89.2 44.1 15.4 2.6 Latvia 15.2 70.7 89.1 67.5 30.0 6.1 0.3 Lithuania 21.9 70.3 87.0 59.1 25.3 4.8 0.2 Macedonia, FYR 21.7 74.5 99.9 68.3 20.4 3.2 0 Moldova 33.8 106.9 87.1 51.5 17.4 3.1 0.2 Montenegro 14.7 87.9 106.9 77.0 32.7 7.6 0.4 Poland 13.9 57.4 90.3 62.5 23.8 5.1 0 Romania 31.2 61.8 86.3 60.7 20.0 3.8 0.2 Russian Federation 25.1 90.3 83.9 51.9 18.9 3.0 0.1 Serbia 22.1 94.9 107.8 66.5 25.7 4.7 1.9 Slovak Republic 20.7 58.0 86.8 63.7 22.5 3.7 0.2 Slovenia 4.9 40.7 101.3 89.3 30.3 4.7 0.2 Tajikistan 28.4 190.3 222.4 155.4 68.0 21.9 3.6 Turkey 38.8 127.0 133.4 73.7 38.4 12.4 2.6 Turkmenistan 19.5 144.6 176.0 108.6 42.2 7.9 1.0 Ukraine 28.3 94.0 78.0 44.3 15.3 2.5 0.1 Uzbekistan 12.9 158.8 154.1 92.1 34.0 5.9 0.6 Source: Data of the united Nations. Note: Mean age at childbearing is the average age mothers would have at the birth of their children if women were subject throughout their lives to the age- specific fertility rates observed in a given year. An age-specific fertility rate is defined as the annual number of births to women in a particular age-group divided by the number of years lived by the women in that age-group. it is expressed as the number of births per 1,000 women in the age-group considered. 94 Opportunities for Men and Women: Emerging Europe and Central Asia Figure A.3 Relation between the Total Fertility Rate and the Mean Age at Birth 4.0 3.5 3.0 Total Fertility Rate 2.5 y = 0.0249x + 1.0111 2.0 1.5 1.0 0.5 0 25 26 27 28 29 30 Mean age at birth Source: World Bank staff. Note: The equation is TFr = 1.01 (2.89) + 0.025 Age (0.10). Annexes 95 Table a.3 | Age at First Birth Country 1980 1990 1995 2000 2001 2002 2003 2004 2005 2006 2007 2008 Albania — — — — — — — — — — — — Armenia 22 23 23 22 22 22 22 23 23 23 23 23 Azerbaijan 23 23 24 24 24 24 24 24 24 24 24 24 Belarus — 23 23 23 23 24 24 24 24 24 24 24 Bosnia and Herzegovina 23 24 — 24 24 24 24 24 24 25 25 25 Bulgaria 22 22 22 24 24 24 24 25 25 25 25 25 Croatia 23 24 25 26 26 26 26 26 27 27 27 27 Czech Republic 22 22 23 25 25 26 26 26 27 27 27 27 Estonia 23 23 23 24 24 25 25 25 25 25 25 26 Georgia — 24 24 24 25 25 25 24 24 24 24 24 Hungary 23 23 23 25 25 26 26 27 27 27 28 28 Kazakhstan — 22 22 23 24 24 24 24 24 25 25 25 Kyrgyz Republic 22 22 22 23 23 23 23 23 23 24 24 23 Latvia 23 23 24 24 25 25 25 25 25 25 25 26 Lithuania 24 23 23 24 24 24 25 25 25 25 25 25 Macedonia, FYR 23 23 24 24 24 25 25 25 25 25 25 26 Moldova 23 — 22 22 22 22 22 22 22 23 23 23 Montenegro — — — 26 26 26 26 26 26 25 26 26 Poland 23 24 24 25 25 25 25 26 26 26 26 26 Romania 23 22 23 24 24 24 24 25 25 25 25 26 Russian Federation 23 23 23 — — — — — — — — — Serbia 23 24 24 25 25 25 25 26 26 26 26 27 Slovak Republic — 21 22 24 24 25 25 25 26 26 26 27 Slovenia 23 24 25 27 27 27 27 28 28 28 28 28 Tajikistan 22 22 22 22 22 — — — — — — — Turkey — — — — — — — — — — — — Turkmenistan — 24 24 24 24 24 24 25 25 25 — — Ukraine 22 23 — 22 23 23 22 23 24 24 26 26 Uzbekistan — 22 22 23 23 23 23 24 24 24 — — Europe and Central Asia 23 23 23 24 24 24 25 25 25 25 25 25 Source: uNeCe Statistical Database, united Nations economic Commission for europe, geneva, http://w3.unece.org/pxweb/. Note: — = not available. 96 Opportunities for Men and Women: Emerging Europe and Central Asia Annex B: Selected Government-Led helps create public recognition and visibility Initiatives in Europe and Central Asia among women entrepreneurs. Germany has launched initiatives to increase wom- The Slovenian government supports women en’s entrepreneurship with a view to mainstream entrepreneurs by facilitating business support ser- women entrepreneurs and increase their visibility vices, as follows: and their access to finance, as follows: ◆ Under business support schemes, women en- ◆ TWIN, the Two Women Win program, encour- trepreneurship promoters have been trained ages mentoring among the ventures of young to build networks of experts. women. After a successful start in North ◆ The Small Business Development Center Rhine–Westphalia, the program has been ex- sponsors business support schemes aimed at tended to the whole of Germany. women entrepreneurs. ◆ Through a loan program in North Rhine–West- phalia, the state government allows applica- The Lithuanian government has initiated sev- tions by women entrepreneurs who have no eral measures to foster the development of small previous industry knowledge. and medium enterprises (SMEs). European Union ◆ In the eastern German state of Mecklenburg- membership has represented an impetus for wom- Pomerania, the state investment bank gives en entrepreneurship in Lithuania, and several SME loans directly to women entrepreneurs, provided development programs have been aimed at women their loan requests have previously been rejected entrepreneurs under the auspices of the Lithuanian by a bank. Development Agency for Small and Medium En- ◆ Various ministries have undertaken initia- terprises, including: tives to encourage women to take advantage of new opportunities and to mainstream women ◆ Subsidized consulting services for SMEs and in chambers of commerce and business associa- start-ups aimed at rural and urban start-ups, tions. especially women and youth business found- ◆ Eight federal and two state awards have been ers established for women to recognize innova- ◆ A government-supported guarantee fund tive business ideas and new products, ser- through the agency Invega to assist SMEs in vices, or innovative ways of combining work gaining access to bank loans and family, as well as employment creation ◆ The development of business incubators in through new enterprises. There are also sev- seven Lithuanian cities eral public-private initiatives to increase the ◆ A separate page on the website of the SME visibility of women entrepreneurs by giving development agency with information on awards to women entrepreneurs and women success stories, statistics on women entrepre- business founders. Many of these awards are neurs, and so on sponsored by well-known public and private companies or mass media companies. This Source: Welter et al. (2006). Annexes 97 Annex C: Selected Donor Initiatives initiatives, nongovernmental organizations in the Kyrgyz Republic and Serbia attempt to raise aware- A publicly owned loan guarantee fund for small and ness about women’s property rights and land own- medium enterprises (SMEs) in Poland assists small ership rights so as to change patterns of ownership. enterprises in gaining access to commercial bank Similarly, the Pink Card Project of the Province of credit by establishing a network of approximately Milan educates women about the importance of es- 100 local and 16 regional credit guarantee funds for tablishing a credit history and managing their per- SMEs and approximately 100 local and 16 regional sonal finances. credit funds for small companies. Entrepreneurship The United Nations Economic Commission First, an enterprise development package in Poland, for Europe and the Regional Cooperation Council includes the following policy components: are supporting a women entrepreneurs’ forum in southeastern Europe. Set up in the context of the ◆ Simplification of the tax system crisis, the forum seeks to tackle the main obstacles ◆ Lowering labor costs and making employ- faced by women entrepreneurs in southeastern Eu- ment relations more flexible rope, to strengthen women’s economic position, ◆ Simplification of the social security system enhance the contribution of women to their econo- ◆ Simplification of the regulations for applica- mies, and help promote economic cooperation and tions for various procedures stability in southeastern Europe. The project aims ◆ New legal regulations to achieve this through the following: ◆ A review of procedures, instructions, and other internal regulations meant to remove ◆ By improving women’s access to networking the potential barriers in the contacts between opportunities administrative offices and entrepreneurs ◆ By facilitating women’s access to credit through innovative financing Women’s World Banking, a Spanish program, ◆ By improving women’s skills and capacity to helps women who lack collateral to access bank develop their businesses loans, and the European Bank for Reconstruction and Development’s lending packages target wom- Among other examples of public-private an-owned SMEs, provide women entrepreneurs partnership to support women entrepreneurs is with information, counseling, and guidance to en- the Alliance for the Support of Women in Busi- sure that they are aware of existing resources, and ness in Astana through programs at the Enterprise help women navigate the loan application process. Development Center of Astana, in Kazakhstan The bank’s program also works directly with the (USAID 2011). The United States Agency for In- staff of financial institutions to educate them about ternational Development partnered with the Exx- women’s financing needs and help them tailor their onMobil Foundation and the Kazakhstan Loan services and financial tools to the needs of women Fund to establish the initiative. The United States entrepreneurs. Agency for International Development and the Among targeted financial tools for women ExxonMobil Foundation provided funding and entrepreneurs are the pilot project of the Nordic technical support. The program is managed by two Investment Bank and the Council of Europe Bank local organizations. By the end of 2009, the En- for the Baltic States, as well as the loan package terprise Development Center had trained about provided by the lending agency Finnvera, which 4,500 entrepreneurs (over 75 percent of whom is funded by the Finnish government (Hellen and were women) in strategic and business planning, fi- Seppäla 2004). The Women’s Cooperative Bank, a nancial management, marketing, human resources nonprofit lending institution in Cyprus, does not management, taxes, and law. Over 130 small and exclusively target women in its services, but priori- medium businesses have since received business tizes women through its mission. Among nonprofit consulting services through the alliance. 98 Opportunities for Men and Women: Emerging Europe and Central Asia References Aculai, E., N. Rodionova, and N. Vinogradova. 2006. “Women Business Owners in Mol- dova: Proprietors or Entrepreneurs.� In Enterprising Women in Transition Economies, ed. F. Welter, D. Smallbone, and N. B. Isakova, 67–92. Aldershot, Hampshire, United Kingdom: Ashgate Publishing Limited. Aidis, R. 2005. “Entrepreneurship in Transition Countries: A Review.� Working Paper 61 (December), Centre for the Study of Economic and Social Change in Europe, School of Slavonic and East European Studies, University College London, Lon- don. ———. 2006. “From Business Ownership to Informal Market Traders: The Character- istics of Female Entrepreneurship in Lithuania.� In Enterprising Women in Transi- tion Economies, ed. F. Welter, D. Smallbone, and N. B. Isakova, 119–42. Aldershot, Hampshire, United Kingdom: Ashgate Publishing Limited. Aidis, R., F. Welter, D. Smallbone, and N. B. Isakova. 2007. “Female Entrepreneurship in Transition Economies: The Case of Lithuania and Ukraine.� Feminist Economics 13 (2): 157–83. Alam, A., P. Anós Casero, and F. Khan. 2008. Unleashing Prosperity: Productivity Growth in Eastern Europe and the Former Soviet Union. Washington, DC: World Bank. Alesina, A. F., F. Lotti, and P. E. Mistrulli. 2009. “Do Women Pay More for Credit? Evidence from Italy.� NBER Working Paper 14202, National Bureau of Economic Research, Cambridge, MA. Allen, I. E., A. Elam, N. Langowitz, and M. Dean. 2008. “Global Entrepreneurship Mon- itor: 2007 Report on Women and Entrepreneurship.� Global Entrepreneurship Research Association, London and Wellesley, MA. http://www.gemconsortium. org/about.aspx?page=special_topic_women. Altonji, J. G., and R. M. Blank. 1999. “Race and Gender in the Labor Market.� In Hand- book of Labor Economics, vol. 3, part 3, ed. O. C. Ashenfelter and D. Card, 3143– 3259. Amsterdam: Elsevier Science. Apostolova, B. 2010. “Gender Wage Gap in Western Balkan Countries.� Paper present- ed at the World Bank’s International Conference, “Poverty and Social Inclusion 99 in the Western Balkans,� Brussels, December Blunch, N.-H., and V. Sulla. 2011. “The Financial 14–15. Crisis, Labor Market Transitions and Earn- Averett, S. L., and L. A. Whittington. 2001. “Does ings: A Gendered Panel Data Analysis for Ser- Maternity Leave Induce Births?� Southern bia.� Paper presented at the Sixth Institute for Economic Journal 68 (2): 403–17. the Study of Labor–World Bank Conference Babović, M. 2008. “The Position of Women on the “Employment and Development,� Mexico Labour Market in Serbia.� Gender Equality City, May 30-June 1. Council, Government of the Republic of Ser- Boeri, T., D. Del Boca, and C. Pissarides. 2005. bia, and United Nations Development Pro- Women at Work: An Economic Perspective. New gramme, Belgrade. York: Oxford University Press. Barber, B. M., and T. Odean. 2001. “Boys Will Be Bowersox, G., D. Sibley, L. Snee, and M. L. Vitelli. Boys: Gender, Overconfidence, and Common 2007. “Assessment of Afghanistan Gemstone Stock Investment.� Quarterly Journal of Eco- Industry: Final Report.� June, USAID, Wash- nomics 116 (1): 261–92. ington, DC. http://www.gems-afghan.com/ Bardasi, E., P. Paci, and N. Pignatti. 2007. “Is It ngo/USAID-SibleyFinalMayReport0607.pdf. What She Is, What She Does, or the Way Brainerd, E. 2000. “Women in Transition: Changes She Moves? Assessing the Role of Segmenta- in Gender Wage Differentials in Eastern Eu- tion in Explaining the Gender Wage Gap in rope and the Former Soviet Union.� Industrial Transition and Emerging Countries.� Paper and Labor Relations Review 54 (1): 138–62. presented at the Associazione Italiana degli Chawla, M., G. Betcherman, and A. Banerji. 2007. Economisti del Lavoro’s “XXII Conference of From Red to Gray: The “Third Transition� of Ag- Labor Economics,� Department of Economic ing Populations in Eastern Europe and the For- Studies and Faculty of Economics, Parthe- mer Soviet Union, with Anne M. Bakilana et al. nope University of Naples, Naples, Septem- Washington, DC: World Bank. ber 13–14. Constant, A. 2006. “Women and Entrepreneurship Becker, G. 1971. The Economics of Discrimination. in Germany.� AICGS Issue Brief 12, Ameri- Chicago: University of Chicago Press. can Institute for Contemporary German Berger, L. M., J. Hill, and J. Waldfogel. 2005. “Ma- Studies, Johns Hopkins University, Washing- ternity Leave, Early Maternal Employment ton, DC. and Child Health and Development in the Darnovsky, M. 2009. “Countries with Laws or Poli- US.� Economic Journal 115 (501): F29–F47. cies on Sex Selection.� Memorandum, April, Blau, F. D., M. Ferber, and A. Winkler. 2002. The Center for Genetics and Society, Berkeley, CA. Economics of Women, Men, and Work. 4th ed. Dimova, R., and I. N. Gang. 2004. “Self-Selection Englewood Cliffs, NJ: Prentice Hall. and Earnings during Volatile Transition.� IZA Blau, F. D., and L. M. Kahn. 1997. “Swimming Up- Discussion Paper 1158, Institute for the Study stream: Trends in the Gender Wage Differen- of Labor, Bonn. tial in the 1980s.� Journal of Labor Economics Dimova, R., I. N. Gang, and J. Landon-Lane. 2006. 15 (1): 1–42. “Where to Work? The Role of the Household ———. 2004. “The US Gender Gap in the 1990s: in Explaining Gender Differences in Labour Slowing Convergence.� NBER Working Paper Market Outcomes.� IZA Discussion Paper 10853, National Bureau of Economic Re- 2476, Institute for the Study of Labor, Bonn. search, Cambridge, MA. Dohmen T. F. A., A. Falk, D. Huffman, U. Sunde, J. Blunch, N.-H. 2010. “The Gender Earnings Gap Schupp, and G. G. Wagner. 2005. “Individual Revisited: A Comparative Study for Serbia Risk Attitudes: New Evidence from a Large, and Five Countries in Eastern Europe and Representative, Experimentally-Validated Central Asia.� Unpublished working paper, Survey.� IZA Discussion Paper 1730, Insti- Institute for the Study of Labor, Bonn. tute for the Study of Labor, Bonn. 100 Opportunities for Men and Women: Emerging Europe and Central Asia Dostie, B., and D. Sahn. 2008. “Labor Market Dy- Fortin, N. M., and T. Lemieux. 2000. “Are Women’s namics in Romania during a Period of Eco- Wage Gains Men’s Losses? A Distributional nomics Liberalization.� Unpublished report, Test.� American Economic Review 90 (2): 456– HEC, Montreal. 60. Drnovsek, M., and M. Glas. 2006. “Women Entre- Fortin, N. M., T. Lemieux, and S. Firpo. 2011. “De- preneurs in Slovenia: By Fits and Starts.� In composition Methods in Economics.� In La- Enterprising Women in Transition Economies, bor Economics, ed. O. Ashenfelter and D. Card, ed. F. Welter, D. Smallbone, and N. B. Isakova, 1–102. Vol. 4A of Handbook of Labor Econom- 143–70. Aldershot, Hampshire, United King- ics. Amsterdam: Elsevier North Holland dom: Ashgate Publishing Limited. Francesconi, M. 2002. “A Joint Dynamic Model of Duflo, E. 2004. “Gender Equality and Develop- Fertility and Work of Married Women.� Jour- ment.� Unpublished paper, Massachusetts In- nal of Labor Economics 20 (2): 336–80. stitute of Technology, Cambridge, MA. Galor, O. 2011. “The Demographic Transition: Eriksson, T., M. Pytliková, and F. Warzynski. 2010. Causes and Consequences.� NBER Working “Increased Sorting and Wage Inequality in the Paper 17057, National Bureau of Economic Czech Republic: New Evidence Using Linked Research, Cambridge, MA. Employer-Employee Dataset.� Working Pa- Ganguli, I., and K. Terrell. 2006. “Institutions, Mar- per 09–5, Aarhus School of Business, Univer- kets and Men’s and Women’s Wage Inequal- sity of Aarhus, Aarhus, Denmark. ity: Evidence from Ukraine.� Journal of Com- EuroChambres. 2004. “Women in Business and parative Economics 34 (2): 200–27. in Decision Making: The Situation in Cham- Golden, L. 2001. “Flexible Work Schedules: What bers across Europe.� Association of Euro- Are We Trading Off to Get Them?� Monthly pean Chambers of Commerce and Industry, Labor Review 124 (3), March. Brussels. http://www.echwomennetwork. Hallward-Driemeier, M., R. Rijkers, and A. Wax- eu/Objects/2/Files/WiB_Chambers_final. man. 2011. “Ladies First? Firm-Level Evidence pdf. on the Labor Impacts of the East Asian Cri- European Commission. 2010. Combating Poverty sis.� Policy Research Working Paper 5789, and Social Exclusion: A Statistical Portrait of World Bank, Washington, DC. the European Union 2010. Eurostat Statistical Hellen, S., and M. Seppäla. 2004. “Financing Wom- Books. Luxembourg: Publications Office of en Entrepreneurs in the Baltic States.� In Ac- the European Union. cess to Finance and ITC for Women Entrepre- Fallick, B., J. Haltiwanger, and E. McEntarfer. 2010. neurs in the UNECE Region: Challenges and “Nonemployment Duration and the Conse- Good Practices, Report ECE/TRADE/336, quences of Job Separations.� Unpublished 64–68. Geneva: United Nations Economic working paper, Center for Economic Studies, Commission for Europe. U.S. Census Bureau, Washington, DC. Hoem, J. M. 1990. “Social Policy and Recent Fertil- Fernández-Kranz, D., and N. Rodríguez-Planas. ity Change in Sweden.� Population and Devel- 2011. “The Part-Time Pay Penalty in a Seg- opment Review 16 (4): 735–48. mented Labor Market.� Labour Economics 18 Holzer, H., and D. Neumark. 2000. “Assessing Af- (5): 591–606. firmative Action.� Journal of Economic Litera- Flabbi, L. 2010. “Prejudice and Gender Differ- ture 38 (3): 483–568. entials in the U.S. Labor Market in the Last Hunt, J. 2002. “The Transition in East Germany: Twenty Years.� Journal of Econometrics 156 When Is a Ten Percent Fall in the Gender Pay (1): 190–200. (Background paper prepared Gap Bad News?� Journal of Labor Economics for World Development Report 2012: Gender 20 (1): 148–69. Equality and Development, World Bank, Wash- IFC (International Finance Corporation). 2006. ington, DC.) “Women Entrepreneurs and Access to Fi- References 101 nance: Program Profiles from around the Transition Economies: A Reader. Transition World.� Document, November, IFC, Wash- and Development Series. Aldershot, Hamp- ington, DC. shire, United Kingdom: Ashgate. ———. 2010. Scaling UP SME Access to Financial Lang, K., and M. Manove. 2006. “Education and Services in the Developing World. Report, Octo- Labor-Market Discrimination.� NBER Work- ber. Washington, DC: IFC. ing Paper 12257, National Bureau of Eco- IFC (International Finance Corporation) and MI- nomic Research, Cambridge, MA. Bospo. 2008. “Voices of Women Entrepre- Lastarria-Cornhiel, S. 2009. “Women’s Role in neurs in Bosnia and Herzegovina.� May, IFC, Agriculture and in Rural Welfare: Access to Washington, DC; MI-Bospo, Tuzla, Bosnia Land and Resources.� Paper prepared for Ex- and Herzegovina. pert Group Meeting, “The Impact of the Im- ILO (International Labour Office). 2007. Equality plementation of the Beijing Declaration and at Work: Tackling the Challenges. Geneva: ILO. Platform for Action on the Achievement of Isakova, N. B., O. Krasovska, L. Kavunenko, and the Millennium Development Goals,� United A. Lugovy. 2006. “Entrepreneurship in the Nations, Geneva, November 11–13. Ukraine: A Male Female Comparison.� In En- Lisowska, E. 2002. “Women’s Entrepreneurship: terprising Women in Transition Economies, ed. F. Trends, Motivations, and Barriers.� In Wom- Welter, D. Smallbone, and N. B. Isakova, 17– en’s Entrepreneurship in Eastern Europe and 44. Aldershot, Hampshire, United Kingdom: CIS Countries, United Nations Economic Ashgate Publishing Limited. Commission for Europe, 17–37. Geneva: Jashi, C. 2005. Gender Economics Issues: The Case UNECE. of Georgia. Tbilisi: United Nations Develop- Mansoor, A., and B. Quillin. 2007. Migration and ment Programme and Swedish International Remittances: Eastern Europe and the Former So- Development Agency. viet Union. Washington, DC: World Bank. Jianakoplos, N. A., and A. Bernasek. 1998. “Are McKenzie, D. 2009. “Why Is More Capital Not Women More Risk-Averse?� Economic Inquiry Enough to Grow Female Businesses.� Finance 36 (4): 620–30. & PSD Impact Note 4, World Bank, Washing- Jurajda, S. 2003. “Gender Wage Gap and Segrega- ton, DC. tion in Enterprises and the Public Sector in McKenzie, D., S. D. Mel, and C. Woodruff. 2007. Late Transition Countries.� Journal of Com- “Returns to Capital in Micro Enterprises: parative Economics 31 (2): 199–222 Evidence from a Field Experiment.� Policy ———. 2005. “Gender Segregation and Wage Gap: Research Working Paper 4230, World Bank, An East-West Comparison.� Journal of the Eu- Washington, DC. ropean Economic Association 3 (2–3): 598–607. Miluka, J., and C. Grown. 2010. “Not Just Educa- Karabchuk, T., and A. Krause. 2010. “Determinants tion: Gender Wage Gap in the Albanian La- of Job Stability in East Germany and Russia: bour Markets through Occupational Segre- A Comparative Analysis Using Micro Panel gation, Work Experience, and Child Care.� Data.� Unpublished report, Centre for La- Paper presented at the International Political bour Market Studies, National University Science Association’s International Confer- Higher School of Economics, Moscow. ence, “Is There a European Model of Gover- Kepler, E., and S. Shane. 2007. “Are Male and Fe- nance? A Comparative Perspective,� Luxem- male Entrepreneurs Really That Different?� bourg, March 18–20. Office of Advocacy Working Paper 309 (Sep- Minniti, M., I. E. Allen, and N. Langowitz. 2006. tember), Office of Advocacy, Small Business “Global Entrepreneurship Monitor: 2005 Administration, Washington, DC. Report on Women and Entrepreneur- Kirby, D. A., and A. Watson, ed. 2003. Small Firms ship.� Global Entrepreneurship Research and Economic Development in Developed and Association, London and Wellesley, MA. 102 Opportunities for Men and Women: Emerging Europe and Central Asia http://www.gemconsortium.org/about. tional Conference, “Poverty and Social Inclu- aspx?page=special_topic_women. sion in the Western Balkans,� Brussels, De- Montenegro, C., and M. Hirn. 2009. “A New Dis- cember 14–15. aggregated Set of Labor Market Indicators Omori, Y. 1997. “Stigma Effects of Nonemploy- Using Standardized Household Surveys from ment.� Economic Inquiry 35 (2): 394–416. Around the World.� Background paper pre- O’Neill, J. 2003. “The Gender Gap in Wages, Cir- pared for World Development Report 2009: ca 2000.� American Economic Review 93 (2): Reshaping Economic Geography. World Bank, 309–14. Washington, DC. Özcan, G. B.. 2006. “Djamila’s Journey from Kolk- Muravyev, A., O. Talavera, and D. Schäfer. 2009. hoz to Bazaar: Female Entrepreneurs in Kyr- “Entrepreneurs’ Gender and Financial Con- gyzstan.� In Enterprising Women in Transition straints: Evidence from International Data.� Economies, ed. F. Welter, D. Smallbone, and N. Journal of Comparative Economics 37 (2): B. Isakova, 93–118. Aldershot, Hampshire, 270–86. United Kingdom: Ashgate Publishing Lim- Narain, S. 2006. “Mainstreaming Gender in First ited. Microfinance Bank Kabul.� Unpublished Paci, P. 2002. “Gender in Transition.� Report (May working paper, International Finance Corpo- 21), World Bank, Washington, DC. ration, Washington, DC. Paci, P., and B. Reilly. 2004. “Does Economic Lib- ———. 2009. “Gender and Access to Finance.� eralization Reduce Gender Inequality in the Analytical working paper, World Bank, Wash- Labor Market: The Experience of the Transi- ington, DC. tion Economies of Europe and Central Asia.� Nearchou-Ellinas, L., and I. S. Kountouris. 2004. Report, World Bank, Washington, DC. “Women Entrepreneurs in Cyprus: A New Paul, S., and S. Sattar. 2008. “What Accounts for Dynamic in Cyprus Economy; Measures to the Lower Participation of Women in En- Integrate Women Refugees in the Society in trepreneurship? Evidence from Transition Cyprus and to Promote a Multi-cultural So- Countries.� World Bank, Washington, DC. ciety.� Women in Management Review 19 (6): Phelps, E. 1972. “The Statistical Theory of Racism 325–32. and Sexism.� American Economic Review 62 Nolte, E., M. McKee, and A. Gilmore. 2004. “Mor- (4): 659–61. bidity and Mortality in Transition Countries Posadas, J., and M. Vidal-Fernández. 2011. “Grand- in the European Context.� Background paper, parents’ Childcare and Female Labor Force thematic session “Morbidity, Mortality, and Participation.� Unpublished working paper, Reproductive Health: Facing Challenges in Institute for the Study of Labor, Bonn. Transition Countries,� European Population Quillin, B., C. Segni, S. Sirtaine, and I. Skamnelos. Forum, Geneva, January 12–14. 2007. “Remittances in the CIS Countries: A Ñopo, H., N. Daza, and J. Ramos. 2011. “Gender Study of Selected Corridors.� Chief Econo- Earnings Gaps in the World.� Background mist’s Regional Working Paper 2 (2), Europe paper prepared for World Development Re- and Central Asia Region, World Bank, Wash- port 2012: Gender Equality and Development, ington, DC. World Bank, Washington, DC. Ramalho, R., J. Rodríguez-Meza, and J. Yang. 2009. Oglobin, C. 2005. “The Gender Earnings Differ- “How Are Firms in Eastern and Central Eu- ential in Russia after a Decade of Economic rope Reacting to the Financial Crisis?� Enter- Transition.� Applied Econometrics and Interna- prise Note 8, Enterprise Surveys, Enterprise tional Development 5 (3): 5–26. Analysis Unit, World Bank, Washington, DC. Ognjenović, K. 2010. “Wage Differences in Serbia: Rankin, G., and S. Narain. 2009. “Centre for the An Application of Quantile Regressions.� Provision of Credit–Financial Advice, Bos- Paper presented at the World Bank’s Interna- nia and Herzegovina Feasibility Study for References 103 the Provision of Preventative and Remedial ———. 2009b. “2009 World Survey on the Role Measures for Indebtedness in Microfinance of Women in Development: Women’s Con- Clients.� Unpublished report, World Bank, trol over Economic Resources and Access to Washington, DC. Financial Resources, Including Microfinance.� Reva, A., and V. Sulla. 2010. “Gender Inequality in Document ST/ESA/326, Division for the the Labor Market in Serbia.� Unpublished re- Advancement of Women, Department of port, World Bank, Washington DC. Economic and Social Affairs, United Nations, Ruhm, C. J. 1998. “The Economic Consequences New York. of Parental Leave Mandates: Lessons from Eu- UNECE (United Nations Economic Commission rope.� Quarterly Journal of Economics 113 (1): for Europe). 2002. Women’s Entrepreneurship 285–317. in Eastern Europe and CIS Countries. Geneva: Sabarwal, S., and K. Terrell. 2008. “Does Gender UNECE. Matter for Firm Performance: Evidence from UNECE (United Nations Economic Commission Eastern Europe and Central Asia.� Policy for Europe) and World Bank Institute. 2010. Research Working Paper 4705, World Bank, Developing Gender Statistics: A Practical Tool. Washington, DC. Report ECE/CES/8. Geneva: United Na- Sabarwal, S., N. Sinha, and M. Buvinic. 2011. “How tions. Do Women Weather Economic Shocks? A USAID (United States Agency for International Review of the Evidence.� Economic Premise Development). 2011. “Global Partnerships.� 46 (January), World Bank, Washington, DC. USAID, Washington, DC. http://www.us- Scandura, T. A., and M. Lankau. 1997. “Relation- aid.gov/our_work/global_partnerships/gda/ ship of Gender, Family Responsibility and egat_guide/egat1.html. Flexible Work Hours to Organizational Com- Welter, F., and S. Kolb. 2006. “Women and Entre- mitment and Job Satisfaction.� Journal of Or- preneurship in Latvia.� TeliaSonera Institute ganizational Behavior 18 (4): 377–91. Discussion Paper 4 (May), Stockholm School Schultz, P. 1990. “Women’s Changing Participation of Economics in Riga, Riga, Latvia. in the Labor Force: A World Perspective.� Eco- Welter, F., D. Smallbone, D. Mirzakhalikova, N. J. nomic Development and Cultural Change 38 Schakirova, and C. Maksudova. 2006. ‘Wom- (3): 457–88. en Entrepreneurs between Tradition and Mo- Shariari, H., and A. Danzer. 2008. “A Gender Per- dernity.� In Enterprising Women in Transition spective on Access to Land and Finance in Ta- Economies, ed. F. Welter, D. Smallbone, and jikistan.� World Bank, Washington, DC. N. B. Isakova, 45–66. Aldershot, Hampshire, Sondergaard, L. M., and M. Murthi. 2012. Skills, United Kingdom: Ashgate Publishing Lim- Not Just Diplomas: Managing for Results in ited. Education Systems in Eastern Europe and Cen- WHO (World Health Organization). 2010. World tral Asia. Directions in Development Series. Health Statistics 2010. Geneva: WHO. Washington, DC: World Bank. World Bank. 2005a. Dying Too Young: Addressing Tominc, P., and M. Rebernik. 2006. “Gender Dif- Premature Mortality and Ill Health Due to ferences in Early-Stage Entrepreneurship In Non-Communicable Diseases and Injuries in the Three European Post-Socialist Countries.� Russian Federation. Washington, DC: World Drustvena Istrazivanja: Journal for General Bank. Social Issues 16 (3): 589–611. Institute of So- ———. 2005b. Enhancing Job Opportunities: East- cial Sciences Ivo Pilar, Zagreb, Croatia. ern Europe and the Former Soviet Union. Wash- UN (United Nations). 2009a. World Population ington, DC. Prospects: The 2008 Revision. CD-ROM. New ———. 2006. Turkey: Labor Market Study. Report York: Population Division, Department of 33254-TR, April 14. Washington, DC: World Economic and Social Affairs, United Nations. Bank. 104 Opportunities for Men and Women: Emerging Europe and Central Asia ———. 2009. Business Environment in Tajikistan as ———. Forthcoming. “In from the Shadow: In- Seen by Small and Medium Enterprises. Invest- tegrating Europe’s Informal Labor.� World ment Climate Advisory Services, Internation- Bank, Washington, DC. al Finance Corporation, Dushanbe, Tajikistan. World Bank and IFC (International Finance Cor- ———. 2010. “Roma Inclusion: An Economic poration). 2009. Doing Business 2010: Reform- Opportunity for Bulgaria, Czech Republic, ing through Difficult Times. Washington, DC: Romania, and Serbia.� ECA Policy Note, Sep- World Bank; New York: Palgrave MacMillan. tember 30, Human Development Unit, Eu- Zhelyazkova, N., and M. Valentova. 2009. “Wom- rope and Central Asia Region, World Bank. en’s Perceptions of Consequences of Career ———. 2011a. World Development Report 2012: Interruption Due to Childcare in Central Gender Equality and Development. Washing- and Eastern Europe.� IRISS Working Paper ton, DC: World Bank. 2009–01, Integrated Research Infrastructure ———. 2011b. “Regional Strategy Update 2011: in Social Sciences, Centre d’Etudes de Popu- Europe and Central Asia Region.� Slide show lations, de Pauvreté et de Politiques Socio- presentation, World Bank, Washington, DC. Economiques–International Networks for ———. 2011c. “Project Appraisal Document, Ro- Studies in Technology, Environment, Alterna- mania: Social Assistance System Moderniza- tives and Development, Luxembourg.^9- tion Project.� Report 58280-RO, World Bank, Washington, DC. References 105 The world bank 1818 H Street, NW Washington, DC 20433