TURKEY COUNTRY GENDER ASSESSMENT 2017 The World Bank January, 2018 TURKEY GENDER COUNTRY ASSESSMENT 2018 Table of Contents SUMMARY ..................................................................................................................................... 6 CHAPTER 1: ENDOWMENTS .......................................................................................................... 8 1.1 Health ................................................................................................................................ 8 1.2 Education......................................................................................................................... 14 CHAPTER 2: ECONOMIC OPPORTUNITIES ................................................................................... 21 2.1. Female labor market participation.................................................................................. 21 2.2. Employment and outcomes in the labor market ............................................................ 26 2.3. Gender gap in earnings ................................................................................................... 30 2.4. Women’s entrepreneurship and access to finance ........................................................ 32 2.5. Welfare and Income ........................................................................................................ 35 CHAPTER 3: AGENCY ................................................................................................................... 39 3.1. Social and gender norms ................................................................................................. 39 3.2. Violence against women ................................................................................................. 43 3.3. Women’s political participation ...................................................................................... 50 3.4. The legal, institutional and policy framework for gender equality ................................. 52 CHAPTER 4: CONCLUSIONS AND POLICY RECOMMENDATIONS ................................................. 58 REFERENCES ................................................................................................................................ 60 ANNEX ......................................................................................................................................... 63 2 TURKEY GENDER COUNTRY ASSESSMENT 2018 List of figures Figure 1.1: Share of Population by Age Groups and Gender in 2014, percent ............................................ 9 Figure 1.2: Life expectancy at birth, years .................................................................................................... 9 Figure 1.3: Regional differences in gender gaps in life expectancy ............................................................ 10 Figure 1.4: Main cause of death by sex, and prevalence of health problems in old age by sex ................ 11 Figure 1.5: Population increases and fertility rates .................................................................................... 12 Figure 1.6: Married women use of contraception (ever used and current use) ........................................ 13 Figure 1.7: Maternal, infant and under-5 child mortality........................................................................... 13 Figure 1.8: Gross enrollment in Primary, Secondary and Tertiary Education in Turkey and ECA region average ....................................................................................................................................................... 14 Figure 1.9 Educational attainment, population over 25 years old (percentage) ....................................... 14 Figure 1.10 Enrollment in Higher Education Among Population Aged 17-24 by Income per capita Quartiles in 2014 ........................................................................................................................................................ 16 Figure 1.11: Regional differences in higher education gender gaps .......................................................... 17 Figure 1.12 Tertiary education fields of study, female ............................................................................... 19 Figure 2.1 Male and Female Labor Force Participation, percent ............................................................... 21 Figure 2.2: Women’s participation in the labor market by cohort ............................................................. 22 Figure 2.3: U-shaped relationship between GDP per capita and FLFP ....................................................... 22 Figure 2.4 Regional gaps in female and male labor force participation ..................................................... 23 Figure 2.5 Reasons for inactivity by sex and educational level .................................................................. 25 Figure 2.6 Share of regular employees, employers, self-employed and unpaid-family-workers across gender (percent), 2013 ............................................................................................................................... 27 Figure 2.7: Employment Rates across Gender, Age and Education in 2015 ............................................... 30 Figure 2.8: Log Monthly Wage Across Gender, 2015 ................................................................................. 31 Figure 2.9 Female Management and Ownership of Firms in 2013, percent .............................................. 33 Figure 2.10: Female Ownership and Management by Economic Sectors in 2013, percent ....................... 34 Figure 2.11: Use of Bank Accounts in Turkey and the ECA Region, 2014 ................................................... 35 Figure 2.12: Purposes of Accounts and Sources of Loans in Turkey, 2014 ................................................. 35 Figure 2.13: Share of male and female headed households in population by age groups, percent .......... 36 Figure 2.14: Share of men and women headed households among single person and single parent families by age groups, percent ............................................................................................................................... 36 Figure 2.15: Monthly income per capita by gender of head of household and the type of household ..... 37 Figure 2.16: Monthly income per capita by gender of head of household and age group among single person and single parent households ........................................................................................................ 37 Figure 2.17: Income per capita in female headed single parent households by the number of children below 14 years, 2013 ................................................................................................................................. 37 Figure 3.1: Gender related social norms - Mean Value of Indicators Showing Agreement with the Views on Gender Related Statements Across Gender, 2008 ..................................................................................... 41 Figure 3.2: Views on Gender Related Statements Across Men and Women and Age Groups, percent of Agreement, 2008 ........................................................................................................................................ 42 Figure 3.2: Physical and/or sexual violence by a partner in Turkey ........................................................... 44 Figure 3.3 Prevalence of physical and/or sexual violence by a partner in the 12 months prior to the interview in Turkey, ever married women aged 15-59, by provinces ........................................................ 45 Figure 3.4 Estimated probabilities of experiencing physical and/or sexual violence by a partner in the 12 months prior to the survey ......................................................................................................................... 46 Figure 3.5 Share of ever married women that agree that domestic violence by the husband is justified if the woman neglects the children, argues with him, refuses sex or burns the food .................................. 47 3 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 3.6 Percentage of ever married women having experienced controlling behavior by their husbands .................................................................................................................................................................... 48 Figure 3.7: Percentage of ever-married women (15-59) who reported having being subject to d controlling behaviors by their husbands or intimate partners (lifetime prevalence by type of behavior) .................. 49 Figure 3.8 Women in political representation positions in Turkey ............................................................ 51 Figure 3.9: Policy and institutional framework for gender equality in Turkey ........................................... 56 List of boxes Box 1.1: The Impact of the Extension in Compulsory Education in Turkey ................................................ 15 Box 1.2: Increasing Schooling Rates Especially for Girls ............................................................................. 18 Box 1.3: Promoting Gender Equality in Education ..................................................................................... 19 Box 2.1. Programs Promoting Female Labor Market Inclusion in Turkey .................................................. 26 Box 3.1. Early and Arranged Marriages in Turkey ...................................................................................... 39 List of tables Table 2.1: Women’s Representation in Employment and Earnings by Sector, 2015 ................................. 28 Table 2.2: Oaxaca Decomposition of Monthly Wages (with Heckman correction for selection), 2015 .... 32 Table 3.1: Distribution and Proportion of Women Who Have Experienced Physical and/or Sexual Violence by Their Husband in the 12 Months before the Survey (married women 15 –59) ..................................... 45 Table A1. Ordered Probability Model for Life Satisfaction, 2008 ............................................................... 63 Table A2. Returns to Education and Experience, 2015 ................................. Error! Bookmark not defined. 4 TURKEY GENDER COUNTRY ASSESSMENT 2018 Acronyms BEEPS Business Environment and Enterprise Surveys BMI Body mass Index CEDAW Convention on the Elimination of All Forms of Discrimination against Women CKA Cukurova Development Agency DGSW Directorate General on the Status of Women EBRD European Bank for Reconstruction and Development ECA Europe and Central Asia EU European Union FRA EU Agency for Fundamental Rights HBS Household Budget Survey LFS Labor Force Survey MoD Ministry of Development MoFSP Ministry of Family and Social Policies MoI Ministry of Interior MoLSS Ministry of Labor and Social Security MoNE Ministry of National Education MFAL Ministry of Food, Agriculture and Livestock NCD Non-communicable diseases OAW Own-account worker OECD Organization for Economic Cooperation and Development PTP Priority Transformation Programs SILC Income and Living Conditions Survey STEM Science, Technology, Engineering and Math TDHS Turkish Demographic and Health Survey TDVS National Research on Domestic Violence against Women in Turkey UNICEF United Nations International Children´s Emergency Fund WDI World Development Indicators WHO World Health Organization 5 TURKEY GENDER COUNTRY ASSESSMENT 2018 SUMMARY1 Gender equality is a core development objective in its own right, and it is also a smart development policy. Gender equality is also a key pathway to ensure lasting poverty reduction and shared prosperity.2 Identifying the main gender gaps a country faces across different domains, contributes to better inform policy design. To that effect, this report seeks to identify where progress has been achieved with regard to increasing opportunities for women and men in Turkey and where further policy action is required. It focuses on three areas that are critical for gender-equal access to opportunities, namely endowments such as health and education; economic opportunities, such as access to labor, land and financial markets; and agency, including norms, representation, and freedom from violence (World Bank 2016).3 The report takes advantage of different sources of publicly available data for the country, including the World Bank’s World Development Indicators (WDI), the Global Financial Inclusion (FINDEX) Database, the Business Environment and Enterprise Performance Survey (BEEPS), the World Health Organization (WHO) Statistics, as well as data from the Organization for Economic Co-operation and Development (OECD), and others. National surveys by the Turkish Statistical Institute include the Income and Living Conditions Survey (SILC) 2015, the Turkish Labor Force Survey (LFS) 2016 and the Household Budget Survey (HBS). In addition, the Ministry of Family and Social Policies (MoFSP) and Hacettepe University Institute of Population Studies conducted the National Research on Domestic Violence against Women in Turkey. The Turkish Demographic and Health Survey (TDHS) collected by the Ministry of Development (MoD) and Hacettepe University Institute of Population Studies, and the Research on Family Structure in Turkey (TAYA) by MoFSP were also used. The report aims to provide a panorama of the prevailing gender gaps and areas for work to close those gaps in the country, covering a wide range of outcomes. As such, it seeks to serve as a guiding document for policy action and dialogue, further research, and public discussion. Gender equality is increasing in many outcomes in Turkey, however the current levels have plenty of room for progress in several areas, particularly in educational attainment, participation in the labor market and overall access to economic opportunities, when compared with the situation in countries with similar levels of income per capita. Between 2008 and 2015 the maternal mortality ratio halved, falling from 32 to 16 per 100,000 live births. Net female secondary enrollment rates grew from 76 percent in 2008 to 86 percent in 2015 almost approaching gender parity. More women enroll in university every year (88 percent gross enrollment ratio in 2015 compared with 35 percent in 2008), although they still enroll at lower rates than men (gross enrollment ratio for men was 101 percent in 2015). Female labor force 1 This report was produced by Ana Maria Munoz Boudet (Senior Social Scientist, GPVDR), Lourdes Rodriguez Chamussy (Economist, GPVDR), Cristina Chiarella, Carmen de Paz, Paola Buitrago, and Isil Oral (Consultants, GPVDR). The team benefited from the comments and reviews from different agencies of the Turkish Government (via comments provided to the Ministry of Finance), as well as from reviewers including Arzu Uraz (Social Development Specialist, GPSURR), Facundo Cuevas (Senior Economist, GPVDR), and Meltem Aran (Development Analytics). 2 World Bank Gender Strategy 2016–2020. 3 As specified by the World Development Report 2012 on Gender Equality and Development, and the World Bank Gender Strategy 2016–2020. 6 TURKEY GENDER COUNTRY ASSESSMENT 2018 participation has increased to 32,5 percent in 2016 from 24,5 percent in 2008. However it represents about half of the female labor force participation in the Europe and Central Asia (ECA) region or the OECD. Despite progress over the recent decade, gender differences are still a challenge, and are frequently wider among vulnerable groups. Girls from lower-income backgrounds have significantly lower school enrollment rates than boys, while among the higher-income bracket gender parity has been achieved. Whereas gender gaps in educational attainment have decreased in the western parts of the country, they remain substantial in the eastern regions. Female labor force participation also shows marked gradients with respect to income brackets. Increasing women’s economic participation and addressing disadvantages among vulnerable groups and regions remain two main challenges for Turkey. Gender gaps in participation in the labor market and entrepreneurship are common across countries and entail substantial economic losses for women and their families in the form of foregone income, but also losses in aggregate terms, for the societies where women are deprived of those opportunities (World Bank 2012). Estimates suggest the loss associated with the gender gap in labor participation reaches 22 percent of income per capita (Cuberes and Teignier 2015). This is especially relevant in the current demographic context of Turkey in which higher participation of all those in the economically active age group represents a unique window of opportunity for growth and savings. If the trend toward rising female labor force participation is sustained and women from lower-income brackets join the labor force, this will, over time, make a significant contribution to reduced poverty and inequality. As trends in other regions suggests—for example, Latin America and the Caribbean over the past decade—women can play a key role in strengthening the links between access to jobs and poverty reduction and shared prosperity. The current pace of increase in Turkey would need to accelerate to close the existing gap with comparable countries, and allow the country to reap the benefits and returns of its human resources. Supporting the equality of women and men is a smart development strategy for Turkey. Policy efforts must encompass the several fronts on which barriers to gender equality persist, from influencing norms to ensuring equal access to opportunities. Working to remove the barriers to women’s paid work is fundamental to increase their economic opportunities and the prospects of the overall economy. This would involve facilitating a beneficial school-to-work transition by tailoring the content of higher education more closely to the needs of the private sector and thus making university education more relevant for the job market. Investing in active labor market programs tailored to men and women’s needs and vocational training would help closing the gender gap in the labor market. Business-oriented education and financial inclusion will help strengthen women’s involvement in entrepreneurial activities. Better childcare options constitute another fundamental policy to help parents combine work and the care of young children. Two aspects of childcare are fundamental to successful programs: quality and convenience. Building up the quality and availability of kindergartens to meet the demand for preschool enrollments is therefore an important policy priority and one 7 TURKEY GENDER COUNTRY ASSESSMENT 2018 that can not only expand women’s opportunities but simultaneously address equality and efficiency challenges. CHAPTER 1: ENDOWMENTS Investments in human capital have a demonstrated a large impact on the capability of individuals to benefit from life-long opportunities and to make a positive social and economic contribution. Differences in basic endowments such as health and education between men and women, especially at an early age, can thus lead to the perpetuation of gender gaps in access to opportunity throughout the life cycle. The persistence of unequal opportunities with regard to such a large share of the population that women represent entails similarly large costs, not only for individuals and families, but also for societies in the long term. Evidence also shows that these gaps can play a relevant role in the inter-generational transmission of gender inequalities (World Bank 2012) and bear substantial costs for societies. This chapter analyzes the situation of women and men in Turkey with regard to their access to and accumulation of basic endowments such as health and education. 1.1 Health In contrast with other countries in the Europe and Central Asia (ECA) region, Turkey is a young country. that has recently entered the last stage of demographic transition, namely, a falling fertility and mortality rate at all ages, and a substantive young working-age population bulge, placing the country at the ‘demographic window of opportunity’.4 This youth bulge could yield important benefits in the development and growth process of the country, but compared with now ‘old’ European economies, Turkey’s window is predicted to be shorter, as the demographic transition in the country will accelerate—a recent report predicts that the average population age in Turkey of about 28 years will increase by 7 years over the next three decades, a pace more accelerated than the one observed in Europe (World Bank 2015). While currently enjoying the benefits of the youth bulge, Turkey’s focus going forward will need to be twofold: promote healthy aging by tackling specific morbidities and risk factors that affect men and women differently, as well as continue focusing on maternal and reproductive health, with attention to declining fertility rates. Life expectancy has increased but gender-specific morbidities are a future challenge The Turkish population has been increasing at a constant growth rate of 14 percent on average per year, driven by increases in life expectancy and constant fertility rates. Like other countries in Europe, life expectancy has increased both for women and for men with regard to gains in years, but on average women are expected to live 7 years more than men (Figure 1.2), a gap that has remained constant over the last 10 years. For both men and women, life expectancy remains 2 years lower than the ECA average, and 5 years lower than the average for 4 The United Nations Population Division defines the window of opportunity as a period in which the ratio of under 15 population is less than 30 percent of the total population and the ratio of 65 years and older population is under 15 percent of the total population. 8 TURKEY GENDER COUNTRY ASSESSMENT 2018 OECD countries. An average Turkish newborn girl has the chance of living an additional 7.5 years if born in 2014 as compared with 1995, and a newborn boy has 8.8 additional years.5 Figure 1.1: Share of Population by Age Groups and Gender in 2016, (thousands) 75+ 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5-9 0-4 (4,000) (3,000) (2,000) (1,000) 0 1,000 2,000 3,000 4,000 Women Men Source: TURKSTAT 2016 Figure 1.2: Life Expectancy at Birth, Years 82 2005 2015 80 78 76 74 years 72 70 68 66 64 62 Turkey ECA Turkey ECA female male Source: WDI 2016 Average gains in life expectancy mask wide variations across regions in the country. Figure 1.3 depicts the existing differences with regard to gender gaps in life expectancy. Although gender gaps in life expectancy tend to be higher in some lower-income regions in the 5 World Bank Health Nutrition and Population Statistics. October 2016. In the same period of time, Turkey increased its per capita health expenditure in 3 percentage points (GDP share), a similar increase than the one observed for OECD countries, although Turkey still spends less than half of these countries do in health. Gains in life expectancy among women can also be linked to improvements in tacking infant mortality rates and under five mortality rates, which have significantly gone down in the country. 9 TURKEY GENDER COUNTRY ASSESSMENT 2018 east, they are also large in some western wealthier regions; with the traditional rural-urban and east-west divide in outcomes and in access to services and resources being less stark when it comes to life expectancy. Although geographical differences in the health status of both Turkish men and women (measured by self-reported health and disability prevalence) have been found to be associated with their educational level and the wealth of their household (Ergin and Kunst 2015), these are less reflected in life-expectancy gaps, but in their absolute levels for both women and men. However, an evaluation of the Turkey Health Transformation Program (for the period 1995-2008) shows that the association between women’s background variables and health care utilization –particularly in relation to maternal health- weakened after the program’s introduction (Atun et al 2013), particularly thanks to improvements in the supply of health services (see also Aran et al 2015). Figure 1.3: Regional Differences in Gender Gaps in Life Expectancy Source: World Bank team’s based on TURKSTAT data, 2016 The observed gender gap in life expectancy in Turkey is largely explained by differences in adult mortality rates between men and women. Adult mortality rates are significantly higher among men than women in Turkey—almost double at 142 per 1,000 adults for men compared with 73 per 1,000 for women—both levels in line with observed ones for the Europe and Central Asia region. The most common cause of death for both men and women are cardiovascular diseases. However, men register higher death rates from cardiovascular diseases than women (384 per 100,000 people for men compared to 256 for women), along with much higher death rates from malignant neoplasms and chronic respiratory diseases (Figure 1.4b). However, reported mortality from coronary heart disease among Turkish women was the highest in Europe (WHO 2014). For women, gains in life expectancy do not always mean gains in healthy life expectancy. While in Turkey the risk of death by non-communicable diseases (NCDs) is not particularly high compared with other countries, the burden of disease, and NCDs impact on healthy aging is important. As of 2012, the majority of the disability-adjusted life years (DALYs) losses estimated 10 TURKEY GENDER COUNTRY ASSESSMENT 2018 in the country could be explained by NCDs, in particular, related to cardiovascular disease. (Figure 1.4a) Further insights with regard to risk factors point to the change of eating habits (according to health data by TURKSTAT in 2014, 38 percent of men and 30 percent of women were overweight based on body mass index [BMI]), smoking—although on the decline—and other lifestyle-related behaviors. In addition, the share of women reporting age-related difficulties, particularly with regard to mobility (walking or walking up or down stairs) unaided, or remembering things, almost doubles men’s reports of the same problems, while there are much smaller differences in age-related vision or hearing problems (Figure 1.4b). Many morbidities are lifestyle related, and not only with risk behaviors such as smoking or drinking. For example, Etiler (2015) found that although the physical health status of women who are out of the labor force (housewives) and working women is similar, there are significant differences in mental health, with depression being higher among housewives. And the same difference applies to other morbidities such as hypertension and diabetes, which are more frequent among housewives. Further analysis of the connections between lifestyles and morbidities, and how to prevent NCDs and other specific health issues via changes in lifestyle would be important as Turkey’s population ages. Figure 1.4: Main Cause of Death, by Sex; and Prevalence of Health Problems in Old Age, by Sex a. Age-standardized death rates per 100,000, by cause b. Main disability-related problems among the and sex, 2012 elderly, by age and sex Problems/age groups Male Female 800 Vision problems 700 65–74 15.1 19.8 600 75+ 23.1 32.5 Total Male Female Hearing problems 500 65–74 15.4 14.1 400 75+ 37.7 34.6 300 Difficulty in walking and walking up or down stairs without using 200 any aid or assistance 100 65–74 15.3 31.9 75+ 38.3 59.7 0 All causes Cardiovascular Malignant Chronic Difficulty in remembering diseases neoplasms Respiratory 65–74 6.3 12.0 diseases 75+ 15.6 27.3 Source: WDI and TURKSTAT Turkey Health Interview Survey Fertility has seen a slight decline as reproductive health has improved While in the past 15 years the Turkish population has gained 14 million people, fertility rates have stayed almost constant at about 2 children per mother on average since the decline observed in the early 2000s (Figure 3.1). The lack of change in the fertility rate is also explained by the desired fertility rate; according to TDHS the desired mean number of children has stayed at around 2.6 children per woman since 2003, with some differences between regions, with the east having a mean ideal number of children of 3.1 during the same period. And while 71.6 percent of women with one child want to have another child, 60 of women with two or more children do not want to have more children. 11 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 1.5: Population Increases and Fertility Rates 100,000,000 3 80,000,000 2.5 2 60,000,000 1.5 40,000,000 1 20,000,000 0.5 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Population Fertility rate Source: TURKSTAT 2016 The changes in desired fertility are not explained by delays in age of marriage by women or by changes in the overall number of marriages in the country, and it is similar across education levels. Greulich et al. (2015) analyze the potential determinants of women’s and their partners’ decision to continue or stop having children. They identify women with higher education— tertiary or above—with labor market attachment, to be the group with the highest likelihood to stop fertility at two children, with a particular emphasis of marriages where men are not in stable employment. In European countries that have gone a strong fertility decline, the re-increase in fertility is strongest in countries providing institutional support to work-life balance, particularly where child care coverage is high (Greulich and Thévenon 2013, 2014), and where both partners are in the labor market.6 The prevalence of contraceptive use is high in Turkey, ranging from a lifetime use among married women of 92 percent to 73 percent of current use among married women. However, while the rate of use of modern techniques has increased by 12 percentage points, the use of traditional methods has stayed at around 27 percent among current users after a decline was observed in the 1990s (Figure 1.6); unmet need for contraceptives remains low at 6 percent in 2013 compared with 15 percent in 1993-987. Women report that the public sector is the main provider of contraceptives. The positive changes observed in access to modern methods can be attributed to government-level efforts such as the National Strategic Action Plan on Women’s Health and Family Planning in place between 1996 and 2000 and revised for the 2005–2015 period, focusing on the reduction of maternal deaths and the prevention of unwanted pregnancies through addressing the unmet needs in contraceptive services.8 6 In Turkey, almost 80 percent of children live in households with a single earner —this is the largest proportion among European countries (EU average 40 percent) 7 Unmet need is defined as currently married, fecund women who either want to postpone the next birth (need for spacing) or who want no more children (need for limiting), but who are not using a contraceptive method. As a reference, the median among all countries most recent round of Demographic and Health Surveys is at 19.6 percent. With Turkey being lower than levels registered for countries such as Peru and Colombia (between 8 and 9 percent), or Indonesia (11 percent) 8 The 2005 revision changed the plan name to the National Strategic Plan for the Health Sector on Sexual Health and Reproductive Health. 12 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 1.6: Married Women Use of Contraception (ever used and current use) 100 90 80 70 60 29 27 26 38 26 50 40 30 20 43 46 47 35 38 10 0 1993 1998 2003 2008 2013 Current use-modern methods Current use-traditional methods Ever used Source: TDHS The decline in fertility and increase in contraceptive use has come along with a strong decline in maternal and child mortality, largely thanks to relevant health sector improvements to increase supply of health care with a strong focus on maternal and child chld care, improvements in the provision of health services, and demand side interventions that make health services more accessible (Aran et al 2015). Maternal mortality decreased from 97 to 14 per 100,000 live births from 1990 to 2015, with a steep decline starting in year 2000 (Figure 1.7a), owing to strong deliberate government policy, reaching the average level for OECD countries by 2015. This trend comes paired with a decline of home births, from 16 to 2 percent between 2003 and 2013.9 Alongside this decline, there was also a 65 percent reduction in infant and child mortality during the same period. However, under-5 female mortality and under-5 male mortality are still higher than for ECA—by 24 and 17 percent, respectively (see Figure 1.7b). Figure 1.7: Maternal, Infant, and Under-5 Child Mortality a. Maternal mortality rate (per 100,000 births) in b. Infant and under-5 child mortality in Turkey Turkey, ECA, and OECD, 1990–2015 and ECA region in 2000 and 2015 100 45 2000 2015 90 40 80 35 70 30 60 25 50 20 40 15 30 10 20 5 10 0 0 1990 1995 2000 2005 2010 2015 Infant U-5 U-5 Infant U-5 U-5 female male female male Turkey ECA OECD Turkey ECA Source: WDI 9 According to TDHS data for ever married women 15 –49 years of age. 13 TURKEY GENDER COUNTRY ASSESSMENT 2018 1.2 Education Enrollment rates have increased for both girls and boys but gender disparities persist Turkey has substantially narrowed the existing gender gaps in education over the last decades, especially at the primary level where gender parity in enrollment already exists. Although differences persist at the secondary and tertiary levels, enrollment rates have increased significantly for both male and female students and especially so for women since 2008 (see Figure 1.8). Turkey continues to show higher gender gaps than other countries with similar income levels (World Bank 2015), however the positive trends in levels of enrollment compare favorably to those observed for the ECA region on average; for example, 80 percent of women in the relevant age range enroll in tertiary education, compared to 70 percent on average for the ECA region. Figure 1.8: Gross Enrollment in Primary, Secondary, and Tertiary Education in Turkey and ECA Region Average 120 female 100 male 80 60 40 20 0 Turkey ECA Turkey ECA Turkey ECA Turkey ECA Turkey ECA Turkey ECA Primary (% Primary (% Secondary (% Secondary (% Tertiary (% Tertiary (% gross), 2008 gross), 2014 gross), 2008 gross), 2014 gross), 2008 gross), 2014 Source: WDI, 2017. Note: Enrollment rates can exceed 100 percent due to the inclusion of overage and underage students because of early or late school entrance and grade repetition. As expected, given the recent evolution of enrollment rates, gender gaps in attainment among the population older than 25 years persist at the higher levels. More women than men above 25 years have completed primary education, while the opposite is true of lower- and upper-secondary, and tertiary education (bachelor’s or equivalent). Despite improvements over time for both groups, the gender gap has remained largely unchanged to date, and has even slightly increased for lower-secondary education (Figure 1.9). The expansion in the number and capacity of the universities as a consequence of the changes to the law implemented in 2008 may have contributed to increase enrollment and attainment in higher education for both men and women (Yilmaz 2014), although the impact on quality in university degrees is ambiguous. In turn, the impacts on gender equality of the 2012 reform of the Fundamental Law of National Education are not yet fully clear. Among other measures, the reform extended the duration of compulsory education from 8 to 12 years. The existing evidence suggests that this measure could have a mixed effect with regard to gender 14 TURKEY GENDER COUNTRY ASSESSMENT 2018 parity in education, given the dynamics and existing gender biases in the labor market and the corresponding incentives to schooling decisions (see Box 1.1). Figure 1.9: Educational Attainment, Population Over 25 Years (percentage) 60 50 2005 40 2010 2014 30 20 10 0 Female Male Female Male Female Male Female Male Female Male Primary Lower secondary Upper secondary Bachelor´s or Master´s or equivalent equivalent Source: WDI 2016 Box 1.1: The Impact of the Extension in Compulsory Education in Turkey Owing to the 2012 reform change that increased compulsory schooling to 12 years, Dincer and Erten (2015) find evidence of a 3.8 percent increase in high school attendance on average and to a large reduction in child labor among boys. For girls, on the other hand, the increase in schooling seems to be accompanied by a reduction in idleness, given the different characteristics of the child labor based on gender and due to the presence of underemployment. Dincer and Erten (2015) also point to a higher prevalence of distance education among girls that appears to have driven a significant decrease in idleness, and thus a smaller tradeoff between school attendance and child labor in services. In relation to previous reforms, available evidence suggests that the extension of compulsory schooling from 5 to 8 years, established in 1997, led to higher number of completed years of schooling with spillovers beyond the compulsory age —for example, the completed years of schooling by age 17 increased by 1.5 years for rural women. The reform seems to have had important equalizing effects on the educational attainment of urban and rural children. According to the study of Kirdar, Dayioglu- Tayfur, and Koc (2014), the urban-rural gap in the completed years of schooling at age 17 fell by 0.5 years for men and by 0.7 years for women. However, no evidence of an effect of the policy was found on narrowing the gender gap. That is probably related to the stronger increases in labor market earnings associated with the completion of a diploma for men due to their much higher labor force participation rate. With regard to gaps in performance, the most recent results from the Programme for International Student Assessment (PISA) indicate that the education system in Turkey has accomplished important progress improving scores and mitigating the associations between 15 TURKEY GENDER COUNTRY ASSESSMENT 2018 family background and educational success between 2003 and 2012. Girls on average have higher scores in mathematics and sciences according to the Trends in International Mathematics and Science Study (TIMSS) 2011 and 2015. However, this advantage does not translate into women entering Science, Technology, Engineering, and Math (STEM) fields of study and careers (this is analyzed in more detail later in this chapter). Low-income groups show higher gender gaps in education but girls and women are particularly disadvantaged in some geographical regions Gender gaps in enrollment to post-secondary education decrease and even reverse along the income distribution. Enrollment in higher education increases consistently by income quartile and youth from the poorest quartiles are less likely to continue education after graduating from secondary school. A gender gap exists with regard to youth enrollment, especially in the poorest quartiles: While more men than women are enrolled in post-secondary school in the second quartile, the gap reverses to the advantage of women in the bottom and third quartiles and then decreases with a slight advantage towards women in the top quartile (Figure 1.10). Figure 1.10: Enrollment in Higher Education Among Population Aged 17–24 by Income Per Capita Quartiles in 2015 70 men women 60 50 percent 40 30 20 10 0 Bottom II III Top income per capita quartiles Source: SILC 2015, World Bank staff calculations. Note: A quartile is one of the three points that divide a range of data or population into four equal parts. The first (bottom) quartile contains the poorest individuals, while the last (top) the richest. Income per capita refers to the equivalized household disposable income which is the total income of a household, after tax and other deductions, that is available for spending or saving, divided by the number of household members converted into equalized adults; household members are equalized or made equivalent by weighting each according to their age, using the modified OECD equivalence scale. (TURKSTAT, 2015). Higher education refers to college, Master, or Doctorate levels. Substantial differences can be observed with regard to the gender gap in secondary and tertiary educational attainment across regions. The lower-income eastern and especially southeastern parts of the country show much higher gender gaps for both indicators (Figure 1.11). Gaps in educational attainment between men and women are lowest in urban and wealthier parts of the country including Istanbul and Ankara. Past trends in average years of education for boys and girls follow a similar evolution: Whereas gender gaps decreased in the 16 TURKEY GENDER COUNTRY ASSESSMENT 2018 western parts of the country, such as Ankara, Istanbul, Antalya, Aydın, and Izmir over the 1975– 2000 period, they widened substantially in southeastern regions including Mardin, Van, Sanlıurfa, Agrı, and Malatya (Tomul 2009). In recent years, public policy efforts to continue raising girls’ enrollment have indeed focused on the regions that lag behind, for instance through the implementation of the “Increasing the School Attendance Rate of Girls (KEP I and II)” and the “Vocational Skills Development (MESGEP)” projects (see Box 1.2). To improve enrollment and educational outcomes for girls—especially in the eastern regions, where the gender gap in enrollment rates is still 4 to 8 percent—it is necessary not only to increase the supply of education services but also to promote the attendance of girls. Figure 1.11: Regional Differences in Higher Education Gender Gaps (a) Ratio of the proportion of men with high school education over the proportion of women with high school education (b) Ratio of the proportion of men with tertiary education over the proportion of women with tertiary education Source: TURKSTAT 2016 17 TURKEY GENDER COUNTRY ASSESSMENT 2018 Box 1.2: “Increasing Schooling Rates Especially for Girls” and “Vocational Skills Development” The Ministry of National Education (MoNE) has implemented the project ‘Increasing Schooling Rates Especially for Girls (KEP I and II)’ to increase girls’ schooling rates and to raise awareness among families on the importance of education predominantly in East and Southeast Anatolia regions. The project was launched in 2012 and finalized in 2013. The activities concentrated mainly in the 16 eastern provinces.10 Within the scope of the project,  1,638 teachers and 430 university students were trained to visit houses and increase schooling rates.  Spot films were prepared, in which the President, the Prime Minister and Minister of National Education took part for the first time in Turkey, and these were featured 10,680 times in 45 different national channels as well as 16 local channels. The films were viewed 1,250,500 times through the MoNE website.  765 children were directed to education through 147 call centers, which received more than 1,000 calls.  20,000 booklets and 6,000 posters were prepared and distributed in 16 provinces. Another relevant effort is Vocational Skills Development Project (MESGEP) which is implemented in close collaboration with the private sector. MESGEP included activities aimed at improving employability of all disadvantaged groups including women. The needs of target groups were identified and they were offered employment-oriented courses in line with these needs. Under the project, 955 courses were organized in 35 pilot provinces between 2013 and 2017, and 15,382 people received and continue receiving training in these courses. Women are underrepresented in STEM fields of study Like in many other countries, a pattern of gender segregation by field of study can be observed at the tertiary level in Turkey. Women are disproportionately represented in certain fields including health, education, and humanities and arts; similar shares of women and men are enrolled in social sciences programs while engineering, manufacturing, construction, and services remain male-dominated (Figure 1.12). The share of women in services has increased substantially since 2005, while it has fallen slightly in agriculture. The socioeconomic background of women seems to contribute to these decisions, however more comprehensive studies on the various factors determining decisions on field of studies is still needed. Younger female students from towns or villages, with a lower-educated mother and who graduated from a vocational school in a rural region are more likely to study education, while females whose father is an unpaid family worker in a family business, who graduated from a secondary school in an urban region, attended kindergarten, and received private tutoring courses for longer are more likely to study engineering (Cevik 2015). 10 Ağrı, Adıyaman, Batman, Bayburt, Bingöl, Bitlis, Diyarbakır, Erzurum, Gaziantep, Kars, Muş, Mardin, Siirt, Şanlıurfa, Şırnak, and Van. 18 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 1.12: Enrollment in Tertiary by Fields of Study, Percentage Female 70 60 50 40 30 20 10 0 Engineering, Agriculture Education Health Humanities Science Services Social science, manufacturing and arts business and and law construction 2005 2010 2012 2013 Source: WDI, 2016 Gender-biased stereotypes and sexist attitudes prevail in the educational system in Turkey, although improvements have taken place after the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW). For instance, lower-secondary school textbooks show more men than women, and the professional affiliations featured differ between them, with more women than men taking part in family-related roles and generally being represented with objects of private life (Kukrer 2015). Women report that they particularly experience gender differentiation in the schoolroom during secondary education, mainly in the form of sexism related to educational progress and practice by teachers; for men, gender- differentiated treatment is more commonly reported in primary education, and in the form of physical punishment (Esen 2013). Within the latest reform, the curricula have been updated in Turkey, and gender-related training is being provided to teachers to address such discriminatory behaviors and stereotypes (see Box 1.3). Box 1.3: Promoting Gender Equality in Education The project ‘Promoting Gender Equality in Education (PGEE)’, launched in 2015, aims to:  Develop a Gender Equality Assurance Tool for Schools (GEATS) as a guideline and an assessment instrument for teaching and non-teaching staff at schools;  Improve MoNE´s capacity to promote equality and a gender-sensitive approach throughout the education system;  Raise awareness among students, parents, teaching and non-teaching staff on gender equality in education through a media campaign and local actions. 19 TURKEY GENDER COUNTRY ASSESSMENT 2018 The project was put into practice as a pilot scheme at 40 schools in 10 provinces;11 where all the curricula ranging from preschool to secondary education, textbooks, and training materials are being reviewed and 60,000 of instructors and managerial staff members have been trained accordingly. Moreover, it is targeted to reach out to 100,000 of people by means of the media activities and regional campaigns aimed at raising awareness.12 11 Izmir, Karaman, Samsun, Sivas, Malatya, Sanliurfa, Mardin, Batman, Erzurum, and Trabzon. 12 http://etcep.meb.gov.tr 20 TURKEY GENDER COUNTRY ASSESSMENT 2018 CHAPTER 2: ECONOMIC OPPORTUNITIES Economic opportunities for both men and women lead to dignified lives and to fulfill their individual and social potential. Equality in opportunities to access productive employment and income generation is not only about fairness: Gender gaps in participation in the labor market and entrepreneurship are common across countries and entail substantial economic losses for women and their families in the form of foregone income, but also losses in aggregate terms for the societies where women are deprived of those opportunities.13 This is especially relevant in the current demographic context of Turkey that represents a unique window of opportunity for growth and savings through higher participation of all those in the economically active age group. This chapter will examine differences in access to economic opportunity between men and women in Turkey around: (a) labor market participation and employment, (b) wages and earnings, (c) entrepreneurship and access to finance, and (d) poverty reduction. 2.1. Female labor market participation Despite recent progress, Turkey shows one of the highest gender gap in labor force participation among upper-middle-income countries from the ECA region and in the world (Figure 2.1). The gender gap in participation has declined over the last decade: In 2006 the rate of male labor force participation was 3 times higher than female labor force participation rate; in 2016, it was down to 2.2 times. However, still very few women participate in the labor market in Turkey. While for men there are only slight differences in participation between Turkey and the average among upper-middle-income countries, Turkish women participate at rates of about 32 percent compared to 62 percent on average in upper-middle-income countries and 63 percent in ECA upper-middle-income countries in 2014 (see Figure 2.1). It has indeed been estimated that around 26 million women remain out of the labor force in the country (World Bank 2015). Figure 2.1: Male and Female Labor Force Participation, Percent 100 Female labor force participation, 90 80.0 Turkey ECA ECA 70.0 80 Middle 60.0 70 income percent All middle 50.0 60 income 40.0 50 countries 30.0 40 20.0 30 TURKEY 10.0 20 0.0 20 40 60 80 100 2006 2016 2006 2016 Male labor force participation, percent Male labor force Female labor force participation participation Source: WDI, 2017. Note: Only upper-middle-income countries are used in figure on the left. 13 World Bank (2012); Cuberes and Teigner (2015). 21 TURKEY GENDER COUNTRY ASSESSMENT 2018 In the last decade, labor force participation rates have shown an upward trend— particularly participation has increased among younger women cohorts (Figure 2.2). Women born in 1963–1967 participated at rates near 20 percent by the time they reached ages 30–34; the rate almost doubles to about 38 percent for the same age group in the younger generation born between 1978 and 1982. This is partly due to the increase in educational achievement of the younger cohorts. As more women enter the educational system and stay in it for longer, their aspirations and drive to work increase, together with the opportunity costs of leaving the labor market (World Bank 2015). However, it must be noted that Turkey continues to be one of the few OECD countries where female labor force participation rates have decreased since the 1980s (Figure 2.3). The U-shape relationship between income per capita and female labor force participation, common in most countries, shows however that Turkey has not yet picked up consistently on a positive trend that ensures female attachment to the labor market and rates of participation comparable to countries with the same income levels. Figure 2.2: Women’s Participation in the Labor Market by Cohort 50 Female labor force participation 45 40 35 1979-1983 1969-1973 30 1984-1988 1974-1978 1964-1968 rate 25 1959-1963 20 1954-1958 15 1989-1993 1949-1953 10 5 0 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 Age Source: LFS 2008, 2013, and 2015 Figure 2.3: U-shaped Relationship Between GDP Per Capita per capita and Female Labor Force Participation 50 1980 Female labor force participation (in 45 1985 40 35 1989 1988 1991 1990 1992 percent, 15+) 30 1994 1995 1996 1999 2014 2013 1997 1998 2012 2011 20012002 2000 1993 2003 2010 25 2009 2004 2005 2006 2007 2008 20 15 10 5 0 0 2000 4000 6000 8000 10000 12000 GDP/capita (current US$) Source: WDI 2016 22 TURKEY GENDER COUNTRY ASSESSMENT 2018 Important regional differences can be observed in female labor force participation compared to men in Turkey (See Figure 2.4). On the one hand, labor force participation rates among women and men appear to match to some extent the regional differences and rural- urban socioeconomic divide in Turkey. However, women’s engagement in farming and agriculture explain a pattern in which at the same time many urban and more industrialized areas show low female labor force participation rates, and the most rural and poorer regions deploy considerably higher rates. More specifically, female labor force participation rates tend to be higher compared to those of men in some rural regions when only looking at the agricultural sector; whereas female labor force participation rates are particularly lower than those of men in some urban regions when focusing on the nonagricultural sector (Inan and Asik 2015). Female labor force participation rates have been largely driven by changes in the economic structure and the increase in rural-to-urban migration in Turkey over the last decades. As a result of such trends, one in every three Turkish women has become a migrant since the late 1980s. Many of these women used to work in agriculture and withdrew from their economic activity when they migrated to cities. Consequently, a considerable part of the low female labor force participation of women in Turkey is explained by women with low levels of schooling staying out of the market in urban areas. Policies that promote affordable care services in urban areas are likely to have an impact on mobilizing some of inactive labor force. Particularly among the semi-skilled women facing less cultural and economic barriers to employment, labor supply might be more easily activated through care provision. Consistently, female labor force participation rates in rural areas have been decreasing, in line with the overall decline in agricultural production, and due to changes in preferences in lifestyle choices (World Bank 2009). Figure 2.4: Regional Gaps in Female and Male Labor Force Participation (a) Female labor force participation rates 23 TURKEY GENDER COUNTRY ASSESSMENT 2018 (b) Gap between male and female labor force participation rates Source: LFS 2015 Similar to many other countries, the main reason reported by women for not joining the labor force is household and family-related duties. Although most women declare being inactive due to household chores, and disproportionately more so than men, there are differences within women depending, for example, on education level (Figure 3.5). The proportion of women reporting household responsibilities as the main reason to remain out of the labor market is nearly 80 percent among those with less than high school education and 14 percent among those with higher education degree. In contrast, for men the most common reasons for inactivity—especially among those with less than high-school education—are discouragement or participation in non-formal education. Existing evidence indicates that marital status and the number of children are important determinants of female labor force participation; women in Turkey tend to leave the labor force much earlier than men mostly in connection with family formation and childbearing (Okten 2014; Gökşen et al. 2014; World Bank 2015). Marriage is decisive in the early labor market exit by female workers, although its effect decreases with women’s education (Cebeci 2014b). Only 20 percent of women with three or more children are working in urban areas of Turkey (World Bank 2015). In connection with these trends, the evidence across many countries of a positive effect of institutional support and polices to help women better reconcile family and work lives is very robust (Goux and Maurin 2010; Del Boca 2015). The conflicting demand on women’s time for care and work activities represents a fundamental barrier to economic participation and generates a vicious circle of low labor market attachment and prominence of the care provider 24 TURKEY GENDER COUNTRY ASSESSMENT 2018 role that leads to increased vulnerability and gender-based inequalities. Maternity leave in Turkey is 180 days, of which generally only half can be taken after a child’s birth. Flexible work arrangements are constrained and only 14 percent of 0–5-year-olds are enrolled in early childhood education centers.14 Rates of enrollment to early education and care centers in countries with similar income levels are much higher in general. For example, in 2012 among 3- years old children only 5% were enrolled in Turkey compared to 39% and 45% in Mexico and Chile respectively.15 Regional disparities exist in childcare provision: the supply tends to be higher in provinces with higher female literacy rates and female labor force participation, and in less densely populated urban areas. The legal requirements for those activities are high, which substantially increases the costs of expanding supply.16 Figure 2.5: Reasons for Inactivity by Sex and Educational Level Other Family and personal reasons Disabled, old or ill Participation in non-formal education Household chores Seasonal worker Others not seeking a job but available to start Discouraged 0.00 0.20 0.40 0.60 0.80 Less than high school Women Less than high school Men General and vocational/technical high school Women General and vocational/technical high school Men Higher education Women Higher education Men Source: TURKSTAT, Detailed Labor Statistics 2015 Lately, there have been efforts in Turkey to enhance women’s employability and the reconciliation between personal and professional life. The legislations published in April 2016 and November 2016 states that by the time the official maternity leave ends, mothers that recently gave birth (or women who adopted a child younger than 3-yeasr-old) and fathers are entitled to work part-time for i) 2 months/60 days at their first pregnancy, ii) 4 months/120 days at their second pregnancy, iii) 6 months/180 days for the third or further pregnancy. The amount of time taken off can be considered as unpaid leave under this legislation. Also, the period of breastfeeding leave has been rearranged, and female workers are entitled to daily breaks for breastfeeding their babies (3 hours per day in the first six months and 1.5 hours up to one year of age). Initiatives for increasing the number of daycare centers and preschools have also been 14 World Bank (2015a) 15 OECD “Education at a Glance” 2014. 16 Aran et al (2016) and World Bank (2015a and 2015b). 25 TURKEY GENDER COUNTRY ASSESSMENT 2018 implemented, such as ‘My Mom’s Job is My Future’. Under the program, ‘Operation on Provision of Support to Women’s Employment’ more than 10,000 women were provided with vocational skills and personal development training across 43 provinces (see Box 2.1). Box 2.1. Programs Promoting Female Labor Market Inclusion in Turkey The project ‘My Mom’s Job is My Future’ was launched in 2013 by Borusan Holding in collaboration with the MFSP and the Ministry of Science, Industry and Technology, to increase women’s employment rate in industry and support the empowerment of women. The project will be carried out up to 2019. Within this program, daily childcare facilities called ‘Borusan Joy Factory’ are being opened to facilitate the needs of children ages 0–6 in 10 organized industrial zones (OIZ) of cities17 across Turkey. Borusan Joy Factories are intended to make life easier for women who cannot go to work because of their childcare obligations. The centers have started functioning in the organized industrial zones of Adıyaman, Afyonkarahisar, Balıkesir and Malatya.18 The purpose of the ‘Project on Promotion of Women’s Access to Economic Opportunities’, which was launched in 2013, is to contribute to the ongoing efforts by the government to improve women’s access to more and better jobs by (a) strengthening evidence -based policy making; (b) facilitating knowledge sharing (including Turkey’s experience with other countries) and raising awareness; and (c). Several research projects and technical notes relating to women’s employment and entrepreneurship, women’s cooperatives and sector-based analysis of women’s circumstances as well as surveys are conducted within the scope of the project. 19 The program, ‘Operation on Provision of Support to Women’s Employment’, was run in 43 provinces by Ministry of Labor and Social Security (MoLSS) between 2010 and 2013. Within the scope of the program, 131 projects received support and more than 10,000 women were provided with vocational skills and personal development trainings. Labor Force Market Analyses were conducted in 5 provinces and the employment opportunities which are accessible by women target groups were identified, which all resulted in drafting of an Active Model of Labor Force Policies. 20 Another recent project developed to keep women in the workforce is called the “Grandmother Project.” It is a pilot project that aims to make monthly payments to grandmothers who take care of their grandchildren while their mothers are at work. The project will initially cover 6,000 families in 10 provinces, and be monitored by the Ministry of Family and Social Policies. The ministry mentions that the goal of the project is to support employment of women, with grandmothers assuming the role of caretakers of their grandchildren receiving 425 liras ($114) a month.21 2.2. Employment and outcomes in the labor market Mainly as a consequence of the low labor force participation rate of women in Turkey, female employment rates are also comparatively low (26.7 percent compared to an average 47.7 percent in ECA in 2014), and much lower than those of men (64.8 percent in Turkey and 61.2 percent in ECA). It must, however, be noted that the country has managed to generate employment during the crisis years and that women have particularly benefited from it—female employment rates have increased by 5 percentage points since 2008. This may be partly related to the provision of public subsidies to cover the social security costs for the first year of employment of newly hired women from 2008 to 2015 (World Bank 2015a). 17 Adıyaman, Afyonkarahisar, Balıkesir, Malatya and 6 other provinces. 18 http://www.anneminisi.org/en/Home.aspx 19 Research projects and technical notes as outputs of the project as well as information on other activities can be found at: http://kadininstatusu.aile.gov.tr/projeler/yurutulmekte-olan-projeler/kadinlarin- ekonomik-firsatlara-erisiminin-artirilmasi-projesi 20 http://www.iskur.gov.tr/ 21 https://buyukanne.csgb.gov.tr/ 26 TURKEY GENDER COUNTRY ASSESSMENT 2018 Despite this progress, unemployment rates are higher for women than for men in 2014 (11.8 compared to 9 percent). Although this difference is not wide, it becomes quite significant when examining the educational level of unemployed women in comparison to men: 32 percent of unemployed women had completed tertiary education relative to only 14.6 percent of unemployed men. Gender differences with regard to employment are also evident when examining the types of work that women tend to do compared to men in Turkey. While 71 percent of all employed Turkish men are regular employees and 22 percent of them are self-employed, this is only the case for 64 and 9 percent of women, respectively. This is related to a sharp contrast in participation as unpaid family workers: 31 percent of employed women are in this category compared to only 4 percent of men (Figure 2.6). Figure 2.6: Share of Regular Employees, Employers, Self-employed, and Unpaid Family Workers Across Gender (Percent), 2016 80 70 60 50 40 30 20 10 0 regular employee employer self-employed unpaid family worker men women Source: LFS 2016. A higher proportion of employed women compared to men are engaged in agriculture, although the gender gap has decreased since 2000; in turn, a substantially higher share of employed men work in industry compared to women, and the gap today is even larger than 15 years ago. In services, although a small gender gap exists, it has been narrowed over time and today almost the same share of both working women and men are employed in that sector. 27 TURKEY GENDER COUNTRY ASSESSMENT 2018 Women are more present in traditionally female-dominated occupations: 95 percent of those engaged in household work (as employers of domestic staff) are women, 65 percent in the case of human health and social work and 47 percent in agriculture (Table 2.1). Conversely, the distribution of all employed women across sectors, shows that 28 percent work in agriculture and forestry, 15 percent of them in manufacture, 11 percent in wholesale and retail, 10 percent in education, and 9 percent in human, health and social work. Table Error! No text of specified style in document.2.1: Women’s Representation in Employment and Earnings by Sector, 2016 Ratio of Percent of Percent of Ratio of Men Women Women in Women Wage to Wage to Sector Employers Mean Wage Mean Wage Total 30.64 8.79 97.6 100.9 Agriculture, forestry 46.41 7.33 64.9 119.1 Mining and quarrying 3.49 — 87.1 100.5 Manufacturing 24.6 5.27 86.4 103.7 Electricity, gas, steam, and air 11.12 — 109.9 99.0 conditioning Water supply, sewerage 10.33 — 91.5 101.1 Construction 4.22 1.9 130.0 98.5 Wholesale and retail 23.93 8.57 93.4 102.1 Transportation and storage 9.45 5.27 117.3 97.8 Accommodation and food 23.52 6.43 95.7 101.2 Information and communication 25.22 15.69 89.0 103.9 Financial and insurance activities 44.06 25.21 90.8 107.4 Real estate activities 17.91 7.24 83.6 103.3 Professional, scientific, and technical 40.55 16.86 91.4 107.2 Administrative and support 33.6 7.37 94.4 102.0 Public administration 17.28 36.37 95.8 100.8 Education 52.86 42.24 92.5 107.7 Human health and social work 70.63 33.28 85.8 132.7 Arts, entertainment activities 24.61 17.59 127.8 91.7 Other service activities 20.91 23.15 85.7 103.8 Activities of households as employers 90 — 90.0 173.7 Extraterritorial organizations 37.91 — 147.4 71.7 Source: LFS, 2016. Employment gap ratios in favor of men are especially high among older people and those with upper-secondary education. While among individuals younger than 44 years the share of employed men is 2.3 times higher than that of women on average, this rate increases to 2.5 among people above 45 years. Gender differences are statistically significant across all age groups (Figure 2.7a). Around 3 times more men than women with an upper-secondary degree are employed, although this trend can also be observed to a lesser extent among those with lower-secondary, primary, or higher education.22 The gender gap only decreases for men and women with graduate education: the share of men with a master’s or doctorate degree who 22 Higher education refers to two, three or four-year programs after secondary school. 28 TURKEY GENDER COUNTRY ASSESSMENT 2018 are employed is 9 percent higher than that of women with the same level of education (Figure 2.7b). Indeed, different sources highlight the relevant role that education plays for women´s labor market inclusion in Turkey (Gökşen et al. 2014). As highlighted before, low participation rates among poorly educated women in urban areas accounted for a large share of the low female labor force participation rates observed in Turkey in comparative terms (World Bank 2009). Turkish women who participate in the labor market show higher attainment levels than men (25.1 percent of the female labor force has tertiary education compared to 17.5 percent of the male labor force), and thus the average female employee is more educated than the average male (Cebeci 2014b). Bozkaya (2013), Dublen (2014), Er (2013) and Kilic and Ozturk (2014), all find that education, especially higher education, and particularly in urban areas, is the most important factor affecting female labor force participation and employment decisions in Turkey. Differences in employment across age groups confirm that the positive progress experienced with regard to female labor market inclusion over the last years has been mostly driven by a generational shift associated with changes in education (World Bank 2014, 2015; A2F Consulting 2014). The weak labor market attachment of Turkish women may also be partly explained by skills mismatches and the availability of lower quality or not suitable jobs (Gökşen et al. 2014). Some sources highlight a growing skills mismatch in Turkey, which seems to be especially affecting women. The share of working-age women that report withdrawing from the labor market due to the lack of adequate education and training has been increasing. Indeed, a large number of tertiary graduates are unable to find jobs, and this is especially the case for women. On the other hand, low-skilled women more often than men find themselves working in low- productivity sectors of the economy (World Bank 2015). The jobs that many women find are not accommodating to their needs and offer poor conditions (Gökşen et al. 2014). Informal work rates and the ratio of vulnerable employment are higher for women (19.8 percent and 38.5 percent) than for men (16.4 percent and 25.6 percent). 29 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 2.7: Employment Rates Across Gender, Age, and Education in 2016 (a) Employment ratios across age groups and gender, (b) Employment ratios across education percent categories and gender, percent 100 100 men women men women 90 90 80 80 70 70 60 60 percent percent 50 50 40 40 30 30 20 20 10 10 0 0 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 age groups Source: LFS 2016, World Bank staff calculations. Note: Population ages 20–64. Education refers to highest level completed, 2.3. Gender gap in earnings As in many other countries, women in Turkey earn on average less than men with the same level of education, working in the same sectors or occupations, or within the same company (Cebeci 2014a, b). The distribution of female wages is shifted toward the left (Figure 2.8), which indicates that a higher proportion of women receive lower salaries than men. The density distribution at the top of the wage distribution needs to be interpreted carefully given that there are very few women participating in the labor market and receiving wages above the median level. The economic activities where women concentrate (with the exception of education) are characterized by important gender gaps in wages. In female-dominated sectors, wage gaps are to the advantage of men. In all the areas where women represent either the majority of all employees, or at least more than 40 percent of all employees, women wages are, on average, below average wages for men engaged in the very same activities. However, in male-dominated activities, it is women who generally earn more as they mostly work in white-collar positions compared to men—who are largely engaged in blue-collar jobs (Mercan and Karakas 2015). This is mainly explained by the fact that the few women participating in these sectors are those with outstanding education and sociodemographic characteristics. Earnings differentials are also observed among employers: the average female employer earns around 61 percent of average male employer’s earnings (Cebeci 2014a). 30 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 2.8: Log Monthly Wage Across Gender, 2016 2 1.5 Density 1 .5 0 2 4 6 8 10 Log monthly pc income Men Women Source: LFS, 2016. Note: Kernel density distributions. The gender difference in the distributions is statistically significant at the 5 percent level based on Kolmogorov-Smirnov equality-of-distributions test. Returns to education and experience are high in Turkey, with education paying more for women.23 Once again, the scarce representation of women in the labor market implies that education and sociodemographic characteristics of women who participate in paid employment are very different from the average women in the same age group.24 Consistent with previous results, data from LFS 2016 indicate that the majority of the observed gap in wages arises from ‘unobserved’ factors that are potentially attributable to labor market discrimination; as opposed to those that arise from ‘observable’ characteristics such as education, experience, sector of employment, and region. The decomposition of the gender wage gap actually shows that if the wage gap is not even larger than the one already observed in Turkey, it is because the limited number of employed women are actually more educated and participate in better-paying (male-dominated) sectors (Table 2.2) The model presented actually is constrained by a limited number of observable variables, the inclusion of omitted variables such as ability, quality of education, education of the parents, and so on, would have likely reduced the size of the unexplained magnitudes (and more accurate described discrimination). 23 These results are consistent with previous literature. See for example, Dayioğlu and Kasnakoğlu (1997); Gürler and Üçdoğruk (2007); Günalp, Cilasun, and Acar (2015) 24 Results from OLS regressions are presented in Table A2 in Appendix. 31 TURKEY GENDER COUNTRY ASSESSMENT 2018 Table 2.2: Oaxaca Decomposition of Monthly Wages (with Heckman correction for selection), 2015 (a) Threefold and twofold (b) Detailed twofold decomposition of decomposition of wage gap, original wage gap, log of monthly wage scale Threefold decomposition Gap 100 Gender gap, Explained percent 11.68*** -33.77 Endowments -5.67*** Experience 13.51 Coefficients 13.90*** Education -76.95 Interaction 3.46*** Region -6.48 Sector -0.47 Twofold decomposition Mills ratio 36.92 Explained -3.94*** Unexplained 133.77 Unexplained 15.63*** Experience -198.20 Education -2.69 Region -0.90 Sector -3.61 Mills ratio -70.19 Constant 409.37 Source: LFS 2016, World Bank staff estimation. Note: *** Significant at 1 percent, ** significant at 5 percent, * significant at 10 percent. Explanatory variables include age, age-squared, years of education, economic sector and regional dummies. Variables used for Heckman correction for selection include age, years of education, household size, head of household and region. Positive sign of components indicates increase of wage gap. 2.4. Women’s entrepreneurship and access to finance Based on recent studies, the level of entrepreneurship is low in Turkey among both men and women. While the share of employers in the total employed (6 percent) in Turkey is high compared to the ECA average (3.2 percent), the gender gap is particularly large: 7 percent of men are employers compared to only 1 percent of women). Most self-employed women are own-account workers (OAW), and although the share of the overall population in OAW has decreased in the last decade, among women it has increased over the same period. As with employment, women employers tend to have higher educational attainment, to be much younger, and to earn less than male employers. Significant gender segregation by sector exists, as seen in section 3.1. Most female employers operate in one of three sectors: 29 percent in retail and wholesale trade, 28 percent in manufacturing, and 25 percent in other services sectors. In addition, firms owned by women have a smaller size than those owned by men (World Bank 2014, Cebeci 2014a; Okten 2014). Turkey is also one of the countries where women are less represented among firms’ managers and owners. Based on BEEPS 2013 data, women appear to have fewer chances than men to participate in firm’s ownership, unless is very small ownership (less than 50 percent) and are even less likely to manage them. The share of firms with some female ownership or management in Turkey is only 33 percent, compared to the average for the ECA region of 36 percent. However, only 5 percent of Turkish firms have women with 50 percent ownership or a 32 TURKEY GENDER COUNTRY ASSESSMENT 2018 female manager compared to an ECA average of 27 percent. Similarly, only 5 percent of firms in Turkey have a woman among top management in contrast to 20 percent on average for ECA (Figure 2.9a). Although in the ECA region female ownership is almost equally likely regardless of firm’s size, in Turkey medium-size firms are more likely to have female participation in ownership (40 percent). The share of women in female management is low across all sizes, but especially more so in larger firms (Figure 2.9b). Figure 2.9: Female Management and Ownership of Firms in 2013, percent (a) Female ownership and management in Turkey (b) Female ownership and management by firm and ECA size (employees) 40 45 percent of firms 35 40 30 25 35 20 30 percent 15 25 10 5 20 0 15 Female Female Female 10 manager or manager or manager 5 some female 50% and 0 ownership more female Small 5-19 Medium 20- Large 100+ ownership 99 Turkey ECA Female ownership Female management Source: BEEPS (European Bank for Reconstruction and Development [EBRD] and World Bank 2013). BEEPs 2013 data also confirm the persistence of gender segregation in female compared to male firm’s management and ownership. Women tend to concentrate in sectors including textile, garments, machinery and equipment, nonmetallic minerals, food, wholesale, and chemicals. In a lower proportion, women also show higher ownership rates in sectors such as hotels and restaurants, retail, services, and manufacturing (Figure 2.10). High female ownership and management is observed in traditionally feminized sectors as chemicals, plastic and rubber, textiles, machinery, and equipment. High female ownership and low management is the trend in the majority of the remaining sectors. Construction and other services are the sectors where low female ownership and management rates concur. The low rates of entrepreneurship among women appear to be related to barriers to enter and remain in those activities, and not to higher failure rates. Such barriers are associated with financing, skills, and social norms related to marriage and childbearing. Indeed, the number of children appears to decrease the probability of becoming an employer or OAW for women (World Bank 2014; Cebeci 2014a; Kizilaslan and Karaomer 2015). As is the case for other outcomes, education decreases the gender gap in entrepreneurship in Turkey since it has a much larger positive differential impact for females than males (Okten 2014). 33 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 2.10: Female Ownership and Management by Economic Sectors in 2013, Percent Hotel and restaurants Garments Retail Textiles Machinery and equipme Other services Food Non metallic mineral Wholesale Construction Other manufacturing Plastics & rubber Fabricated metal prod Chemicals 0 10 20 30 40 50 60 70 80 Female management Female ownership Source: BEEPS (EBRD and World Bank 2013). World Bank staff calculations. Although financial inclusion, measured as access to financial accounts and loans, is similar in Turkey than the average for ECA countries, the gap in financial inclusion between men and women in Turkey remains comparatively large, even after the positive progress registered since 2011. In 2014, 70 percent of men had formal accounts compared to 44 percent for women; this difference was even larger in 2011 (82 percent of men compared to 33 percent of women). People borrow above average shares in ECA, and mainly from stores buying on credit, and more so among men: 12 percent of men compared to 10 percent of women have a loan in Turkey. Yet access to credit seems to be crucial for economic participation, and especially for entrepreneurial activities (World Bank 2014). The difference between ECA and Turkey in one or more monthly deposits and withdrawals is lower in the case of men (14 and 7 percent, respectively) than for women (25 and 10 percent, respectively) (see Figure 3.11). These results are aligned with those of previous studies (Klapper and Singh 2014; World Bank 2014). 34 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 2.11: Use of Bank Accounts in Turkey and the ECA Region, 2014 (a) Turkey (b) ECA 120 120 100 100 80 80 percent percent 60 60 40 40 20 20 0 0 men women men women male female male female Deposits Withdrawals Deposits Withdrawals 0 1-2 3+ No answer 0 1-2 3+ No answer Source: FINDEX (Demirguc-Kunt and Klapper 2012). Workplace participation is an important factor in explaining differences in financial inclusion between men and women (Klapper and Singh 2014). Many more men than women used an account to receive wages, and this gap is much larger than the ECA average. Most men who do not have an account report they do not need financial services, while for women the main reported reason is the lack of money (Figure 2.12b). Savings are low among both women and men, although the percentage of these that have saved in a financial institution has doubled between 2011 and 2014 (from 6 to 13 percent among men and from 2 to 5 percent among women). Figure 2.12: Purposes of Accounts and Sources of Loans in Turkey, 2014 (a) Source of loans, percent (b) Reasons for not having an account , percent account has a loan no need for financial… cannot get one family and friends Family member… Lack of money store credit women Lack of trust men Lack of documentation financial institution Too expensive Too far away private lender 0 50 100 0 10 20 30 40 women men percent Source: FINDEX 2014 2.5. Welfare and Income Although most households are headed by men in Turkey across age groups, women are more likely to head single-person and single-parent households, especially for the 40 and over 35 TURKEY GENDER COUNTRY ASSESSMENT 2018 age groups. Lower wages by the main earner of the household can have important impacts on the well-being of the entire household. Among young people (24 years or younger), 40 percent of households are headed by a woman compared to 60 percent that are headed by a man. This gap increases for the middle- aged cohorts—on average 90 percent of households are headed my males for those aged 35– 55 years.25 For the oldest age group (65+) the number of households headed by women increases, probably as a result of higher female life expectancy. Despite the low prevalence as household head, about 66 percent of all single-person and single-parent households are headed by women for all age groups (19+). The percentage of female-headed households increases with age. While in the 19–39 years cohorts an average 42 percent of single-parent households are headed by women, for the population group above 40 years, around 73 percent of households are headed by a woman (Figures 2.13 and 2.14). Households headed by a pensioner male as well as single-parent household headed by women have the lowest income per capita among all types of households. Only the households headed by pensioner women have higher income per capita than those headed by men (Figure 2.15). Moreover, women heads of single-person households are poorer than men across all age groups. The gap is larger for middle-aged head of households (40–56 years old). In that group, men’s income is on average 28 percent higher than that of women. The income gap among single-person households decreases for older people. Among 65+ year olds, the gap in the income levels between male- and female-headed single person households decreases but the advantage to men still holds (Figure 2.16). Figure 2.13: Share of Male- and Female-Headed Figure 2.14: Share of Male- and Female- Households in Population by Age Groups, Percent Headed Households Among Single-Person and Single-Parent Families by Age Groups, Percent 100 100% percent head of households percent of men and women 80% single or one in parent… 80 60% 60 40% 40 20% 20 0% 0 19-24 35-39 40-55 56-64 65+ women is head of household men women man is head of household Source: SILC 2015. World Bank staff calculations. 25 Estimations based on the Survey of Income and Living Conditions (SILC) 2015. 36 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 2.15: Monthly Income Per Capita by Gender Figure 2.16: Monthly Income Per Capita by of Head of Household and the Type of Household Gender of Head of Household and Age Group Among Single-Person and Single-Parent Households 2500 2500 Male head Female head men women 2000 2000 Turkish Lira 1500 Turkish Lira 1500 1000 1000 500 500 0 Single person Single parent Head of household household household is 0 pensioner 19-24 35-39 40-55 56-64 65+ Source: SILC 2015. World Bank staff calculations. Note: Income per capita refers to the household disposable income divided by the equivalised household size (TURKSTAT, 2015); includes average monthly cash and in-kind income. *** Gender difference significant at 1 percent. Figure 2.17: Monthly Income Per Capita in Female-Headed Single-Parent Households by the Number of Children Below 14 years, 2015 1800 1600 1400 1200 Turkish Lira 1000 800 600 400 200 0 no children one child two children three children+ number of children 0-14 years Source: SILC 2015, World Bank staff calculations. Note: Income per capita refers to the household disposable income divided by the equivalized household size (TURKSTAT, 2015); includes average monthly cash and in-kind income. The risk of poverty of single-parent households increases with the number of children. Analysis of single-parent households headed by women shows that their income per capita drops sharply with children (Figure 2.17). Single-parent female-headed household without children below 14 years have an average income per capita of around TL 1,590 per month, while a household with 37 TURKEY GENDER COUNTRY ASSESSMENT 2018 3 children or more has less than a quarter of such income per capita (687 Turkish Lira).26 Comparing to the mean per capita income during 2015 in the country, which amounted TL 1,376 per month, and the median of TL 1,041 (TURKSTAT, 2015), single-parent female-headed households with children do better than the national average 26 Economies of scale may entail that less money is needed per each additional household member to meet consumption needs. Results in Figures 2.15-2.17 were checked by calculating equivalised income per capita using the following equivalence scale: 1 to the household respondent, 0.5 to each additional person 14 and older, 0.3 to children below 14 years (TURKSTAT, 2015)). The results do not change and the risk of poverty still increases with the number of children. 38 TURKEY GENDER COUNTRY ASSESSMENT 2018 CHAPTER 3: AGENCY Women’s agency refers to their capacity to make choices free of constraints due to gender differences and to transform those choices into actions, so they can take advantage of the opportunities societies offer. Women’s constraints to agency, for instance in the form of social norms or institutional biases, often underlie discriminatory practices that prevent their equal access to endowments and economic opportunities compared to men in all areas of life, and from the womb up to old age.27 Agency is thus the key basis for women to stand on an equal footing with men in society, as it will largely affect decisions on human capital investments and, ultimately, women’s chances to become active social and economic agents. This chapter explores the existing differences in agency between Turkish men and women, around three areas: social and gender norms; an in-depth look at the phenomena of violence against women; and gender gaps in voice and representation. Finally, the chapter also reviews the legal, institutional, and policy framework for gender equality in place in Turkey, as a key element of the enabling environment for agency exercise by Turkish women. 3.1. Social and gender norms Societal views of the role of women in Turkey have undergone slow changes over the last decades, along with their progressive inclusion in education and labor markets. The country continues to combine egalitarian views in some areas with unequal views in others. Patriarchal and conservative values are still prevalent in the Turkish society, particularly in the east of the country, and appear to be quite resilient to social transformations in other dimensions. Different sources report, for instance, that traditional practices such as arranged or early marriages continue to be common and widely accepted, especially in rural areas. One-third of the highly educated and over one-half of the wealthiest Turkish population believes that women should marry young—before the age of 25 years; and according to World Values Survey (WVS) 2011, the majority of men and women do not find divorce justifiable (50.2 and 48.9 percent of men and women, respectively report finding it never justifiable). Box 3.1. Early and Arranged Marriages in Turkey In Turkey, arranged marriages are still a frequent practice, particularly among some regions in the country, and although family-arranged marriages have decreased, they remain significant. While by the 2008 TDHS, 51 percent of ever-married women declared that their marriage was arranged by their family, only 7 percent is reported to be without the woman’s consent; TDHS 2013 shows that the proportion of ever-married women declaring that their marriage was arranged by their family lowered to 46.5 percent (the level of nonconsensual marriage remains at 7 percent). Arranged marriage is over half of the marriages of women of 35 years and older, but still 40 percent for women ages 18–34, and 30 percent for men in same age group. Several of these arranged marriages are of women that got married before the age of 18. According to TDHS, 15 percent of women in the 20 –29 age group married at age 18, a rate that increases to 20–26 percent for women 30–49 years old. Less than 3 percent of women married by age 15. Child and adolescent marriages come with important age differences between spouses. In 2013, 76 percent of 15–19-year-old married girls had husbands who were at least 5 years older, and 22 percent 27 World Bank (2012). 39 TURKEY GENDER COUNTRY ASSESSMENT 2018 were more than 10 years older than their wives. Data from the 2011 TAYA shows that marriage before 18 years of age is more prevalent in the southeast and central Anatolia regions (18 and 16 percent of women ages 25–64), while the lowest levels are in the west and east Marmara regions (8 and 10 percent, respectively). Child marriage among men is rare. Between 2006 and 2011, only 0.2 percent of men were married before the age of 18. Until the reform in 2001 that increased legal age of marriage for women to 17 years of age, the Civil Code law established the legal age of marriage for women at 15 years. Reduction of early marriage responds to changes such as the increase in women’s educational attainment (according to Kirdar et al. (2014), the 1997 reform on compulsory education had a higher impact on early marriages than the Civil Code law of 2001). Indeed, early marriages are correlated with lesser educational attainment, with women married before age 18 being more likely to have lower levels of completed education.28 Views about women’s economic opportunities remain mixed, and restrictive to a certain extent not only among Turkish men, but also, although to a lesser extent, among Turkish women (see Figure 3.1). Men and women alike have started agreeing with the importance of a university education for boys and girls, however, for many of them, women’s primary role should be in relation to motherhood and homemaking. As many as one in every two women agree with the fact that “a woman with very young children must not work.”29 Men and women alike still believe that “men make better political leaders and business executives than women,” and that “men should be given priority over women in times of job scarcity” (Figure 3.2). However, the share of women who agree with those statements is significantly lower than that of men, which may be interpreted as a sign of some progress regarding social views on Turkish women. With regard to the intra-family distribution of responsibilities, a significantly larger proportion of women do not agree with the belief that “women earning more than their husbands is almost certain to cause problems” (57 percent compared to 51 percent), while the opposite is true in connection with “having a job is the best way for a woman to be an independent person”: 65 percent of women compared to 61 percent of men agree with this statement. Economic empowerment brings a transformation to many norms, for example, only 9 percent of employed married women declared that their husbands were the only persons to decide how their earnings will be used, compared with a higher share among nonworking women.30 28 Own analysis based on TAYA 2011 29 According to TDHS 2013 responses. 30 According to TAYA 2011 results. 40 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 3.1: Gender-related Social Norms - Mean Value of Indicators Showing Agreement with the Views on Gender-Related Statements Across Gender, 2008 **A university education is more important for a Children boy than for a girl Child suffers with working mother ***Men better political leaders ***Men better business executives Labor Market **Woman earning more than husband is problematic ***When jobs scarce, giving men priority ***Being housewife as fullfilling as paid job ***Job best way for independence women Marri age Divorce acceptable 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 women men fully agree fully disagree Source: WVS (2011). World Bank staff calculations. Note: *** Gender difference significant at 1 percent, ** gender difference significant at 5 percent, * gender difference significant at 10 percent. Respondents have to (1) strongly agree, (2) agree, (3) disagree, or (4) strongly disagree with each of the statements above. T-test is conducted for the mean value of indicators. This progressive transformation is more evident when examining responses by age groups, as younger generations’ views conform to gender-equality principles to a larger extent (see Figure 3.2). For instance, as women get older, they show increasing agreement on the statements that “children suffer with a working mother” or that “being a housewife is as fulfilling as a paid job.” Agreement with the facts that “when jobs are scarce men should have priority” and that “women earning more than their husbands is problematic” consistently increases with age for both men and women. Further confirming the higher incidence of tradition among older groups, agreement with the statement that “a job is the best way for women’s independence” declines with age (Figure 3.2). Additional evidence confirms the relevance of age and education in reported views on the role of women. A study conducted in Ankara, for instance, concluded that although women do not tend to assess family life in terms of social gender roles, married women’s perceptions of basic social and economic activities depended on their age, educational attainment, employment status, and family’s incomes (Gulay and Bener 2011). A similar analysis using TDHS 2013 data shows that both younger and older women tend to agree more with the fact that men should make important decisions, while middle-aged female respondents were more likely to disagree. This trend was also observed with regard to the preference for an educated son over an educated daughter. However, age did not affect the response on the statement that women with young children should not work. 41 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 3.2: Views on Gender-Related Statements Across Men and Women and Age Groups, Percent of Agreement, 2008 Children suffer when mother works Being housewife is as fulfilling as paid job 76 76 74 75 75 73 82 67 64 68 64 73 76 73 75 78 78 62 63 61 60 62 62 61 71 70 68 67 69 68 56 58 58 65 63 65 63 63 62 64 63 47 50 56 Women Men Women Men Job is the best way for women's Women earning more than husband is independence problematic 71 71 69 68 73 76 69 65 68 66 63 59 63 51 55 52 53 53 48 49 46 50 54 55 54 55 53 58 58 58 57 55 55 54 44 49 39 38 41 41 39 33 26 Women Men Women Men When jobs are scarce: giving men priority 74 71 73 67 65 69 62 61 58 57 61 65 65 54 54 50 53 44 48 43 31 Women Men Source: WVS (2011). Despite the trend toward greater gender equality in social norms among younger females, greater effort needs to be placed to advance an agenda where norms allow for women’s access to opportunities. Data from 2011 shows that the share of men reporting that it is not appropriate for women to work in the Family Structure Surveys increased from 23 percent 42 TURKEY GENDER COUNTRY ASSESSMENT 2018 in 2006 to over 26 percent in 2011. Among all those reporting that working is not appropriate for women in the last round, 56 percent believed that “a woman’s primary duties are household management and childbearing.” The share of women that reported believing that men should have priority in times of job scarcity also increased to over 55 percent in the 2011 round of the WVS, after having fallen from about 52 percent to 46.7 percent between 2001 and 2007, probably in response to the economic crisis facing several countries and Turkey, but efforts toward stopping further reversals are necessary.31 The government has introduced specific programs that offer special emphasis on the equal distribution of household and family care duties between men and women, including some recent reforms to expand the access to early childhood education and care, and extend maternity leave for women, signaling the value of their economic participation. Among several new trainings being set in place for young couples, the Family Training Program (AEP) led by the MoFSP and the Pre-Marriage Training programs32, which are carried out in all 81 provinces include several modules on co-responsibility. Additionally, the Family Training Program for parents with 0–18-year-old children managed by the MoNE includes a support component to inform fathers about the needs of a child, promotes the adoption of a democratic and equal approach in the distribution of the household responsibilities, and aims to develop and enhance intra-familial and marital communication. 3.2. Violence against women33 Turkish women, as many other women across the world, regardless of their country’s level of income, face the threat of violence against them. The prevalence of gender-based violence in Turkey is not easy to estimate and compare with other countries—TDHS does not cover prevalence question (ever experienced) but only justifications of violence, and the Turkey Domestic Violence Survey (TDVS) is a stand-alone country survey. However, all these data sources provide rich data to look into the problem. Prevalence of violence among women With regard to physical and/or sexual violence incidence, in 2008, 13.7 percent of ever- married Turkish women reported having experienced physical and/or sexual violence by their husband or intimate partner in the 12 months before to the survey. This number went down to 11 percent by 2014. The prevalence of lifetime physical and/or sexual domestic violence has also gone down from 41.9 percent in 2008 to 37.5 percent in 2014 (see Figure 3.3). Such declines are important, although relatively small. The prevalence rate observed for European countries in 31 In relation to this, government public discourse has changed in the past 3 to 4 years to a more restrictive view of women’s role, operating in dissonance with several policies being set in place. Concerns have been raised about the potential institutionalization of conservative views on families and the role of women as mothers to the detriment of gender equality through legislative and policy reforms (Okten 2014, CEDAW 2016) 32 Dasre, A., A. Greulich, and C. Inan (forthcoming). 33 This section is based on the findings by Dasre et al. (forthcoming). The paper was commissioned especially for this gender assessment to look at the socioeconomic determinants of domestic violence in Turkey. The analysis combines different data sources (TDHS waves 2003, 2008 and 2013 and the Domestic Violence Survey waves 2008 and 2014). 43 TURKEY GENDER COUNTRY ASSESSMENT 2018 the survey conducted by the EU Agency for Fundamental Rights (FRA) in 2012 shows a lifetime prevalence of 33 percent as an average of the EU-28 countries included, and an 8 percent average of women reporting having experienced violence in the 12 months before the survey.34 Figure 3.2: Physical and/or Sexual Violence by a Partner in Turkey a. Turkey. Physical and/or Sexual Violence by a Partner b. Turkey and EU-28, Physical and/or Sexual (Share of Ever-Married Women 15–59) Violence by a Partner Since Age 15 (Share of Ever-Married Women 15–59) 39 year 2008 year 2014 33 41.9 39 13.7 11 since the age of 15 in the 12 months prior to the interview EU 28 Turkey Source: Survey on Domestic Violence Against Women in Turkey, 2008 and 2014; FRA gender-based violence against women survey dataset, 2012. Among the specific characteristics of women who have experienced violence by their partner, based on TDVS, while there are more women in the age group 30 –39 (36 percent of those having experienced violence in the last 12 months), the highest incidence is among the youngest group at 20 percent for the ever-married women ages 15–19, compared with 12 percent for those in the 30–39 age group. In line with this, the younger the woman’s age at marriage, the higher the prevalence of domestic violence (14 percent for those marrying at age 17 and younger, 12 percent for those of ages 18–21, and 9 percent for those in the age group 22 and above). Women whose marriage is arranged are also at a higher risk of experiencing domestic violence than women who decided themselves who to marry; two-thirds of women who have experienced violence in the past 12 months were part of an arranged marriage. The prevalence of physical and/or sexual violence is lowest among women with tertiary education, and also when the woman’s partner has that same level of education, regardless of the woman’s education. Economic empowerment does not always protect women from violence. Table 3.1 summarizes key descriptive information in this regard. Women who are in formal employment are less likely to experience violence than those in informal/irregular employment, and as likely as nonworking women to experience violence. Ownership of the home, although very low, is linked to a much lower level of violence incidence, but women who provide the main income source for the household are at a higher likelihood of violence than other levels of income contribution. 34 The highest reported level, of 11 percent, same as Turkey, in the past 12 months was reported in Belgium, Denmark, France, the Netherlands, and Sweden. 44 TURKEY GENDER COUNTRY ASSESSMENT 2018 Table 3.1: Distribution and Proportion of Women Who Have Experienced Physical and/or Sexual Violence by Their Husband in the 12 Months Before the Survey (Married Women 15– 59) Distribution (%) Proportion (%) Activity Status Formal employment 14 10 Informal/irregular employment 17 14 Not working 69 11 House ownership No 85 13 By herself 7 7 With other people 8 9 Contribution to family income Sole income of household 4 16 More than others 2 10 About the same than others 13 11 Less than others 3 6 Nothing 78 12 Source: TDVS 2014 Although differences between rural and urban regions seem negligible, when comparing violence levels among the 12 regions in the country, the prevalence appears to be somewhat higher in the southeast and somewhat lower in the northwest. Figure 10, which distinguishes between provinces, illustrates however, that the prevalence of domestic violence in Turkey is quite heterogeneously spread over the country. Figure 3.3 Prevalence of Physical and/or Sexual Violence by a Partner in the 12 Months Before the Interview in Turkey, Ever-Married Women Ages 15–59, by Provinces Source: TDVS 2014 45 TURKEY GENDER COUNTRY ASSESSMENT 2018 Education and economic empowerment as preventive factors As initial data shows, a more in-depth analysis of the determinants of violence incidence confirms that education is one of the critical protective factors against violence.35 Figure 3.4 shows the estimated probabilities of experiencing domestic violence according to education and illustrates that at 7.5 percent, the probability is significantly lower (and significant) for women with tertiary education in comparison to women with secondary and primary education or less. When it comes to women’s contribution to the household income, contributing in equal levels as other household members lowers the probability of experiencing violence, while being the sole contributor or not contributing at all—both extremes between the least and most frequent status of women in relation to the household finances in Turkey—is associated with an increase in the probability of being exposed to violence. Educated women with a sufficient degree of economic independence, with educated partners, are the most protected from violence. Figure 3.4: Estimated Probabilities of Experiencing Physical and/or Sexual Violence by a Partner in the 12 Months Before the Survey (a) Estimated probabilities by woman’s education (b) Estimated probabilities by women’s contribution to family income 0.140 0.300 0.250 0.109 0.114 0.250 0.200 0.120 0.180 0.200 0.165 0.100 0.125 0.150 0.075 0.080 0.100 0.060 0.050 0.000 0.040 Only she More than About the Less than Nothing 0.020 has income others same than others in the others 0.000 household low education middle high education education Source: TDVS 2014 Note: Reference group: women aged 20–29, age difference with Note: Reference group: women ages 20–29, age partner 3 years, age at marriage 22–29, marriage not arranged, 2 children, first marriage, partner's mother tongue is Turkish, nuclear difference with partner 3 years, age at marriage 22–29, family, no alcohol consumption, urban, middle educated, no real marriage not arranged, 2 children, first marriage, and estate property, informal/irregular employment. partner’s mother tongue is Turkish, nuclear family, no alcohol consumption, urban. Definition of education categories: Low education (pre- primary and primary completed), middle education (secondary), high education (tertiary). 35 To estimate women’s propensity of experiencing physical and/or sexual domestic violence (last 12 months), the authors apply a multivariate logit model with robust standard errors. The model includes a range of demographic background characteristics, and then successively add socioeconomic characteristics. Both women’s and their partners’ features are considered, as well as province fixed- effects. 46 TURKEY GENDER COUNTRY ASSESSMENT 2018 Factors that “justify” violence One of the biggest challenges to increasing awareness and reporting of domestic violence, relates to views about its acceptability, as ‘justified’ violence is assumed as deserved and/or expected under certain situations, receiving not societal punishment. With regard to prevalence, in Turkey over the last decade, there has been an important decrease in the share of women agreeing that a husband is justified in physically punishing his wife. TDHS data from 2003 to 2013, indicate that among ever-married women ages 15–49 years the acceptance of domestic violence has gone down to 13 percent of women accepting at least one of the reasons in the survey as a justification for physical violence, almost half of the 2008 results.36 The decline is particularly steep for reasons such “she argues with him” (Figure 3.5). By 2013, the reason women agree most as justification for violence (9 percent) is if the woman neglects the children. TDVS 2014 data indicate that the reason that is seen as most justifiable is for the woman to be unfaithful (36,4 percent), while other reasons show similar levels than those in TDHS. Figure 3.5: Share of Ever-Married Women That Agree That Domestic Violence by the Husband is Justified If the Woman Neglects the Children, Argues with Him, Refuses Sex, or Burns the Food 35 % of ever maried women who agree with 30 she neglects the children (DHS) 25 she argues with him (DHS) the statement she refuses to have sex with 20 him (DHS) she burns the food (DHS) 15 responds to her husband (DV) 10 refuses sex (DV) 5 husband learns woman being unfaithfull (DV) 0 2002 2004 2006 2008 2010 2012 2014 Sources: TDHS (2003, 2008, 2013) and TDVS (2014) The decrease in acceptance is mainly driven by young women. For those married at young ages (15–19), the acceptance of violence by one of the causes covered in TDHS decreases from 63 percent in 2003 to 9 percent ten years later for that age group, and to 20 percent for the same cohort of women. Acceptance has also decreased for all education levels. Indeed, education is the most important driver for the reduction in women’s acceptance of domestic violence over the last decade. Once controlled for education and geographic location, all the age categories appear to be statistically insignificant for the probability to agree with at least one of the statements. The high level of acceptance seen for the 15–19 age group in the 2003 36 The stated reasons in TDHS include “burns the food,””, “goes out without telling him,””, “argues with him,””, “neglects the children,”” and “refuses to have sexual intercourse with him.””. 47 TURKEY GENDER COUNTRY ASSESSMENT 2018 TDHS is caused by the relatively low level of education of women in this age group, and not only their age. Men’s controlling behaviors are related to the reasons for violence, are. In Turkey, according to THDS around 40 percent of ever-married women ages 15–49 are subject to some level of control by their husbands, and while the prevalence of controlling behaviors has decreased between 2008 and 2013, from 43 to 38 percent (Figure 3.6), levels slightly above those observed for the EU-28 countries—35 percent of psychological violence is due to controlling behaviors—covering a wider range of behaviors than TDHS.37 The majority of women experiencing controlling behavior do so in relation to their freedom of movement, with their husbands wanting to know there whereabouts. The 2014 TDVS data confirms that limited freedom of movement is the most experienced controlling behavior by women (Figure 3.7) at 62 percent. TDVS also collects information on other behaviors related to women’s economic participation, and finds that 24 percent of ever-married women (ages 15–59) declare that their husband prevented them from working or causing them to quit their job, with 10 percent of these women reporting this behavior occurrence in the last 12 months. Figure 3.6: Percentage of Ever-Married Women Having Experienced Controlling Behavior by Their Husbands 50 % ofwomen that experienced 40 controlling behavior 30 2008 2013 20 10 0 limit contact prevent to distrusting accusing her knowing at least one to family see friends her with being where she is money unfaithful Source: TDHS (2008, 2013) As with other related factors, education seems to be the most important preventive factor associated with a lower incidence of violence. The more the woman is educated, the lower her probability of experiencing a controlling behavior by her husband. In this case, the husband’s level of education also matters. Controlling by the husband’s level of education renders statistically insignificant the woman’s education level for all behaviors, except for the “accusing her to be unfaithful” one. As women’s education level is highly correlated with the level of education of their husbands, the protective role of education doubles its relevance. 37 In the FRA survey, women were asked the following questions concerning controlling behavior of their partner: How often does your current partner/did your previous partner ever: “Try to keep you from seeing your friends,””, “Try to restrict your contact with your family of birth or relatives,””, “Insist on knowing where you are in a way that goes beyond general concern,””, “Get angry if you speak with another man,””, and “Become suspicious that you are unfaithful?” 48 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 3.7: Percentage of Ever-Married Women (15–59) Who Reported Having Been Subject to Controlling Behaviors by Their Husbands or Intimate Partners (Lifetime Prevalence by Type of Behavior) Blocking social network sites Not sending woman to health institutions… Accusing her of being unfaithful Getting angry when she talks with other men Interfering with her clothing Ignoring Always wanting to know where she is Preventing from seeing her family Preventing from seeing her friends 0 10 20 30 40 50 60 70 Source: Survey on Domestic Violence Against Women in Turkey, 2014 Policy action to prevent violence against women Initiated by the 2001 Civil Code, important legal and policy reforms and nongovernmental initiatives have been undertaken in Turkey to address violence against women. Turkey ratified the Council of Europe Convention on Preventing and Combating Violence against Women and Domestic Violence (Istanbul Convention) in 2012. Building on it, the Law No: 6284 on Protection of Family and Prevention of Violence against Women came into force in 2012. The law, which broadens the concept of gender-based violence to comprise physical, verbal, sexual, economic, and psychological violence, also includes provisions to protect the identity of victims and families and to increase the efficiency of protection orders through sanctions. The law has led to the improvement of the infrastructure and instruments available for the protection of victims. First, it mandated the establishment of ‘Violence Prevention and Monitoring Centers (SONIM)38, which provide support for preventing violence and monitoring the efficient implementation of cautionary decisions. A total of 49 SONIMs were active by the end of 2016, and over 90.000 women had benefited from them (DGSW 2016). A new protocol was signed between DGSW, the Ministry of Interior (MoI), and the Ministry of Justice for using surveillance systems and technical methods in combating violence against women in 2015. A pilot project has been implemented in İzmir and Ankara where electronic monitoring system infrastructure and electronic bracelet are used to monitor both the victim of violence and perpetrator. Finally, a new Regulation on the Opening and Operation of Women’s Shelters resulted in the renovation of 75 women’ shelters, and the development of 15 new shelters 2013 and 2015 (DGSW 2015). 38 http://www.resmigazete.gov.tr/eskiler/2016/03/20160317-8.htm 49 TURKEY GENDER COUNTRY ASSESSMENT 2018 The DGSW has been implementing the EU project ‘Combating Domestic Violence’, launched on December 27, 2013, which aims to establish new—or develop the capacity of existing—support services for women victims of violence throughout 26 provinces.39 Within the scope of the project, the quality of shelter services as well as their coordination is being enhanced, training for service providers is delivered and a variety of activities and campaigns to raise public awareness are conducted.40 More recently, the DGSW also issued a national action plan specifically on ‘Combating Violence against Women’ covering the period 2016–2020. The plan aims to improve the existing conditions regarding five main fields: legal regulations, awareness raising and transformation of mentality, protective and preventive service provision and empowerment of victims of violence, regulating and implementing health services and inter-institutional cooperation and policy development. .. In view of the activities carried out by DGSW, the MFSP updated The National Strategy Paper on Child Rights and the Action Plan in 2014, to fight early and forced marriages using a holistic approach. The update covered identifying and keeping track of the girls without access to education and ensuring that they return back to school, identifying adolescent pregnancies and adopting due measures for notification thereof and carrying out information activities accordingly. However, despite all these new developments, it must be noted that the implementation of the new laws has not always been effective (Tozlu and Goksel 2016) and further work is required. According to justice statistics, the number of victims of sexual assault has increased from 1,091 in 2001 to 2,524 in 2008 (TSI 2015), while in 2013, legal statistics recorded 16,748 female victims. Most of these are victims of domestic violence (73 percent). 3.3. Women’s political participation Despite the progress registered in Turkey in recent years, the presence of women in public institutions is still quite limited, especially from the international perspective. According to World Bank Gender data the proportion of Parliament seats held by women has increased in Turkey since 2000 (see Figure 3.8); however, at 14.6 percent in 2017, it remains well below the ECA average of 27.6 percent. The UN Economic Commission for Europe (UNECE) also highlights that out of 52 European countries only 11 have a lower proportion of women in the national legislature than Turkey. The situation is even more dramatic when examining female representation in government ministries. The proportion of women in ministerial positions, at 4 percent, compares poorly with the average 23 percent registered in ECA for the year 2016. According to UNECE data, only two countries in Europe have a lower ratio of women as government ministers. 39 Adana, Afyonkarahisar, Ankara, Antalya, Bursa, Çanakkale, Denizli, Diyarbakır, Düzce, Erzurum, Eskişehir, Gaziantep, Isparta, İstanbul, İzmir, Kırşehir, Kocaeli, Konya, Manisa, Mersin, Nevşehir, Sakarya, Samsun, Şanlıurfa, Trabzon, Van 40 For detailed information on those projects, see http://www.siddetlemucadele.net/tr/hibe-projeleri 50 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 3.8 Women in Political Representation Positions in Turkey Women in Parliament (percent) Women in ministerial positions (percent) 30.0 25.0 25.0 20.0 20.0 15.0 15.0 10.0 10.0 5.0 5.0 0.0 0.0 2000 2005 2010 2015 2016 2017 2000 2005 2010 2015 2016 Turkey ECA Turkey ECA Women in local governments (percent) 12% 10% 8% 6% 4% 2% 0% Mayor City Counsilor Member of Head of Member of Head of Member of Provincial Village Village Council Quarter Quarter Assembly Council 1999 2004 2009 2014 Source: WDI 2016 and TURKSTAT-MoI- General Management of Local Authorities At the local level the picture does not change much. Only 4 percent of the representatives in local governing bodies are women, and across different bodies, although increases in the share of female mayors to 4 percent, of city councilors to 11 percent, and to provincial assemblies members to 5 percent; only two countries in Europe show a lower percentage of women in local government positions than Turkey, according to UNECE. Two studies of election show limits faced by women in Turkey when seeking public office. First, an analysis of the results from the local election held in March 2014 in Canakkale Province (Bozatay and Kutlu 2014) shows that the number of female candidates and elected representatives is much lower than that of men, far behind the internationally accepted critical level of 30–40 percent. The study highlights that women tend to be nominated at the very end of the lists, which lowers their chances to be elected, and that the number of women candidates was higher for central districts compared to rural ones. Another study of the content of news reported in local media in Antalya and Isparta during the 2007 elections (Caglar 2013) additionally highlights that media coverage of male candidates was much higher in both provinces, while the focus in the case of women was on personal characteristics such as their marital status and family, and on women issues. 51 TURKEY GENDER COUNTRY ASSESSMENT 2018 In collaboration with UN Women and the government of Sweden the parliamentary Committee on Equality of Opportunity for Women and Men has recently launched (December 2015) a 3-year project aimed at increasing gender equality in political participation in Turkey, mainly through the development of policies and legal reforms that can help facilitate and promote women’s engagement and leadership in this area. 3.4. The legal, institutional and policy framework for gender equality Latest legal reforms on gender issues Reforms aimed at the equalization of women’s and men’s legal status in Turkey had an early start in the past century, with the recognition of women’s right to vote and to be elected to office and Parliament in 1930. Since then, other legal changes have ensued with regard to women’s compared to men’s socioeconomic rights, although the irregular intervals between such measures and the lack of enforcement of some of them suggest that the political commitment toward gender equality in the country has been rather limited or volatile. As a result of many of the advances registered over decades, and according to the OECD Social Institutions and Gender Index (SIGI),41 Turkey (which ranks 35 out of 108 countries) shows a low level of institutional discrimination, at par with countries such as Morocco or Madagascar—or Ukraine and Kazakhstan when only considering ECA. Turkey appears to perform particularly well with regard to institutions regarding access to assets, followed by family law and institutions protecting physical integrity. In a similar vein, the CEDAW Committee, in its latest (June 2016) observations on the seventh periodic report of Turkey especially commends the recent efforts deployed with regard to the legal and policy framework for gender equality in the country.42 Although the 1961 Constitution already prohibited discrimination on the grounds of sex and Turkey became a signatory country to and ratified CEDAW in 1985, some of the basic rights of women, for instance to work without their husband’s consent, were only recognized much later (the right to work, in particular, in 1992). Indeed, it was not until 2001, with the enactment of the new Civil Code, that the country started taking the necessary steps toward the recognition of women’s rights on an equal footing with men, especially with regard to family matters. Prior to the new Civil Code, men were legally deemed responsible for the maintenance of the family, as head of the household. However, the new legislation established that both men and women have the same rights in marriage, particularly with regard to property and parental authority. It also increased the legal marriage age for women from 15 to 17 years and granted unmarried women the legal custody over their children. After initiating divorce, children could be placed under the custody of any of the parents, who would share the economic responsibility over them. Women and men also deemed to have equal inheritance rights. Family courts were created in 2003 to ensure the enforcement of the civil legislation in this regard. 41 Which considers not only formal but also informal laws, social norms, and practices that restrict or exclude women and consequently curtail their access to rights, justice, resources, and empowerment opportunities. 42 CEDAW (2016). 52 TURKEY GENDER COUNTRY ASSESSMENT 2018 In 2004, an addendum to the Constitution included the principle of gender equality, and the State responsibility to ensure it (Gökşen et al. 2014). Since 2011, this central legal instrument explicitly states that family is based on equality between spouses. Starting in 2005, legislation has been enacted on domestic violence, explicitly criminalizing marital rape and addressing sexual harassment. In 2012 the Law on Protection of Family and Prevention of Violence against women was issued, and Violence Prevention and Monitoring Centers were established. With regard to legislation on employment, the new labor law has made it illegal for employers to discriminate among workers on the basis of gender, as it mandates equal treatment in hiring, remuneration, promotion, and termination of job contract. The reform has also shifted the burden of proof in cases of sexual harassment to employers, has extended paid maternity leave from 12 to 16 weeks, and given women the option to take an additional six months of unpaid leave (24 for public sector workers).43 Moreover, the new legislation has prolonged the paid paternity leave from 3 to 10 days and entitled men up to 24 months of unpaid leave. More recently passed legislation is seeking to increase women’s access to childcare services by eliminating taxes to childcare centers, and expanding the coverage of pre-primary education by the MoNE. Social security legislation recognizes social benefits to particular groups of women out of the labor force, such as spouses, widows, and daughters on the basis of their dependency status (Dedeoglu and Elveren 2012). In 2011, new regulation also introduced the opportunity for home-based working women to have access to social security with lower contributions. However, gaps persist in regulation and, particularly, in its enforcement. Progress is especially slow and observed gaps concerning in the fields of civil rights and violence against women. Although early marriage is prohibited by law, different sources44 confirm its high prevalence among girls from disadvantaged backgrounds, and its wide social acceptance as a means to relieve families from the economic burden that girls may still represent for some families. With regard to employment, labor legislation does not guarantee mothers an equivalent position after maternity leave and non-pregnant and non-nursing women cannot undertake the same jobs as men (Women Business and the Law database). The Penal Code still allows genital examinations for virginity testing authorized by a judge or prosecutor45, while the 43 The Law on Public Servants was also amended in 2011, extending the unpaid maternity leave from 12 months to 24 months for women in the public sector - 2014 Analysis of Public Programs Related to Women’s Entrepreneurship and Access to Labor Markets. 44 European Parliament 2011, Federation of Women Associations of Turkey (2010) , 2011 Report by the Parliamentary Commission on Equality of Opportunity for Women and Men, Güsten (2011); U.S. Department of State (2013). 45 Comments provided for this report by the Ministry of Justice states the following about this: “A rticle 76 of Criminal Procedures Law No. 5271 governs the procedures for physical examination to obtain evidence in the scope of an investigation. The objective of subject article is to help unveil material facts in cases where there is suspicion that a sexual offense has been committed. Therefore, the subject arrangement is not a virginity control procedure, but is rather an evidence collection instrument intended to enlighten suspicions of offense. The Law clearly provides that no action that could harm the wellbeing of the victim or no surgical operation can be carried out during the examination. On the other hand, the examination decision taken by the Public Prosecutor in cases where a delay cannot be tolerated is submitted to the Judge for approval, and any decision taken on this subject can be challenged. ” 53 TURKEY GENDER COUNTRY ASSESSMENT 2018 regulation on abortion established that the termination of pregnancy of a married woman is subject to spousal authorization. Main actors in charge of gender equality The institutional and policy framework for gender equality in Turkey is depicted in Figure 3.9. At the central level, the main governmental institution for gender equality in Turkey is the DGSW, which functions under the MFSP. The mission of the directorate is to ensure gender equality through the empowerment of women and the development of strategies and policies to prevent all kinds of discrimination, always working in coordination with other partners.46,47 In accordance with its mandate, the DGSW not only works in priority setting and policy formulation but plays an active implementation role, and has the capacity to legislate. Given the variety of institutions involved in the implementation process of gender equality policies, the DGSW carries out its activities in cooperation with these organizations through joint projects and programs, and through protocols. DGSW additionally monitors the actions of other implementing agencies, assessing progress and results.48 Given that services are provided by the provincial and district governorships, which operate as local extensions of the MoI and other line ministries, both such local levels of government and the MoI are among the most important stakeholders of the DGSW. Some other institutions that contribute to the public efforts regarding gender equality include the MoD, the MoLSS, the MoNE, the Ministry of Health (MoH), the Ministry of Justice, Ministry of National Defense, the Turkish Labor Agency (ISKUR), the MFAL, the Social Security Institution (SSI), the Ministry of Environment and Urbanization (MoEU), the Presidency of Religious Affairs, the Regional Development Agencies (DAs) and the State Personnel Department (SPP). Among these, the MoD is responsible for preparing Development Plans and programs49 in coordination with all relevant stakeholders, and for guiding, coordinating, and monitoring their implementation. At the local level, DAs and regional development administrations conduct their activities as an affiliated organization of MoD. DAs, which are established in 26 NUTS 2 regions, play a specific role in better understanding and analyzing local dynamics, preparing regional action plans, coordinating and implementing projects on-site, and monitoring, 46 http://kadininstatusu.aile.gov.tr 47 The specific duties of the DGSW as mandated by law include: (1) the identification of policies and strategies aimed at preventing gender-based discrimination, protecting and promoting women’s human rights and social status and ensuring that women are active in all fields of social life; (2) the implementation, monitoring and assessment of such policies and strategies; (3) the coordination of the social services´ activities toward the protection, education, development, guidance, rehabilitation and advancement of women and the prevention of discrimination; (4) conducting activities and projects to prevent all forms of violence against women; and (5) the identification and development of principles, procedures and standards relating to social services for women provided by public institutions and agencies, voluntary organizations, natural and legal persons, and ensuring the compatibility of such standards. 48 Under the Decree Nr.633. 49 Medium-Term Programs, Annual Programs, and Investment Programs. 54 TURKEY GENDER COUNTRY ASSESSMENT 2018 evaluating, and coordinating services.50 Another important institution is the MoLSS,51 which contemplates gender equality in labor markets as a relevant objective of the National Employment Strategy and has implemented various programs (for example, the ‘Gender Equality at Workplace’ award) to attain that objective. Policy documents emphasize the need to coordinate and collaborate with civil society organizations (CSOs) during the implementation of projects and programs. CSOs, as part of their mandate to eliminate violence against women, conduct research, gather data on violence, and develop problem-solving strategies to respond to the physical, psychological, social, financial, and legal problems arising from this type of violence. Similarly, the private sector has increasingly engaged in activities related to gender equality, especially through social responsibility projects, although these efforts remain limited. Examples of these initiatives are provided by the projects implemented by Intel Technology Services Limited Company, Borusan Holding and Sabancı Foundation. Other public institutions for the protection and promotion of gender equality in Turkey include the Committee on Equality of Opportunity for Women and Men and the Ombudsman, both attached to the Turkish Parliament. The committee, established in 2009, is responsible for issuing periodic reports on the progress toward gender equality and raising public awareness on this topic through different programs. Citizens have the right to issue individual complaints to the committee on matters related to gender-based discrimination (Gökşen et al. 2014). The Ombudsman was set up in 2012 as an independent and efficient complaint mechanism for the delivery of public services, and includes a woman Ombudsperson specially charged with the issue of “women’s and children’s rights.”52 50 Specifically, they: (a) prepare the regional development plans with contributions from MoD and all local stakeholders in their region, and design programs accordingly; (b) provide financial and technical support contributing to the implementation of regional plans and programs that cover different areas including gender equality; (c) contribute to the improvement of the capacity of the region; (d) improve the cooperation between the public sector, private sector and non-governmental organizations, and (e) carry out research. NUTS stands for Nomenclature of Units for Territorial Statistics. 51 ISKUR, SSI, and SPP are affiliated organizations of the MoLSS. 52 http://www.ombudsman.gov.tr 55 TURKEY GENDER COUNTRY ASSESSMENT 2018 Figure 3.9: Policy and Institutional Framework for Gender Equality in Turkey Policy Framework MFSP, The National Action Plan DGSW for Gender Equality Program for Improving Labor Market Action Plan MoD Effectiveness National Development Plan Conservation of Family and Dynamic Action Plan Population Structure Central Level National Employment Action Plan MLSS Strategy 2014 - 16 Progress Towards Gender Equality Financial and Development Regional Technical Agencies Plans Assistance Local Level Regional Action Plans Development Administrations Implementation Central Implementation Public Institutions Provincial Directories MoI Local Development Agencies Implementation Regional Development MoD Administrations Civil Society Private Institutions 56 TURKEY GENDER COUNTRY ASSESSMENT 2018 Central gender equality policies and programs The National Development Plans, prepared by the MoD for 5-year periods, are the fundamental policy documents in the area of gender equality. The latest—the 10th Development Plan for 2014–2018—features gender equality as a key factor in achieving growth and development targets through female labor market inclusion, and, more explicitly, under the head ‘Family and Women’. Within the later component, the overall objective is “to empower women in all aspects of social, economic and cultural life, to improve the status of family while preserving the institution and to strengthen social integration” (MoD 2013). To achieve gender equality the plan sets forth a number of specific goals and actions: (a) enhancing women’s participation in decision making, (b) increasing women’s employment by promoting the conciliation of family and working life (through parental leave and expanded child care services), (c) enhancing the educational level and skills of women, (d) eliminating all forms of violence and discrimination against women, specifically through awareness raising actions at all educational levels, and (e) raising awareness on gender-responsive budgeting. An important feature of the 10th Plan is the Priority Transformation Programs (PTPs), which focus on structural reform areas that are critical for the achievement of the 2023 targets. The issues tackled through PTPs usually fall under more than one ministerial area of responsibility and thus require the effective coordination and responsibility sharing among institutions. For each PTP, an action plan is developed with the coordination of MoD; contributions from other relevant actors and responsible institutions are determined by subcomponents in each action plan. Women are among the target groups in two PTPs, the Program for Improving Labor Market Effectiveness and the Program on Conservation of Family and Dynamic Population Structure. In addition to national development plans, national sectoral plans developed by the DGSW provide the specific policy framework for gender equality. The National Action Plan for Gender Equality for the period 2008–2013, covered the areas of education, economy, poverty, health, power and decision making, environment and media, and includes targets and implementation strategies. The implementation of The Action Plan for Gender Equality 2008- 2013 was monitored by Monitoring Meetings of each priority area by DGSW. In addition, a National Action Plan on the Empowerment of Rural Women has been adopted for the period 2012–2016. 57 TURKEY GENDER COUNTRY ASSESSMENT 2018 CHAPTER 4: CONCLUSIONS AND POLICY RECOMMENDATIONS In the last 15 years, Turkey has made great strides in reducing gender inequalities, particularly in education, health and life expectancy. Nonetheless, the country has seen, over the same period, a much slower increase in female labor force participation as expected given economic growth and employment creation. Starting from much lower levels than similar income countries, changes in women’s participation in the labor market have yet to meet the level and pace of growth of comparable income countries. The limited participation of women in economic activity reflects systematic constraints and barriers that require decisive policy actions. Removing barriers to women’s access to economic participation will have positive effects on women’s incomes and, in turn, on their households’ well-being and the economy as a whole. This assessment highlights three priorities for policy making in Turkey: 1. Long-term economic sustainability depends on employment of the population at economically active age. By maintaining the current structure of labor participation, the demographic dividend can only be partially capitalized, given that only a little over half of the population ages 15–64 are actively contributing to the economy. 2. Human capital formation continues to be key to improve gender equality not only in education and health but as a means to higher participation of women economically or in public roles but in all aspects of the society. 3. Marriage and childbearing are key determinants for women to transitioning into inactivity and staying out of the labor market, as well as for early discouragement of women in seeking employment. 4. Gender disparities are wider among vulnerable groups in Turkey. Tackling the obstacles that women from less-advantaged segments of society face to access opportunities that are valuable to them will not only improve gender equality but overall equality in the country and growth prospects. The following policy tools may be helpful in strengthening gender equality and overcoming multiple barriers that women face to access economic opportunities:  Turkey has done a commendable job increasing the employability of its young population, yet the challenge remains to increase women’s economic participation and productivity. Policies such as affordable childcare, flexible work arrangements and parental leave may increase women’s attachment to the labor market. Women from low-income backgrounds with less education face additional labor market constraints, and may need vocational training and highly subsidized childcare to join or rejoin the labor force.  International evidence shows that availability of affordable childcare is positively correlated with both female labor force participation and fertility. At the same time, 58 TURKEY GENDER COUNTRY ASSESSMENT 2018 robust evidence in developing and developed countries demonstrates that investing in early childhood education has a significant impact on children’s development and their long-term labor and income outcomes. Affordable childcare lessens women’s home- care burden, and also increases the opportunity cost of leaving the labor force. Childcare needs to be heavily subsidized for women from low-income backgrounds.  Relevant policy options for expanding women’s opportunities and enlarging the overall economic process in Turkey include vocational training for women with low education and scholarship programs for girls in tertiary education. 59 TURKEY GENDER COUNTRY ASSESSMENT 2018 REFERENCES A2F Consulting. 2014. Qualitative Assessment of Economic Mobility and Labor Markets in Turkey: A Gender Perspective. Ministry of Family and Social Policies of the Republic of Turkey and World Bank. Aran, M. N. Aktakke, I. Gurol-Urganci, and Atun R. 2015. Maternal and Child Health in Turkey through the Health Transformation Program. Development Analytics Research Paper Series No. 1501. Aran M., Munoz-Boudet, A.M., and Aktakke, N. 2016. “Can Child Care Regulations Make it More Difficult to Serve the Poor?” Journal of Human Development and Capabilities 17(4): 558-582. October 2016. Askin, E. O., and S. Baris. 2015. Kadın Istihdamının Artırılmasında Mikro Kredi Uygulamasının Etkisi: Tokat Ili Ornegi (The Effect of Micro Credit Implementation on The Increase of Women’s Employment: The Case of Tokat Province) Calısma ve Toplum, 2015/3. Atun, R., Aydın, S., Chakraborty, S., Sümer, S., Aran, M., Gürol, I., and Dilmen, U. 2013. Universal health coverage in Turkey: enhancement of equity. The Lancet, 382(9886), 65-99. Ayvaz Kizilgol, O. 2012. Turkiye’de Egitimde Cinsiyet Esitsizliginin Yoksulluk Uzerindeki Etkisi (The Impact Of Gender Inequality In Education On Poverty In Turkey). Yonetim ve Ekonomi, Cilt: 19 Sayi: 1. Balkız, O. I. ve E. Ozturk. 2013. Neo-Liberal Gelisme Anlayısı ve Kadın: Mikro Finans Uygulamaları Kadınları Guclendiriyor mu? (The Neo-Liberal Development Paradigm and Woman: Do Micro-finance Practices Empower Women?) Mediterranean Journal of Humanities, III/2, 1–21. Berktas, E. 2015. “Fertility Intention of Women in Turkey from a Gender Equity Perspective with Focus on Division of Labor within the Home.” MSc thesis, Hacettepe University. Bozatay, Anbarli S., and S. Z. Kutlu. 2014. Siyasette Kadın Temsili Acısından 30 Mart 2014 Yerel Secimi Sonuclarının Canakkale Ili Orneginde Degerlendirilmesi (The Evaluation of March 30 2014 Local Election Results in Terms of Women’s Representation in Politics: the Case of Canakkale) Canakkale Arastırmaları Turk Yıllıgı Yıl: 12, Bahar 2014, Sayı: 16, ss. 131–156. Bozkaya, G. 2013. Kadinlarin Isgucune Katilimini Belirleyen Faktorler: Turkiye Uzerine Bir Analiz. (Factors Determining the Labour Force Participation of Women: An Analysis of Turkey) Sosyal Bilimler Dergisi, Cilt 3 Sayi 5. Caglar, N. 2013. Yerel Medyada Kadin Adaylar. (Female Candidates in Local Media). Suleyman Demirel Universitesi Vizyoner Dergisi Y.2013, C.4, S.8. s.14-38. Cebeci, T. 2014a. Performance of Female Employers in Turkey. Ministry of Family and Social Policies of the Republic of Turkey and World Bank. ———. 2014b. Gender Earnings Gap in the Formal Labor Market in Turkey . Ministry of Family and Social Policies of the Republic of Turkey and World Bank. CEDAW (Convention on the Elimination of All Forms of Discrimination against Women). 2016. Concluding Observations on the Seventh Periodic Report of Turkey. Çevik, A. 2015. “Higher Education in Turkey in the Context of Gender, Upward Mobility and Reproduction of Inequalities: A Comparative Study of the Education and Engineering Faculties”. Doctoral Thesis. The Graduate School of Social Sciences of Middle East Technical University. February 2015. Cuberes, D. and M. Teignier. 2015. “How Costly are Gender Gaps? Estimations for the Balkans and Turkey”, Policy Research Working Papers Series 7319. The World Bank. Dasre, A., A. Greulich, and C. Inan (forthcoming) Combating domestic violence against women in Turkey. The role of women’s economic empowerment. World Bank Working Paper Series. Dedeoğlu, S., & Elveren, A. Y. 2012. Türkiye'de Refah Devleti ve Kadın. İstanbul: İletişim Yayınları. Del Boca, D. 2015. “Child Care Arrangements and Labor Supply”. Working Paper 569, Inter -American Development Bank, Washington, DC. 60 TURKEY GENDER COUNTRY ASSESSMENT 2018 Demirguc-Kunt, A. and L. Klapper. 2012. “Measuring Financial Inclusion. The Global Findex Database”. Working Paper Series 6025. The World Bank: Washington, D.C. Dincer, M. A., and B. Erten. 2015. “An Early Assessment of Extension of Compulsory School Attendance in Turkey: Evidence from a Natural Experiment.” Working Paper. Dublen, Y. 2014. “Female Labor Force Participation and Determinants in Turkey.” MSc thesis, Istanbul University. Er, S. 2013. Turkiye’de Kadinlarin Isgucune KAtilim Oranini Etkileyen Faktorlerin Bolgesel Analizi. (Spatial An alysis of Factors Affecting Female Labor Force Participation in Turkey) Öneri Dergisi, Cilt: 10, Sayı: 40 (20), s.35 -44. Ergin, I., and A. Kunst. 2015. “Regional Inequalities in Self-rated Health and Disability in Younger and Older Generations in Turkey: The Contribution of Wealth and Education.” BMC Public Health 2015 (15): 987. Esen, Y. 2013. “Gender Discrimination in Educational Processes: An Analysis on the Experiences of Studentship.” International Online Journal of Educational Sciences 5 (3): 757–782. Etiler, N. 2015. Calisan Kadinlar ile Ev Kadinlarinin Saglik Durumlari Uzerine Bir Analiz (An Analysis of Health Statuses of employed women and Housewives). Mesleki Saglik ve Guvenlik Dergisi. Cilt: 15, Sayi:57 Feride, İnan and Güneş Aşık. 2015. Making Economies Work for Women: Female Labor Force Participation in Turkey. OXFAM/TEPAV. Gökşen, F., O. A. Olcay, A. Alnıaçık, G. C. Deniz. 2014. An Analysis of Public Programs Related to Women´s Entrepreneurship and Access to Labor Markets. Goux, D., and E. Maurin. 2010. “Public School Availability for Two-year Olds and Mothers’ Labout Supply”. Labour Economics 17: 951-62. Gulay, G., and O. Bener. 2012. Kadınların Toplumsal Cinsiyet Rolleri Cercevesinde Aile Ici Yasamı Algılama Bicimleri (Perception of Family Life in Frame of Gender Roles of Women) TSA / Yıl: 15 S: 3. Hacettepe University Institute of Population Studies. 2004. Turkey Demographic and Health Survey 2003. Hacettepe University Institute of Population Studies, Ministry of Health General Directorate of Mother and Child Health and Family Planning, State Planning Organization and European Union, Ankara, Turkey. ———. 2009. Turkey Demographic and Health Survey 2008. Hacettepe University Institute of Population Studies, Ministry of Health General Directorate of Mother and Child Health and Family Planning, T.R. Prime Ministry Undersecretary of State Planning Organization and TUBITAK, Ankara, Turkey. ———. 2014. 2013 Turkey Demographic and Health Survey. Hacettepe University Institute of Population Studies, T.R. Ministry of Development and TUBITAK, Ankara, Turkey. Karaoglan, D., and C. Okten. 2012. “Labor Force Participation of Married Women in Turkey: Is There an Added or a Discouraged Worker Effect?” IZA Discussion Papers 6616. Kaya, A., and M. Esen. 2015. “Turkiye’de Yonetsel Ekonomik ve Sosyal Bakımdan Kadın Rolunu Gelistirme Arastırması (Enhancing Woman’s Role Research in Turkey in Terms of Managerial, Economical and Social Respects).” International Journal of Human Sciences 12 (2): 595–610. doi:10.14687/ijhs.v12i2.3296. Kilic, D., and S. Ozturk. 2014. Turkiye’de Kadınların Isgucune Katılımı Onundeki Engeller ve Cozum Yolları: Bir Ampirik Uygulama (The Barriers to the Labor Force Participation of Women in Turkey and Solution Suggestions: An Empirical Application). Amme Idaresi Dergisi, Cilt 47, Sayı 1, s. 107–130. ———. 2014. “Gender Differences in Cigarette Consumption in Turkey: Evidence from the Global Adult Tobacco Survey.” Health Policy 114 (2): 207–214. Kirdar, M. G, M. Dayioglu-Tayfur, and I. Koc. 2014. “Does Longer Compulsory Education Equalize Schooling by Gender and Rural/Urban Residence?” IZA Discussion Paper No: 7939. 61 TURKEY GENDER COUNTRY ASSESSMENT 2018 Kizilaslan, N., and M. Karaomer. 2015. Hatay Ili Kırsalındaki Kadınların Girisimcilik Egilimlerinin Incelenmesi (Determination of The Role of Women Entrepreneurship in The Rural of Hatay City) Turk Tarım – Gıda Bilim ve Teknoloji Dergisi, 3 (6): 371–379. Klapper, L., and S. Singh. 2014. The Gender Gap in the Use of Financial Services in Turkey. Ministry of Family and Social Policies of the Republic of Turkey and World Bank. Kukrer, M. 2015. Toplumsal Cinsiyet Esitligi Yonuyle Ortaokul Turkce Ders Kitaplarinin Incelenmesi (The Analysis of Secondary School Textbooks In Terms of Gender Equality). MSc thesis, Abant Izzet Baysal Universitesi. Mateo Diaz, M. and L. Rodriguez Chamussy. 2016. Cashing In on Education. Women, Childcare and Prosperity in Latin America and the Caribbean. Latin American Development Forum: Washington, D.C. Mercan, and Karakas. 2015. “Industry-Level Female-Male Wage Gap in Turkey.” Bilig, Spring, No 7, pp. 157–170. OECD (Organisation for Economic Co-operation and Development). 2014. Education at a Glance 2014. Paris. Okten. K. 2014. Female Entrepreneurship in Turkey: Patterns, Characteristics and Trends. Ministry of Family and Social Policies of the Republic of Turkey and World Bank. Tomul, E. 2009. “Measuring Regional Inequality of Education in Turkey: An Evaluation by Gini Index.” Procedia Social and Behavioral Sciences 1 (2009): 949–952. Tozlu Ç., Göksel A. 2016. “WAVE: Violence against Women Country Report Turkey” UNFPA (2015) “Combatting Violence against Women and Girls in Eastern Europe and Central Asia.” Issue brief 6, 2015. World Bank. 2009. Female Labor Force Participation in Turkey, Trends, Determinants and Policy Framework . Report number 48508-TR. ———. 2012. World Development Report on Gender and Development. World Bank, Washington, DC ———. 2014. Turkey: Promoting Gender Equity in Labor Market and Entrepreneurship. Report No. ACS9213. ———. 2015a. Turkey Gender Policy Note. ———. 2015b. Supply and Demand for Child Care Services in Turkey . A Mixed Methods Study. The World Bank: Washington, D.C. ———. 2016. World Bank Gender Strategy FY16-FY23 ———. Forthcoming. Awareness and Exposition to Domestic Violence among Women in Turkey. Yilmaz, B. 2014. “The Effects of the University Reform in Turkey on Narrowing Gender Gap in Higher Education.” MSc thesis, Bilkent University. 62 TURKEY GENDER COUNTRY ASSESSMENT 2018 ANNEX Table A1. Ordered Probability Model for Life Satisfaction, 2008 1 2 3 4 Female 0.178*** 0.177*** 0.0587 0.0686 (0.06) (0.06) (0.11) (0.11) Age −0.0208* −0.0217* −0.0215* −0.0223* (0.01) (0.01) (0.01) (0.01) Age squared 0.000234* 0.000240* 0.000252** 0.000260** (0.00) (0.00) (0.00) (0.00) Married 0.161** 0.160** 0.0785 0.0837 (0.07) (0.08) (0.12) (0.12) Married Female 0.159 0.151 (0.16) (0.16) Number of children 0.0205 0.028 −0.0227 −0.0159 (0.08) (0.08) (0.12) (0.13) Children Female 0.0882 0.0937 (0.15) (0.15) Employed 0.242*** 0.252*** 0.298*** 0.312*** (0.06) (0.07) (0.08) (0.09) Employed Female -0.0708 -0.084 (0.12) (0.13) Education level No Yes No Yes dummies Source: WVS (2011). Note: *** Significant at 1 percent, ** significant at 5 percent, * significant at 10 percent. The dependent variables is measured by question: “How satisfied are you with your life on scale from 1 to 10, 1 being dissatisfied and 10 being satisfied. 63 TURKEY GENDER COUNTRY ASSESSMENT 2018 Table A2. Returns to Education and Experience, 2016 (Dependent variable log of monthly average wage) Men Women Men Women Age 0.0815*** 0.0679*** 0.0787*** 0.0647*** (0.00160) (0.00284) (0.00150) (0.00277) Age squared -0.000915*** -0.000778*** -0.000905*** -0.000747*** (0.0000) (0.0000) (0.0000) (0.0000) Secondary school 0.187*** 0.278*** 0.147*** 0.240*** (0.00695) (0.0138) (0.00654) (0.0137) Higher education 0.593*** 0.819*** 0.432*** 0.685*** (0.00648) (0.0116) (0.00675) (0.0130) Post graduate education 1.105*** 1.290*** 0.883*** 1.134*** (0.0129) (0.0192) (0.0132) (0.0202) TR2-Bati Marmara -0.127*** -0.152*** -0.153*** -0.151*** (0.0113) (0.0191) (0.0106) (0.0186) TR3-Ege -0.112*** -0.0987*** -0.130*** -0.0871*** (0.00840) (0.0137) (0.00791) (0.0135) TR4-Dogu Marmara -0.116*** -0.127*** -0.116*** -0.126*** (0.00807) (0.0139) (0.00759) (0.0136) TR5-Bati Anadolu -0.0298*** -0.0840*** -0.0667*** -0.0865*** (0.00822) (0.0146) (0.00773) (0.0144) TR6-Akdeniz -0.153*** -0.168*** -0.164*** -0.160*** (0.00881) (0.0149) (0.00832) (0.0147) TR7-Orta Anadolu -0.0981*** -0.129*** -0.141*** -0.150*** (0.0143) (0.0262) (0.0134) (0.0256) TR8-Bati Karadeniz -0.107*** -0.101*** -0.156*** -0.121*** (0.0132) (0.0233) (0.0124) (0.0228) TR9-Dogu Karadeniz -0.117*** -0.104*** -0.177*** -0.120*** (0.0179) (0.0301) (0.0167) (0.0293) TRA-Kuzey Dogu Anadolu 0.00831 -0.0496 -0.112*** -0.0970** (0.0225) (0.0468) (0.0211) (0.0456) TRB-Orta Dogu Anadolu -0.0432** -0.127*** -0.141*** -0.142*** (0.0177) (0.0317) (0.0167) (0.0310) TRC-Dogu Anadolu -0.0860*** -0.0415* -0.162*** -0.0698*** (0.0125) (0.0225) (0.0118) (0.0222) District Center -0.0478*** -0.0430*** -0.0350*** -0.0341*** (0.00585) (0.00950) (0.00545) (0.00922) Subdistrict or Village -0.133*** -0.119*** -0.0972*** -0.0922*** (0.00708) (0.0132) (0.00663) (0.0129) Dummies for sectors Yes Yes included Constant 5.711*** 5.693*** 5.683*** 5.523*** (0.0320) (0.0538) (0.0375) (0.0640) Observations 28,958 13,036 28,958 13,036 R-squared 0.419 0.450 0.498 0.485 Source: LFS, 2016. Note: *** Significant at 1 percent, ** significant at 5 percent, * significant at 10 percent. 64 TURKEY GENDER COUNTRY ASSESSMENT 2018 65