Women Empowerment for Poverty and Inequality Reduction in Sudan June 2020 Poverty and Equity Global Practice Africa Region Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750- 8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. ACKNOWLEDGMENT This paper was written by Eiman Osman (ETC, EA1PV), Alvin Etang (Senior Economist, EA1PV), and Daniel Kirkwood (Gender Specialist, AFRGI) as part of the Sudan Programmatic Poverty Assessment (P164694). Alison Decker (ETC, AFRGI) provided inputs to the paper. Overall guidance was provided by Pierella Paci (Practice Manager, EA1PV). The authors would like to thank Isis Gaddis (Senior Economist, HGNDR) and Nobuo Yoshida (Lead Economist, EA1PV) for providing very useful comments as peer reviewers. i Table of Contents ACKNOWLEDGMENT .......................................................................................................................... i EXECUTIVE SUMMARY ....................................................................................................................... v 1. INTRODUCTION ..........................................................................................................................1 2. DATA AND METHODOLOGY ........................................................................................................3 3. ANALYSES...................................................................................................................................5 3.1. A profile of poverty and gender in Sudan ..................................................................................... 5 3.1.1. Gender differences in poverty through the lifecycle ............................................................ 5 3.1.2. Gender differences in poverty using a demographic taxonomy of households ................... 6 3.2. Gender gaps in endowments ........................................................................................................ 8 3.2.1. Education .............................................................................................................................. 8 3.2.2. Health and fertility .............................................................................................................. 11 3.2.3. Time use .............................................................................................................................. 13 3.2.4. Physical and financial assets ............................................................................................... 14 3.2.5. Financial inclusion ............................................................................................................... 16 3.3. Access to services........................................................................................................................ 17 3.3.1. Water, sanitation, and hygiene........................................................................................... 17 3.3.2. Electricity............................................................................................................................. 19 3.4. Gender gaps in economic opportunities..................................................................................... 20 3.4.1. Labor force participation .................................................................................................... 20 3.4.2. Wage employment.............................................................................................................. 25 3.4.3. Nonfarm household enterprises ......................................................................................... 26 3.4.4. Agriculture........................................................................................................................... 27 3.5. Food security ............................................................................................................................... 31 3.6. Shocks exposure and coping strategies ...................................................................................... 34 3.7. Voice and agency ........................................................................................................................ 37 ii 3.7.1. Women’s mobility ............................................................................................................... 37 3.7.2. Gender-based violence ....................................................................................................... 38 3.7.3. Voting and women’s rights ................................................................................................. 40 4. POTENTIAL IMPACT OF COVID-19 ON GENDER INEQUALITY IN SUDAN ....................................... 42 5. CONCLUSION AND POLICY OPTIONS .......................................................................................... 44 REFERENCES .................................................................................................................................... 50 APPENDIX ........................................................................................................................................ 54 List of Figures Figure 1: Male and female poverty rates by location, age groups, and marital status ................................ 6 Figure 2: Households’ poverty and demographic composition .................................................................... 7 Figure 3: Literacy rates among young females and school attendance by area and quintile ...................... 9 Figure 4: Reasons for not attending school and dropping out by gender .................................................. 10 Figure 5: Fertility rate and life expectancy country comparison ................................................................ 12 Figure 6: Maternal mortality ratio, country comparison ............................................................................ 12 Figure 7: Trends in childhood mortality and nutrition outcomes, by gender ............................................ 13 Figure 8: Average time collecting water by gender, location, head of household’s education, and quintiles....................................................................................................................................................... 14 Figure 9: Asset ownership of female- and male-headed households by transport, house, quintile, and tenure.......................................................................................................................................................... 15 Figure 10: Financial inclusion indicators and mobile ownership by gender ............................................... 16 Figure 11: Access to improved drinking water sources and sanitation sources ......................................... 18 Figure 12: Access to both improved drinking water and sanitation source ............................................... 18 Figure 13: Availability of place to wash hands and water .......................................................................... 19 Figure 14: Access to electricity ................................................................................................................... 19 Figure 15: Labor force participation and employment status: male versus female ................................... 20 Figure 16: Unemployment rate by area, quintile, and states: male versus female.................................... 21 Figure 17: Employment by sector: male versus female .............................................................................. 23 Figure 18: Share of employment by detailed sector: male versus females ................................................ 24 Figure 19: Work category by gender .......................................................................................................... 25 Figure 20: Enterprise survey indicators ...................................................................................................... 27 Figure 21: Main crop grown by household head gender ............................................................................ 28 Figure 22: Input usage by household head gender .................................................................................... 29 Figure 23: Farm size, yield, and proportion sold by main crop grown: male-headed versus female-headed households .................................................................................................................................................. 29 Figure 24: Loan purpose and reasons for not acquiring loans: male-headed versus female-headed households .................................................................................................................................................. 31 iii Figure 25: Food security among households, by area, employment sector, and states ............................ 32 Figure 26: Type of help received among food insecure households .......................................................... 33 Figure 27: Prevalence of shocks across female- and male-headed households, by area, quintile, and livelihood..................................................................................................................................................... 35 Figure 28: Number of shocks and specific types of shocks reported: female- and male-headed households .................................................................................................................................................. 36 Figure 29: Effects and coping strategies: female- and male-headed households ...................................... 37 Figure 30: FGM across women and daughters by education, quintile, and area ....................................... 39 Figure 31: Attitudes toward domestic violence among women................................................................. 40 Figure 32: Sudanese perception on women’s rights: male versus female ................................................. 41 Figure 33: Gender differences in voting ..................................................................................................... 41 Figure A.1: Main crops grown by poor/non-poor: male-headed and female-headed households ........... 59 Figure A.2: Input use by crop grown: male-headed versus female-headed households ........................... 60 Figure A.3: Difference in input use between female-headed and male-headed households .................... 60 Figure A.4: Crop revenue proportion from total revenue: male-headed versus female-headed households .................................................................................................................................................. 61 Figure A.5: Reasons for acquiring loans by household head gender .......................................................... 61 List of Tables Table 1: Primary and secondary net enrollments by gender, area, and quintile (%) ................................... 8 Table 2: School attendance determinants: Males versus females, probit estimation (coefficients and marginal effects) ......................................................................................................................................... 10 Table 3: Determinants of labor force participation by gender: marginal effects ....................................... 21 Table 4: Earnings summary by gender (SDG per month)............................................................................ 25 Table 5: Males versus females households’ enterprises by sector, area, and poverty (%) ........................ 26 Table 6: Average yearly input expenses in SDG by of household head gender.......................................... 29 Table 7: Yearly income and revenue in SDG by household head gender ................................................... 30 Table 8: Food insecurity determinates by household head gender: odds ratios ....................................... 33 Table A.1: Primary and secondary GPI by states (%) .................................................................................. 54 Table A.2: Determinants of labor force participation by gender, probit estimations (coefficients) .......... 54 Table A.3: Earnings summary by industry, gender, and place of residence (SDG per month) ................... 55 Table A.4: Oaxaca-Blinder decomposition of gender gaps in monthly earnings ........................................ 55 Table A.5: Monthly earnings by gender: regression results ....................................................................... 56 Table A.6: Yield determinants: regression results by head of household gender ...................................... 57 Table A.7: Enterprise Survey gender indicators: by type and size of firms (%) .......................................... 58 Table A.8: Agriculture and nonagriculture yearly income breakdown ....................................................... 58 Table A.9: Determinants of food insecurity by household head gender, probit estimations (coefficients) .................................................................................................................................................................... 58 iv EXECUTIVE SUMMARY This paper examines how gender equality has evolved in Sudan during the last decade. The analysis comprises various dimensions including the accumulation of endowment in all its forms (human capital [education and health] and physical capital), access to economic opportunities (labor market opportunities and access to income-generating activities), access to services (water, sanitation, and electricity), and voice/representation to make decision at all levels. The paper includes a discussion on the potential impact of COVID-19 on gender inequality, as well as possible policy options to reduce gender inequality in Sudan. Sudanese women live in poorer than Sudanese men during key productive and reproductive years and appear to suffer greater poverty-related impacts of childcare and divorce. Sudanese women are more likely to live in poor households. Among households with heads in their mid-20s to late-30s, those headed by women are more likely to be poor than those headed by men. Moreover, households headed by a divorced female are more likely to be poor than those headed by a divorced men , with respective poverty rates of 28 percent and 11 percent (p < 0.01). On the other hand, living in a large household is linked to higher poverty incidence for both heads, with female heads being less favored. Also, there is a large gender difference between households with only a single adult, households with only one female adult is twice as likely to be poor compared to households with male adults (24 percent versus 11 percent, p<0.01), In education, gender gaps are shrinking as the proportion of girls attending primary school and the proportion of boys attending secondary school both continue to increase. However, girls are still disadvantaged in primary school enrollments in rural areas, the poorest quintiles, and some states, such as Kassala, East Darfur, and West Kordofan. Boys, on the other hand, have lower secondary school enrollments overall, especially in urban areas and better-off quintiles. The progress Sudan has made in increasing girls’ access to education is reflected in an increase in the female literacy rate—from 52 percent in 2010 to 73 percent in 2014—and data suggest that location (urban/rural) and wealth are more important determinants of inequality in access to education than gender. However, historical gender gaps continue to have an impact on the stock of educated people, with women still representing the majority of those who have never attended school. Sudan’s maternal mortality ratio declined from 463 to 311 (per 100,00 live births) between 2004 and 2014, supported by an improvement in access to reproductive care services. In 2014, three in five women were protected against neonatal tetanus, and more than three-fourths of births were delivered by a skilled attendant. However, disparities in accessing health services are pronounced, with women from rural areas and those with lower educational levels having the lowest access to reproductive care services. In addition, boys are doing worse than girls in childhood health care, as the share of boys’ deaths has been higher than the national level. Time spent in collecting water is a burden to both genders, with no significant difference between adult females (47 minutes a day) and adult males (49 minutes a day), which equates to nearly 1.5 months of full-time job over the course of the year. Living in households from the poorer quintiles is linked to spending more time to collect water for both females and male adults. In addition, average time is higher for both genders when living in a household that the head has no education. v A higher proportion of female-headed households are in the lowest asset index quintile compared to male-headed households, while a lower share of female-headed households are in the highest asset index quintile than male-headed households. This situation is partly underpinned by legal inequalities. Women in Sudan do not have the same property or inheritance rights as men, marriage property follows a separate property regime, and there is no explicit legal provision providing equitable division of property based on nonmonetary contributions (unpaid activities). While most data on assets are by household head, so may reflect the specific vulnerabilities of the women are household heads (as opposed to the majority of women who live in male-headed households), data on access to financial services is available at the individual level and suggests that gender gaps in access to assets are also relevant to the broader population; Sudanese men are twice likely to have a bank account compared to women, and women are less likely to borrow to start operate, or expand a farm or business or to invest in human capital (education and health) compared to men. Male-headed households have better access to water, sanitation, and hygiene (WASH) services and electricity. Individuals living in male-headed households have a better chance of accessing both improved drinking water and sanitation sources at the same time than those living in female-headed households, with 28 percent and 25 percent, respectively. Access to electricity is higher in male-headed households, compared to female-headed households (47 percent versus 43 percent, p < 0.05). Sudan has a large gender gap in labor force participation that contrasts starkly to the average for the Sub-Saharan African region. In 2014/15, Sudanese women’s labor force participation was 33 percent for the working-age population (15–64), compared to 76 percent male labor force participation. This equates to a female-male ratio of 43 percent – around half the gender ratio for the region as a whole. The gender gap is large across the entire country, with the exception of the Darfur and Kordofan regions, rural areas, and poorest quintiles, where females are more engaged in agriculture activities. In addition, two-thirds of those who are out of the labor force and more than half of the unemployed population are women. Female labor force participation is linked to poverty, education, and location. Poor women are significantly more likely to participate in the labor force (relative to non-poor), an incidence that increased by 3 percentage points from 2009. Having no qualification is linked to a higher likelihood of labor force participation compared to completing primary and secondary school for females, while obtaining higher education increases women’s probability of participating by 31 percent. Finally, there has been an increase in women’s labor force participation in rural areas (compared to those living in urban areas). The probability of women being in the labor force increased by 3 points between 2009 and 2014, More men are involved in wage employment than women. Of the total paid employees, 79 percent are men, compared to only 21 percent that are women. In addition, there is a large gender pay gap among wage employees, with Sudanese female wage workers earning 45 percent lower salaries than male wage workers on average. Overall, women are more engaged in agricultural activities, while men are more likely to work in services. However, female heads have lower agricultural yields than male heads across all crops except for millet. Female household heads are more likely to be food insecure and experience higher exposure to shocks, compared to male heads (7.2 percent and 6.4 percent, respectively). Living in rural areas and in agrarian households is linked to high prevalence of food insecurity and shocks. More than four-fifths of food insecure female-headed households live in rural areas, and three-fourths are agrarian households. In addition, more than half of female-headed households in rural areas and more than 60 percent who work vi on crop framing are exposed to at least one shock. Both male and female heads receive help and sell their assets to cope with shocks, as their first and second coping strategies. However, female heads are more likely to reduce household expenses or consumption, while male heads are more likely to borrow. Sudan has one of the world’s highest rates of female genital mutilation (FGM), with nearly 87 percent of the women (ages 15–49) having been circumcised, according to the 2014 Multiple Indicator Cluster Survey (MICS). The prevalence is higher in rural areas, among women with less education and in the richest quintiles. In addition, most Sudanese men and women are reticent about women’s rights in many areas. A high proportion of men and women believe that the husband should have the final say in family decisions (80 percent males, 67 percent females), and more than one-third of the women ages 15–49 justified husbands beating their wives for at least one of five possible reasons. In May 2020, the government criminalized the practice of FGM and made it punishable with three years imprisonment and a fine. Potential impact of the COVID-19 pandemic on gender inequality in Sudan While impacts from the coronavirus disease 2019 (COVID-19) pandemic will be severe for the whole population, there are likely to be differential and greater impacts on women and girls, including across health, education, and economic outcomes. Women may be more exposed to the virus because of their vulnerability as frontline health workers, their exposure to sick patients during facility-assisted childbirths, their disproportionate role of caring for sick household members, and their low agency. COVID-19 may also have direct impacts as women’s likely greater role of caring for sick family members and their higher representation among health care workers mean that they are likely to be more exposed to COVID-19. It has been reported that women represent 65 percent of nurses across all of Africa (Boniol et al., 2009). During the Ebola outbreak of 2014–2016, women’s greater role as household caregivers and frontline health care workers led to higher infections rates (Davies and Bennett 2016). The negative impacts of school closures or reduced funding for non-COVID health services will likely disproportionately affect women, especially adolescent girls, given this is a time when young women make key decisions on their education, reproductive health, and aspirations for the future. The closure of schools may be the end of schooling for some girls. During the 2014–2016 Ebola outbreak, for example, many girls in Sierra Leone did not return to school once the education system reopened (Bandiera et al., 2019). Also, the pandemic may expose women to higher GBV risks, both inside and outside the household as has been reported in other countries. In terms of economic opportunities, there are likely to be disproportionate impacts on women. We do not have any data on Sudan yet, but a survey of four countries in the Asia and the Pacific region (Bangladesh, Pakistan, Maldives, and the Philippines) indicates that a higher percentage of women than men have already seen decreases in their incomes across a range of sources including family businesses, farming or fishing, paid work, and remittances. Given Sudanese women’s lower incomes, lower agricultural yields/higher food insecurity, and lower ability to come up with emergency funds, they may be less able than men to cope with additional shocks brought about by COVID-19. Possible policy options The analysis presented in this report points to a range of possible policy responses that could be used to close gender gaps across a wide variety of key outcomes . Some of the options are summarized as follows. (i) Targeting the most vulnerable and helping women cope with shocks: Given the large regional vii and wealth-related disparities in the size of gender gaps in Sudan, policies across all of the areas discussed should target the poorest households and female-headed households in lagging rural areas, especially those headed by divorced women. (ii) Policies to address gender gaps and constraints in education and skills: To improve outcomes and close gender gaps in the education sector, interventions that address households’ financial constraints are a promising option as evidence suggests they can improve household investments in the least favored children, such as girls. Given the large number of girls who have already dropped out of school and are particularly at risk of entering a vicious cycle of early marriage, childbearing, and economic inactivity, education and skills programs that can also reach these vulnerable girls could be especially valuable. Programs that deliver vocational and life skills training through girls’ clubs that act as safe spaces have been found to be effective in addressing early childbearing and marriage and can be designed to reach those who are still in school as well as those who have already dropped out. (iii) Policies to reduce gender gaps in economic opportunities could focus on building on recent legal reforms to improve women’s access to assets, supporting programs that target social norms to increase the chance that legal changes are reflected in changes in attitudes and behaviors on the ground, and designing programs that support entrepreneurship and agriculture keeping in mind the specific constraints women face. Policies focused on improving women’s secure land tenure could be important not only for women’s agricultural productivity by increasing incentives to make productive investments in land but also for their ability to switch into nonagricultural activities (as they no longer need to guard their land as closely) and for their financial inclusion, given that land is an important source of collateral for loans. Policies to promote women’s entrepreneurship could include those focused on developing their skills. Increasing women’s access to business capital means not only increasing the amount of credit available to them but also ensuring that they are able to exert control over the finance they access. In the agriculture sector, large differences in agricultural productivity and incomes between female-headed and male-headed households suggest that interventions to address underlying constraints to women in agriculture could have large welfare impacts. Interventions could focus on the deficits highlighted in this paper, including female-headed households’ access to productive inputs (pesticides) and labor. (iv) Policies to support women’s voice and agency. Given women’s low status in Sudan, as evidenced by legal inequalities, acceptance of GBV, FGM, and attitudes to women’s role in household and public decision making, efforts to support women’s voice and agency are critical and can support progress across all other areas. Such efforts could focus on shifting restrictive social norms, which would increase the likelihood that recent high-level policy reforms regarding women’s rights (such as the banning of FGM and the repeal of the public order law) would be matched by changes in practices on the ground. viii 1. INTRODUCTION Several key gender issues likely act as an impediment to poverty reduction and shared prosperity in Sudan. While many of these issues are common across countries in the Sub-Saharan Africa region, some of them are accentuated by the status of Sudan as a fragile state. Fragility and conflict negatively affect men and women in different ways, resulting in gender-specific disadvantages. While men are often disproportionately affected by the direct effects of conflict (for example, death and disability), women and girls are affected by a range of constraints and protection challenges that fragility and conflict pose. These issues include disrupted access to basic social services and infrastructure, lower access to productive assets, displacement, and increased exposure to gender-based violence (GBV). A global review of 50 countries, conducted as part of the 2011 World Development Report, finds significant increases in the incidence of sexual and gender-based violence (SGBV) following a major war. In the context of conflict and fragility, SGBV often increases because of a breakdown in the social and moral order, the collapse of security/justice institutions, and the use of SGBV as a purposeful strategy of war. While women and girls are most often victims of SGBV, men and boys are also often targeted for assault and have even less access to resources and networks of support. In Sudan, SGBV risks are especially high among internally displaced persons (IDPs). Sudan has 1.9 million IDPs. Risks for IDPs are especially high for women. For example, in Somalia it is estimated that female IDPs make up 80 percent of reported survivors of sexual violence (UNDP 2012). One form of SGBV that is high in Sudan is female genital mutilation/cutting (FGM), including Type II and III.1 The practice has a dramatic impact on maternal and reproductive health indicators, as it contributes to infections, obstructed labor, and fistulas. The 2014 Multiple Indicator Cluster Survey (MICS) data show that 87 percent of women ages 15–49 years report having undergone some form of FGM. This drops to 32 percent for girls ages 0–14 years, suggesting a decline in the practice, yet 41 percent of women ages 15–49 years still state that FGM should be continued. Conflict and fragility also often contribute to a proliferation of female-headed households, whether through displacement, migration, or mortality. This can have positive implications for women’s empowerment and agency, as more women assume responsibility for household decision making and engage in income-generating opportunities. At the same time, female-headed households are among the most vulnerable, in part due to untenable time burdens between domestic and productive responsibilities and because they often lose access to assets and resources, such as land, which may be accessed primarily through male family members. While fertility rates are lower in Sudan than in many comparator countries, they are still high, with 4.5 births per woman and an adolescent fertility rate of 67 births per 1,000 women ages 15–19. High fertility, especially among adolescents, not only brings health risks for women but also disrupts their school-to- work transition and impedes the country from moving towards a demographic dividend. While Sudan outperforms many of the comparator countries across a variety of gender indicators, women’s economic empowerment appears to be very low, impeding the ability of around half the population from contributing to and benefiting from economic growth. The ratio of female to male labor force participation rates is just 34 percent in Sudan, compared to 70 percent in the Central African Republic, 83 percent in 1 According to WHO, Type III FGM/C entails “Excision of part or all of the external genitalia and stitching/narrowing of the vaginal opening (infibulation).” (UNFPA: http://www.unfpa.org/gender/practices2.htm) 1 Chad, and 104 percent in Burundi, for example. When women do work, they tend to do so in more vulnerable forms of employment. In Sudan, only 18 percent of women are formally employed in comparison to 67 percent of men. Gender equality is a desirable objective not only for its contribution to economic development through increased productivity and better human development outcomes, but also as a goal on its own merit. As in many Sub-Saharan African countries, tradition and cultural heritage sometimes restrict the participation of Sudanese women in economic activities and society as a whole. Women have limited access to education, land, and employment, particularly in rural areas, and are the victims of GBV With Sudan now being a country in transition, this opens a window of opportunity to empower women to increase their involvement in the allocation of resources and in decision making. This study aims to examine how gender equality has evolved in Sudan during the last decade, by looking at different dimensions. These include the accumulation of endowment in all its forms (human capital [education and health] and physical capital), access to economic opportunities (labor market opportunities and access to income-generating activities), access to services (water, sanitation, and electricity), and voice/representation to make decision at all levels. The study will highlight the areas in which gender inequality persists and propose policies to reduce gender inequality in Sudan. 2 2. DATA AND METHODOLOGY The report uses data from various sources to cover the different gender dimensions. Data from the Sudan 2009 National Baseline Household Survey (NBHS), the 2014/15 National Household Budget and Poverty Survey (NHBPS), the 2014 Multiple Indicator Cluster Survey (MICS), Afrobarometer (2015), and Arabarometer (2018) are used to document levels and trends of gender disparities in terms of education and health outcomes, access to basic services, ownership of land, housing and other assets. The evolution of GBV and discriminating cultural norms is also documented using these (and additional country-level) data sources. Differences across regions, educational levels, quintiles, and urban/rural areas are emphasized, and the situation in Sudan is compared to other countries in Sub-Saharan Africa. Information from other reliable sources is also used, such as the World Development Indicators (WDI); the Little Data Book on Gender 2019; Women, Business, and the Law (WBL) report; Gender and Land Rights Database; United State Department of State Country Reports on Human Rights Practices for 2015; and the UN Secretary General’s Database on Violence against Women. It should be noted that some data analysis in this report is presented by gender of household head while some is presented by gender of the individual. This has important implications for how our analysis should be interpreted. Data presented at the household head level, such as data on poverty rates, does not tell us how resources are distributed between women and men within a household and, therefore, cannot tell us about the gender gaps between the majority of men and women in Sudan who live in male- headed households. However, household head level data can illuminate the specific vulnerabilities of women who are household heads. For example, in many countries across the region, women tend to become household heads because of divorce or widowhood. These women often have less access to certain assets for a number of reasons - local customs may favor men in terms of ownership or inheritance rights, while even when laws protect the rights of widowed women, property grabbing by in-laws after a husband’s death may make it difficult for a woman to enforce her rights in practice. Understanding the specific vulnerabilities of different types of male and female-headed households can be useful from a policy perspective to identify priority households and assist appropriate targeting. In this report, the use of data gender disaggregated at the household head level versus the individual level is driven primarily by data availability. Data on poverty, access to household services, agricultural activities, food insecurity, shocks, and ownership of assets were available only at the household head level and therefore should be interpreted only as evidence on gaps between female and male-headed households and not indicative of gender inequality more generally. Other data, including data on education, health, employment, women’s voice and agency, and financial inclusion were available at the individual level and are reported as such. The study draws from the ongoing women and youth employment study to capture gender disparities in the labor market and explore women-specific constraints to accessing employment opportunities. This includes looking at indicators such as labor force participation and earnings as well as sectorial and occupational segregation. Traditional earning equations are estimated to determine if women are paid less than men after controlling for education, experience, occupation, and geographical location. It remains to be seen if the findings from the earning equations will lead to different conclusions from what has been observed in other countries. The labor dimension focuses more on examining gender productivity differences rather than estimating the traditional earning equations and decomposition since the latter is generally well-known. 3 Gender gaps are decomposed into an explained component (reflecting differences in the composition of male and female wage workers in terms of education, experience, and so on) and an unexplained component (reflecting differences in the wage structure or returns to these characteristics). Similarly, the analysis studies agricultural productivity differences in (based on the head of household gender) and nonagricultural self-employment (based on the gender composition of the workforce) and the factors driving these differences (such as differences in input use and technology). The study also analyzes if there have been major changes over the past 10 years in the representation of women in high-level political and government bodies, such as Parliament (National Assembly),the cabinet, and state governments. 4 3. ANALYSES 3.1. A profile of poverty and gender in Sudan 3.1.1. Gender differences in poverty through the lifecycle According to the NHBPS 2014/15, there is no statistically significant difference in poverty rates between Sudanese males and females.2 The national poverty rates indicate that both genders have a similar incidence to the national rate, 36.1 percent for males and 36.2 percent for females. However, poverty incidence by gender varies slightly more (though still insignificantly) if we look at urban and rural areas separately. Males lives in poorer households than females in urban areas with 37.8 percent and 36.8 percent incidence of poverty, respectively, and females lives in poorer households in rural areas with 35.8 percent compared to 35.2 percent of males. However, Sudanese women are more likely than men to live in poor households in their mid-20s to late- 30s. Therefore, women in Sudan between the age of 25 and 39 are significantly poorer than men, a core period for their productivity and reproductivity (World Bank 2018b). Men, on the other hand, are significantly poorer than women starting from the age of 45, with no significant difference between men and women after retirement age (Figure 1b). Data also show gender differences at younger ages (0–14), with females being more likely to live in poor households than males, but only significant at the age of 10–14 years (Figure 1b). Looking at the marital status of the head of households, gender gaps exist in favors of males. Divorced female heads are more likely to be poor than divorced male heads, at 28 percent and 11 percent poverty rate (p < 0.01), respectively. Similarly, married female heads are significantly more likely to be poor than married male heads (34 percent and 28 percent, respectively). Male heads have higher poverty rates when they are single or widowed, but the differences are not statistically significant (Figure 1c). 2 Note that these estimates are based on a household-level measure of poverty, thus we are only able to present the share of males and females living in households that are poor/not poor, but can’t investigate more on individual level of poverty. 5 Figure 1: Male and female poverty rates by location, age groups, and marital status a) Poverty rates overall and by location 40% 20% 0% Sudan Urban Rural Male Female b) Poverty rates by age group c) Poverty rates for heads by marital 60% status 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 0-4 5--9 10--14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75+ Never Married Widowed Divorced Married Poverty, Male Poverty, Female Male-headed Female-headed Source: NHBPS 2014/15. 3.1.2. Gender differences in poverty using a demographic taxonomy of households There is a clear link between household size and poverty rates; the larger the household the higher the poverty rate. Both female- and male-headed households’ poverty rates increase with an increase of the household size, with female-headed households having significantly higher poverty rates than male- headed households for all groups except households of 13+ members, and with female-headed households of 10-12 members having the highest poverty rates of all (Figure 2a). Gender differences in poverty vary according to household composition. Households comprising one adult female are more than twice to be poor than households comprising one adult male (24 percent versus 11 percent, p<0.01). This is also the case whether the household has children or not; one female adult households have 30 percent poverty rates when there’re children in the household and 2 percent when there’s no children compared to 15 percent and 1 percent for households with only male adult (. Poverty rates are highest, at 31% among households compromising of 2 adults of same sex or 3+adults, where almost half of Sudan’s poor live in these households. Households compromising 2 adults of opposite sex, which are in most cases married couples, represent 45 percent pf Sudan’s poor population and have a total poverty rate of 26 percent. When considering the impact and incidence of poverty among women, noting that Sudan’s poverty rates tend to be higher in rural areas and that much of the population remains vulnerable is critical. About two- thirds of Sudan’s population is rural, and one-third lives in urban areas; the incidence of poverty remains higher among rural populations, which accounts for roughly 75 percent and 70 percent of extreme and moderate poverty, respectively. Compared to Sudan’s urban population, the rural population is more than 6 6 percent more likely to be in extreme poverty and 10 percent more likely to be moderately poor (World Bank 2019d). People across Sudan remain vulnerable—though the extreme poverty rate decreased to 13.5 percent in 2014 (from 14.9 percent), moderate poverty increased from 40.5 percent to 46 percent (World Bank 2019d). Figure 2: Households’ poverty and demographic composition a) Poverty rates by household size b) Poverty rates by demographic household 90% composition 40% 38% 80% 70% 35% 31% 30% 29% 60% 30% 26% 24% 23% 50% 25% 40% 20% 15% 30% 15% 9% 11% 9% 20% 10% 10% 5% 1% 2% 1% 1% 0% 0% 1--3 4--6 7--9 10--12 13+ 2 adults of 1 adult male, 1 adult male 1 adult Other same sex or 1 adult female combination 3+ adults female Male-headed Female-headed Total With children Without children Source: NHBPS 2014/15.Note: Adults are defined as those between 18-64 years old. All Household taxonomy results are irrespective of the elderly (65+). For the total rates, the results are irrespective of the number of children (<18) the number of children and elderly (65+) living in these households. 7 3.2. Gender gaps in endowments 3.2.1. Education Sudan’s primary and secondary enrollment rates increased from 64 percent to 69 percent and from 20 percent to 27 percent, respectively, between 2010 and 2014, with slightly steeper trends among girls. Overall, gender gaps are shrinking with more girls attending primary school and more boys attending secondary school. The primary enrollment increase was mainly driven by girls’ attendance increase among the poorest quintiles (11 percentage point higher) and urban areas (7 percentage point higher). For secondary enrollments, the increase is driven by boys’ attendance increase among the richest quintiles and urban areas (Table 1). Although enrollment rates have increased, Sudan is still below the average of Sub-Saharan Africa countries at the primary and secondary levels, according to the WDI. However, gender gaps are still pronounced at area, wealth, and state levels. In 2014, the Gender Parity Index (GPI), the ratio of girls’ to boys’ enrollments at primary and secondary levels varied between urban (102 and 112 percent) and rural areas (95 and 102 percent). The difference between the poorest and richest quintiles was 9 percentage points at the primary level and 11 percentage points at the secondary level (Table 1). The GPI also varies across states, with River Nile, South Darfur, North Darfur, and Khartoum recording more than 100 percent GPI compared to Kassala, East Darfur, and West Kordofan which all record less than 90 percent GPI. At the secondary level, the gap is higher in Sinnar which has the highest GPI at 159 percent compared to West Darfur with 63 percent GPI (Table A.1, Appendix). Table 1: Primary and secondary net enrollments by gender, area, and quintile (%) Primary 2010 2014 Total Male Female GPI Total Male Female GPI Total 64 66 62 94 69 70 68 98 Urban 79 79 79 100 86 85 87 102 Rural 58 60 55 92 63 64 61 95 Poorest 43 47 39 82 51 53 49 93 Second 52 54 50 93 56 58 54 92 Middle 65 67 63 94 69 70 68 98 Fourth 82 81 82 102 86 87 86 100 Richest 90 90 90 99 93 92 94 102 Secondary Total 20 18 21 117 27 26 28 107 Urban 31 28 35 127 40 38 43 112 Rural 14 14 15 107 21 21 21 102 Poorest 4 4 4 106 8 8 8 100 Second 5 5 5 87 15 17 13 77 Middle 10 11 9 87 15 14 15 113 Fourth 26 24 28 119 37 35 39 112 Richest 50 43 56 129 64 61 68 113 Source: MICS 2014. Nearly three-quarters of young females are literate. The female youth (15–24 years) literacy rate increased from 52 percent in 2010 to 73 in 2014, with the poor and urban youth having the highest increase (World Bank 2018a). Nevertheless, disparities between rural and urban areas (24 percentage 8 point difference) and between wealth quintiles (49 percentage point difference between the poorest and richest quintiles) are still large. Moreover, despite the increase, Sudan is underperforming compared to the average of Sub-Saharan Africa countries and other neighboring countries (WDI).3 In 2014, one-quarter of the population stated that they never attended school, of which the majority were females. At the national level, 25 percent of the population never attended school, with significant disparities between rural (21 percent) and urban areas (4 percent) as well as between poorest (9 percent) and richest (1 percent). Females were more disadvantaged than males, with 15 percent (out of the 25) never attended school (compared to 9 percent of males), and the majority live in rural areas (13 percent out of the 15 percent) and are from the poorer quintiles (Figure 3). Figure 3: Literacy rates among young females and school attendance by area and quintile a) Young female literacy b) Never attended school 100% 20% 80% 60% 15% 40% 10% 20% 0% 5% Urban Fourth Poorest Richest Total Rural Middle Second 0% Area Quintile Female Male Source: MICS 2014. Lack of funding was one of the main reasons for not attending school and dropping out for boys and girls (4–24 years), but early marriage was an important reason for dropping out only for girls. According to MICS 2014, school fees was one of the main reasons children and adults could not go to school. Distance to school was also self-reported as one of the constraints to not attending school. For dropping out, except for agreement on the importance of the impact of fees, boys and girls reported different reasons: girls dropping out from school are more likely to do so because of marriage , while boys are more likely to do so because of the need to work to support their families (Figure 4). Early marriage was the third most common reason given for girls’ dropout but was not a factor for boys, echoing broader analysis from the region showing that child marriage is a primary reason that girls leave school early (Wodon et al. 2016). MICS 2014 data indicate that 21 percent of girls in Sudan were married by the age of 15, but this rate rises to 26 percent if we look only at girls in rural areas. Though national child protection legislation was introduced in 2010, it does not cover protection for girls against early or forced marriage. Early marriage has negative effects on educational attainment for girls and is strongly correlated with poverty and low mother education (World Bank 2018a). Early marriage can also have cyclical effects, as it often leads to early pregnancies which will prevent a girl from going to school. In line with this tendency, MICS 2014 data show that fertility rates are nearly twice as high for adolescent girls who live in rural areas than for those in urban areas, where marriage rates are lower (World Bank 2018a). 3 World Development Indicators (WDI) 9 Figure 4: Reasons for not attending school and dropping out by gender a) Reasons for not attending school b) Reasons for dropping out by gender others others Fees Fees Child Labour School far Early marraige No School/No education services Disability/Disease Disability/Disease School far No water/Toilet Child Labour No School/No education services Co-education Co-education 0% 20% 40% 0% 20% 40% Female Male Female Male Source: MICS 2014. Poverty, parents’ education, and location of residence are important factors for school attendance for both males and females. Using data from MICS 2014/15, a regression analysis was conducted to study the determinants of school attendance for both males and females by examining the household head’s age and gender, parents’ education, location of residence, and wealth quintile. There is a clear disparity between the rich and the poor; the probability of attending school is 23 and 25 percentage points higher for males and females from the richest quintiles, whereas for both males and females being from the second quintile is linked with only a 5 and 6 percentage points higher probability of attending (compared to the poorest quintiles). Living in rural areas is linked to a lower probability of attendance for both genders, with a 6 and 8 percentage points lower probability for males and females, respectively. Having educated parents (compared to non-educated) is linked with a higher likelihood of attending school for both girls and boys. The education of the father seems more important than the mother’s when looking at parents with primary and secondary education, but the reverse is true for those with higher education. (Table 2). Table 2: School attendance determinants: Males versus females, probit estimation (coefficients and marginal effects) Male Female Coefficient Marginal effect Coefficient Marginal effect Household head age 0.0199*** 0.00474*** 0.0168*** 0.00406*** (0.001) (0.000) (0.001) (0.000) Household head gender (Male=1) −0.0979 −0.0227 0.000408 0.000099 (0.109) (0.025) (0.112) (0.027) Mother’s education - primary 0.221*** 0.0551*** 0.286*** 0.0732*** (0.038) (0.009) (0.038) (0.010) Mother’s education - secondary 0.465*** 0.107*** 0.524*** 0.124*** (0.064) (0.013) (0.064) (0.014) Mother’s education - higher 0.587*** 0.128*** 0.530*** 0.125*** (0.162) (0.028) (0.156) (0.031) Father’s education - primary 0.355*** 0.0902*** 0.357*** 0.0932*** (0.037) (0.009) (0.037) (0.010) Father’s education - secondary 0.514*** 0.124*** 0.565*** 0.138*** 10 Male Female Coefficient Marginal effect Coefficient Marginal effect (0.049) (0.011) (0.050) (0.012) Father’s education - higher 0.469*** 0.115*** 0.403*** 0.104*** (0.106) (0.023) (0.110) (0.026) Location (Rural=1) −0.271*** −0.0625*** −0.327*** −0.0766*** (0.042) (0.009) (0.042) (0.009) Second quintile 0.163*** 0.0506*** 0.189*** 0.0604*** (0.036) (0.011) (0.037) (0.012) Middle quintile 0.552*** 0.152*** 0.600*** 0.171*** (0.044) (0.012) (0.044) (0.013) Fourth quintile 0.826*** 0.205*** 0.916*** 0.234*** (0.057) (0.013) (0.058) (0.014) Richest quintile 0.993*** 0.231*** 0.995*** 0.246*** (0.078) (0.015) (0.076) (0.016) Constant −0.528*** −0.573*** (0.129) (0.132) N 11,444 11,444 10,991 10,991 Pseudo R-squared 0.185 0.204 Note: Reference categories: Household head gender = female; Education = none; Location = urban; Wealth quintile = first quintile. Standard errors in parentheses, *p < 0.05, **p < 0.01, ***p < 0.001. 3.2.2. Health and fertility With the declining fertility rate and estimated life expectancy at 65 years, Sudan is expected to enter into a demographic transition period, where the number of working-age population increases, and the number of dependents decreases (World Bank 2019d). Sudan’s fertility rate has declined from nearly seven children born per women in the 1960s to about four births in 2017. This trend is similar to that for Sub-Saharan Africa as a whole, yet the level of fertility remains high when compared to the average for lower-middle-income countries. Life expectancy, on the other hand, has shown an increase, and is now higher than the average for Sub-Saharan Africa, though it still lags behind lower-middle-income countries. With the majority of Sudanese being young (World Bank 2019d), Sudan is potentially entering into a demographic transition, a stage that calls for investments toward human capital and job creation, so that the large pool of youth that are due to reach working age have the opportunity to contribute to boost their country’s productivity. Without such investments, this population bulge could otherwise prove to be a risk rather than an asset, with dangers to social stability that come with large-scale youth unemployment. More important for a sizable dividend are changes in worker productivity. Smaller family sizes mean that both families and governments have more resources to invest in the health and education of each child. It also means that women are more able to enter the labor force. If the economic environment is conducive, and this large and better-educated cohort finds well-paying work, a first dividend comes as this productive labor boosts family and national income. Longer life spans mean that this large, better-earning cohort will also want to save for retirement. And with the right policies and a well-developed financial sector, a second dividend can come from higher savings and investments, leading to further productivity gains (World Bank 2019d) (Figure 5). 11 Figure 5: Fertility rate and life expectancy country comparison b) Life expectancy a) Fertility rate 70 8 65 6 60 4 55 50 2 45 0 40 1968 1996 1960 1964 1972 1976 1980 1984 1988 1992 2000 2004 2008 2012 2016 1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012 2016 Sudan Sub-Saharan Africa Sudan Sub-Saharan Africa Lower middle income Sudan, Females Lower middle income Sudan, Males Source: WDI. The maternal mortality ratio declined from 463 to 311 (per 100,00 live births) between 2004 and 2014, reflecting improvements in access to reproductive care services. Sudan’s maternal mortality ratio remained lower than the average ratio for Sub-Saharan Africa but still lags behind other lower-middle- income countries, such as Morocco, Tunisia, and Egypt. In 2014, three in five women were protected against neonatal tetanus, and more than three-fourths of the births were delivered by a skilled attendant, although 71.3 percent of the deliveries took place at home4 (Figure 6). Disparities are pronounced in accessing health services, with women from rural areas and those with lower educational levels having the lowest access to reproductive care services (CBS and UNICEF 2016’ 2016). Figure 6: Maternal mortality ratio, country comparison a) Maternal mortality ratio b) Maternal mortality ratio 2014, country 1000 600 510 comparison 500 405 400 353 343 315 500 311 300 200 121 62 33 100 0 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Sudan Sub-Saharan Africa Lower middle income Source: WDI and MICS 2014/15. Boys are doing worse than girls in childhood health outcomes. Although mortality rates declined in Sudan from 2010 to 2014 (World Bank 2018a), the share of boys’ deaths has been higher than the national level. One in every 13 newborn boys died between birth and their fifth birthday, compared to 1 in 17 for girls. The infant mortality rate, which is the probability of dying between birth and the first birthday, 4 MICS 2014/15 12 represented more than three-quarters of the under-five mortality rate. These trends are important for the country’s population health status and are suggestive of the need for further progress in ensuring effective, safe, and good quality health care (World Bank 2018a) (Figure 7a). Overall, Sudan performs poorly in terms of under-five child anthropometric indicators, with pronounced disparities between genders. In 2014, one-third of children under five were underweight (low weight for age), 38 percent were stunted (low height for age), and 16 percent were wasted (low weight for height) (MICS 2014). Boys are more likely to be affected by childhood malnutrition, especially among the severe cases. Malnutrition increases the risk of death and is linked to long-term effects on cognitive abilities and therefore productivity over time (Figure 7b). Figure 7: Trends in childhood mortality and nutrition outcomes, by gender b) Stunting, wasting, and underweight for a) Childhood mortality by gender 100 children less than 5, by gender 25% 78.7 20% 80 68.4 15% 10% 59.4 57.6 5% 60 52 0% 44.2 Yes,severely Yes,severely Yes,severely Yes,moderately Yes,moderately Yes,moderately 40 17.320.5 20 14.1 0 Infant mortality Child mortality Under five Stunting Wasting Underweight Male Female Sudan Boys Girls mortality Source: WDI and MICS 2014. 3.2.3. Time use According to the African Gender Equality Index 2015, 40 billion hours a year is spent collecting water in Sub-Saharan Africa. In Sudan, three out of five households are without any water on premises, according to MICS 2014/15. The presence of water on household premises is less common in rural areas, households where the head has no education, and in poorer quintiles. Time spent in collecting water is a burden to both genders, with no significant difference between adult females (47 minutes a day) and adult males (49 minutes a day), which equates to nearly 1.5 months of full-time job over the course of the year..5 Females are significantly more likely to spent more time to collect water in rural areas compared to urban areas (49 minutes versus 33 minutes) . Living in households from the poorer quintiles is linked to spending more time to collect water for both females and male adults. In addition, average time is higher for both gender when living in a household that the head has no education. 13 Figure 8: Average time collecting water by gender, location, head of household’s education, and quintiles a) Average time to collect water by household members b) Average time collecting water,by 50 49 households members and area 47 48 60 49 50 46 42 43 42 37 44 40 33 29 41 42 42 40 20 38 0 36 Woman (15+) Man (15+) Girls Boys Woman Man (15+) Girls Boys (15+) Urban Rural d) Average time collecting water,by c) Average time collecting water , by by households members and quintile** households members and head of HH 97 100 education** 90 90 82 80 68 80 70 65 59 70 60 50 45 42 48 60 53 50 40 43 49 46 47 35 37 39 38 4543 50 39 3938 414343 42 39 40 30 31 29 32 37 22 40 30 19 30 20 20 10 10 0 Woman Man (15+) Girls Boys 0 (15+) Woman (15+) Man (15+) Girls Boys Poorest Second Middle Fourth Richest None Primary Secondar Higher Source: Own calculations based on MICS 2014. **Note: households with boys collecting water where the head has higher education attainment represent less than 1% of the total sample. In addition, households in the richest quintile where female adults collect water represent less than 1% of the sample. 3.2.4. Physical and financial assets Asset ownership A higher proportion of female-headed households are in the lowest asset index quintile compared to male-headed households, while a lower share of female-headed households are in the highest asset index quintile than male-headed households (Figure 9d). The household asset index is an indication of household wealth and accumulation of assets, which is constructed by weighting different assets (transport and household properties such as trans-animals, mobile phones, television, radio, and so on) owned by households (World Bank 2019a). Overall, male-headed households tend to own more valuable assets compared to female headed-households. On the other hand, female and male-headed households are equally likely to own the land they are cultivating, though female-headed households are more likely 14 to rent (25 percent) compared to male-headed households (21 percent). Partial ownership and communal lands were less than 15 percent and 10 percent for male heads and female heads, respectively. Figure 9: Asset ownership of female- and male-headed households by transport, house, quintile, and tenure a) Transport assets ownership b) House assets ownership 40% 80% 30% 60% 40% 20% 20% 10% 0% 0% Male-headed Female-headed Male-headed Female-headed c) Household asset index quintile d) Tenure ownership 80% 30% 25% 60% 20% 40% 15% 10% 20% 5% 0% 0% Poorest Poor Middle Rich Richest Owned Rented Partially Communal owned Male-Headed Female-Headed Male-headed Female-headed Source: Own calculations based on NHBPS 2014/15. In Sudan, women do not have the same property rights as men (World Bank Group 2018). According to the 2020 WBL report, Sudan scores 40 out of 100 in the “assets” indicator, compared to a median score of 80 across all Sub-Saharan African countries as well as across all lower-middle-income countries. The indicator analyzes gender differences in property and inheritance rights. It also examines whether legislation accounts for nonmonetary contributions, such as unpaid care for children or the elderly, in distributing assets upon the dissolution of marriage. While statutory law provides for equal ownership rights over immovable property for women and men (WBL), customary law and practices play an important role in secure access to land in Sudan, and the Food and Agriculture Organization of the United Nations (FAO) Gender and Land Rights Database6 finds that women in Sudan tend to only acquire indirect access to land through their husbands or male family members and that these derived rights are lesser than primary male rights. This is important as women’s lesser access to land restricts their access to an important source of collateral for business loans and reduces their incentives to make productivity enhancing investments in land (for example, see Goldstein and Udry 2008). 15 Although the default marriage property regime is a separate property regime in Sudan, where each spouse retains ownership and control over the property s/he paid for, land law has established joint titling for married couples. Having titled property is particularly important for women in low-income economies, where entrepreneurship offers a chance to overcome poverty (World Bank Group 2018). Providing an incentive for women to make more productive investments in their agricultural land, secure land tenure/strong property rights are also important for promoting women’s entrepreneurship, as land/property is an important source of collateral for business loans. Nonetheless, Sudanese women are restricted in accessing shares of marital property after marriage dissolution, as there is no explicit legal provision allowing for equal or equitable division of property based on nonmonetary contributions (unpaid activities7). 3.2.5. Financial inclusion Women are still lagging in financial inclusion indicators, while men are more likely to bear the cost of living. There are also key gender differences in how and why Sudanese access different financial services. Sudanese men are twice likely to have a bank account, savings, and pay remittances compared to women. Moreover, women face more difficulties in coming up with emergency funds. Overall, access to a financial institution in Sudan is behind Sub-Saharan Africa and lower-middle-income countries (Figure 10 a, b, c). When women do access financial products, they are more likely to use informal services. For example, only 4 percent of women saved at a financial institution, compared to 11 percent of men. Instead, women prefer to save using a savings club or with a person outside of the family: 28 percent of women used this method, compared to 17 percent of men. Men and women are equally likely to save (41 percent), however they choose different methods to save. Women and men also borrow for different reasons. While women are much less likely to borrow to start, operate, or expand a business or farm, the gender gap in borrowing is much smaller when we look at borrowing for education (9.3 percent of women versus 11 percent of men) or health care (18.2 percent of women versus 22.3 percent of men). This may indicate that financial inclusion could be a particularly pressing constraint for women entrepreneurs. Figure 10: Financial inclusion indicators and mobile ownership by gender a) Financial inclusion by gender Saved any money in the past year Paid utility bills in the past year Borrowed for health or medical purposes (% age 15+) Sent domestic remittances in the past year Account at a financial institution Borrowed for education or school fees (% age 15+) Debit card ownership Saved to start, operate, or expand a farm or business Saved at a financial institution Borrowed to start, operate, or expand a farm or business 0% 20% 40% 60% Female Male 7 Such as child and elder care 16 b) Owning an account, country comparison c) Difficulty in coming up with 60% emergency funds 47% 50% 40% 39% 40% 20% 30% 20% 0% Sudan Sub-Saharan Lower middle 10% Africa income 0% Male Female Male Female Source: Global Findex 2014. 3.3. Access to services 3.3.1. Water, sanitation, and hygiene Access to improved drinking water, sanitation, and hygiene (WASH) sources is essential for good health. Lack of access to WASH sources is linked with a range of diseases such as cholera and diarrheal diseases, as well as determinants of malnutrition such as stunting. In addition, WASH affects women and girls disproportionally, due to both biological and cultural factors (CBS and UNICEF 2016’ 2016). More than two-thirds of the population lives in households with access to improved sources of drinking water, while over two-fifths has access to improved sanitation sources (Figure 11a).8Disparities between head of households’ genders are pronounced. Male-headed households have better access to sanitation sources9 than female-head households (41 percent versus 36 percent), while female-headed households have marginally better access to water sources (69 percent versus 68 percent). Disparities by area and wealth quintile are pronounced, with 70 percent of urban households having access to sanitation compared to 28 percent of rural households. Also, households in the richest quintile are about twice as likely to have access to an improved source of drinking water and 15 times as likely to have access to improved sanitation as households in the poorest quintile (Figure 11 a, b). 8 According to UNICEF, improved sources of drinking water are piped water (into dwelling, compound, yard or plot, to neighbor, public tap/standpipe), tube well/borehole, protected well, and protected spring8. On the other hand, an improved sanitation facility is defined as one that hygienically separates human excreta from human contact, which includes flush or pour flush to a piped sewer system, septic tank, or pit latrine; ventilated improved pit latrine, pit latrine with slab, and use of a composting toilet. 9 The WHO / UNICEF Joint Monitoring Program (JMP) for Water Supply and Sanitation classify otherwise acceptable sanitation facilities which are public or shared between two or more households as unimproved. 17 Figure 11: Access to improved drinking water sources and sanitation sources a) Water sources b) Sanitation Sources 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Total Female Urban Fourth Poorest Richest Male Rural Female Headed Male Headed Middle Second Female Urban Poorest Fourth Rural Richest Male Sudan Female Headed Male Headed Middle Second Gender Area Head of Quintile Gender Area Head of Quintile HH HH Gender Gender Improved sources Unimproved sources Open defecation Improved sources Unimproved sources Source: Own calculations based on MICS 2014. However, individuals living in a male-headed household have a better chance to access both improved drinking water and sanitation sources at the same time (28 percent), than those living in a female- headed household (25 percent).10 There are also significant differences between rural and urban areas as well as wealth quintiles, with only 19 percent in rural areas, compared to 48 percent in urban areas having access to both services at the same time. The poorest household’s access was only 3.4 compared to 75 for the richest households (Figure 12). Figure 12: Access to both improved drinking water and sanitation source Sudan 28% Area Urban 48.4% Rural 19.0% Gender Male Headed 28% of HH Head Female Headed 25% Poorest 3.4% Second 6.5% Quintile Middle 16% Fourth 39.7% Richest 74.7% 0% 10% 20% 30% 40% 50% 60% 70% 80% Source: Own calculations based on MICS 2014. More than half of the population lives in households without a facility for handwashing. Only 42 percent of the population lives in households with a place to wash hands, a similar percentage for male and female heads, with significant disparities between urban and rural areas and wealth quintiles. Poorest quintiles 18 are three times less likely to have a place for washing hands compared to the richest quintiles. More than four-fifths of the population has access to water, yet only two-thirds have soap/detergent at the place for handwashing (Figure 13 a, b). Figure 13: Availability of place to wash hands and water a) Place for handwashing b) Water availability 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Urban Fourth Female Poorest Richest Rural Male Middle Sudan Second Female Poorest Urban Fourth Richest Rural Male Middle Sudan Second Area Head of Quintile Head of Area Quintile HH HH Gender Gender Observed Not Observed Water is available Water is not available Source: Own calculations based on MICS 2014. 3.3.2. Electricity Ensuring access to electricity helps in promoting economic development. Households’ electrification raises women’s employment by releasing women from domestic tasks and enabling microenterprises. In rural South Africa, for example, it increased women’s employment by 9 percent and women worked 8.9 hours more per week, with growth on both the extensive and intensive margins (Dinkelman 2011). Nearly half of the population reported living in households with access to electricity, with female- headed households less likely to have access. Overall, 47 percent of the population has access to electricity, with a significant difference between female-headed and male-headed households (43 percent versus 47 percent, p < 0.05). The highest access was recorded in urban areas and the richest quintile: less than 10 percent in the bottom-40 reported having access compared to 99 percent in the top quintile, and only one-third in rural areas reported having access to electricity compared to more than three quarters of those in urban areas (Figure 14). Figure 14: Access to electricity 100% 80% 60% 40% 20% 0% Female Urban Fourth Poorest Richest Total Rural Male Middle Second Area Head of HH Quintile Gender Yes No Source: Own calculations based on MICS 2014. 19 3.4. Gender gaps in economic opportunities 3.4.1. Labor force participation In 2014/15, Sudanese women’s labor force participation11 was 33 percent for the working-age population (15–64), compared to 76 male labor force participation (Figure 15a). The labor force is those who are currently working, or working but were absent from work, or not working but are looking for job. Overall, females in the labor force are likely to be from rural areas, the Kordofan and Darfur regions, and the poorer quintiles. The gender gap for the labor force participation is large, with the exception of Darfur and Kordofan regions, with females in Kordofan being more than twice likely to be in the labor force compared to males. Indeed, Sudan’s ratio of female to male labor force participation rates (43 percent) is only around half the average gender ratio for the Sub-Saharan African region as a whole. In addition, contrary to males, female labor force participation is higher among poorer quintiles, falling from 44 percent for the bottom quintile to 30 percent for the top quintile. Finally, being a woman is a detriment, as the majority (two-thirds) of those who are out of the labor force and more than half of the unemployed population are women (Figure 15b). Figure 15: Labor force participation and employment status: male versus female a) Labor force participation b) Employment status 100% Out of Labor 50% Force Unemployed 0% Eastern Darfur T20 Urban Khartoum Central 3 2 4 Rural Northern Sudan Kordofan B20 Employed 0% 50% 100% Area Region Quintile Male Female Male Female Source: NHBPS 2014/15. Although the overall unemployment rate in Sudan has decreased,12 women’s rates were more than twice those of men, at 19 percent and 8 percent, respectively, in 2014 (Figure 16a). Unemployment generally is more pronounced in the urban areas, among the youth and the richest quintiles. There is a significant difference between women’s and men’s unemployment in urban areas (19 percent point gap), among the two richest quintiles (14 percent point gap) and among both youth and adults (7 and 11 percentage point gap, respectively). The magnitude of unemployment is less in the rural areas and poorest quintiles, with significant difference between women and men. It is also noticeable that, although youth unemployment rates are higher than those of adults for both women and men, the gender gap is much larger among adults, with adult women’s unemployment rates three times higher than those of adult men. In addition, there are significant differences between the states; female unemployment rates are higher in the northern, Khartoum, and central states of Sudan, but much lower in the west side (Kordofan and Darfur). In 2014/15, the states with the lowest poverty incidence had almost the highest female adult 11 Labor force participation rate is the proportion of the population ages 15-64 that are economically active. 12 from 12.5 in 2009 to 11.3 in 2014, according to the (World Bank,2019d) 20 unemployment rate: Al Gazira (58.3 percent), River Nile (45.3 percent), northern (39.9 percent), Khartoum (30.4 percent), Kassala (19.1 percent), and Sinnar (17.4 percent) (Figure 16b). In those states, adult female unemployment rates were two to five times the unemployment rates of males (Figure 16b). Figure 16: Unemployment rate by area, quintile, and states: male versus female a) Unemployment rate by area b) Unemployment rate by state and quintile 70% 35% 60% 30% 50% 25% 40% 20% 30% 15% 20% 10% 10% 5% 0% White Nile Kassala Blue Nile Khartoum Al-Gadarif Central Darfur Al-Gezira River Nile Northern West Darfur West Kordufan Red Sea North Darfur North Kordufan Sinnar East Darfur South Darfur South Kordufan 0% Urban Poorest Poor Rich Youth Rural Richest Total Middle Adult Area Quintile Age Male Female Total Male Female Source: NHBPS 2014/15. Being poor and living in a rural area are associated with an increased probability of a woman participating in the labor force. Regression results shows that poor women are significantly more likely to participate in the labor force (relative to non-poor), an incidence that witnessed an increase of 3 percentage points from 2009. Although there is a decrease from 2009, having no qualification is still linked to a higher chance of participation in the labor force compared to completing primary and secondary school for both males and females. However, in 2014, having higher education increased women’s probability by 31 percent, compared to those with no education. Living in rural areas rather than in urban areas is increasingly linked to a higher likelihood of labor force participation for women and men. On the other hand, marital status plays different roles for women and men; women who are married or widowed are significantly less likely to participate in the labor force, while men’s probability increases (relative to being single). This is also reflected by results of NHBPS 2014/15 which show that the majority of women (85 percent) who are not participating in the labor market cite being a homemaker/housewife as the main reason—young men, in contrast, primarily cite no hope of finding a job, while older men are either retired or too disabled to participate in the labor market. Finally, adults have higher probability to be in the labor force compared to youth for both genders, except the last category (55–64) where male adults have lower chances to be in the labor force compared to male youth (Table 3). Table 3: Determinants of labor force participation by gender: marginal effects 2009 2014 Male Female Male Female 25–34 0.353*** 0.0773*** 0.275*** 0.0736*** (0.018) (0.013) (0.016) (0.013) 35–44 0.418*** 0.0852*** 0.338*** 0.111*** (0.030) (0.015) (0.023) (0.015) 45–54 0.0912 0.0698*** 0.156* 0.0998*** 21 2009 2014 Male Female Male Female (0.077) (0.018) (0.062) (0.018) 55–64 −0.178* −0.0188 −0.224** 0.0162 (0.082) (0.019) (0.074) (0.022) Rural −0.0454* 0.0444*** 0.0384* 0.0910*** (0.019) (0.010) (0.017) (0.011) (own) Some/completed primary −0.260*** −0.108*** −0.206*** −0.0719*** (0.019) (0.011) (0.018) (0.012) (own) Secondary −0.374*** −0.0648*** −0.175*** −0.0091 (0.021) (0.015) (0.023) (0.017) (own) Post-secondary and above −0.237*** 0.228*** −0.0259 0.307*** (0.037) (0.030) (0.027) (0.022) (head) Some/completed primary −0.029 −0.0391*** −0.0697*** −0.0148 (0.022) (0.012) (0.019) (0.012) (head) Secondary −0.0791* −0.0287 −0.177*** −0.0613*** (0.034) (0.017) (0.032) (0.018) (head) Post-secondary and above −0.180*** −0.0177 −0.160*** −0.0289 (0.039) (0.022) (0.035) (0.021) Married 0.252*** −0.0315* 0.182*** −0.0784*** (0.030) (0.013) (0.025) (0.014) Widowed −0.319*** −0.0295 0.16 −0.139*** (0.090) (0.031) (0.093) (0.031) Divorced 0.0783 0.0612 −0.0658 0.00919 (0.124) (0.033) (0.087) (0.034) Poor=1 −0.0185 0.0708*** −0.0327* 0.102*** (0.018) (0.010) (0.017) (0.011) Observations 5,994 11,582 6,229 14,241 Source: Own calculations based on NHBPS 2009 and 2014/15. Note: Marginal effects after probit estimation (see Table A.2, Appendix). Reference categories: Poor = not poor; Marital status = single; Head’s/own education = primary; Age = 15–24. Standard errors in parentheses, *p < 0.05, **p < 0.01, ***p < 0.001. Women in Sudan are more likely to work in agriculture, while men tend to work in services and manufacturing (Figure 17). Women’s employment is skewed toward agriculture, which accounts for 60 percent of all women’s employment, followed by 38 percent in services. This pattern is reversed for men, with 50 percent employed in services and 38 percent in agriculture. Men are also more likely than women to work in services, 12 percent versus 2 percent. If we dig down into more detailed sub-sectors, we continue to see significant sector-based sex segregation. For example, women are more likely to work in sectors such as ‘activities of household as employers’, ‘education’, and ’human health and social workers’, while men aee more likely to work in ‘construction’ and ‘transportation and storage’ (Figure 18). 22 Figure 17: Employment by sector: male versus female Male Female 38% 38% 50% 60% 12% 2% Agriculture Manufacturing Services Agriculture Manufacturing Services Source: NHBPS 2014/15. 23 Figure 18: Share of employment by detailed sector: male versus females Transportation and storage Construction Elcetricity, gas, steam and air conditioning supply Defence Mining and quarrying Real estate activities Animal husbandry Forestry Water supply, sewerage, waste management and remediaton activities Wholesale and retail trade, repair of motor veichles and motor cycles Fishing Manufacturing Professional, scientific and technical activities Arts, entertainment and recreation Other service activities Accomodation and food service activities Adminsitrative and support service activities Public administration and pulsory social security Information and communication Crop farming and horticulture Financial and insurance activities Human health and social work activities Education Activities of households as employers; for households consumption 0% 20% 40% 60% 80% 100% Male Female Source: NHBPS 2014/15. 24 3.4.2. Wage employment More men are involved in wage employment than women. Of the total paid employees, 79 percent are men, compared to only 21 percent women. Generally, women are underrepresented in most of the job categories; they represent 18 percent of employers and 24 percent own account workers. On the other hand, more women than men work as unpaid workers, either contributing as a family worker or working for others without getting paid (64 percent) (Figure 19). Figure 19: Work category by gender Unpaid Worker 36% 64% Own Account Worker 76% 24% Paid Employee 79% 21% Employer 82% 18% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Male Female Source: NHBPS 2014/15. On average, female-to male wage ratios is 55 percent, meaning that Sudanese female wage workers earn 45 percent lower salaries than male wage workers. According to NHBPS 2014/15, there is a big gender gap among wage employees. This gap is larger at the bottom of the earnings distribution, where women earn 53 percent less than men (Table 4). On the other end, wages vary across industries and location. Working in agriculture is linked to lower wages, compared to working in services and manufacturing. Similarly, working in rural areas is associated with lower wages compared to working in urban areas. Sudanese women’s wages are on the lower end, since they are mainly engaged in agricultural activities in rural areas (Table A.3, Appendix). It is worth noting that the gender gap in earning seems to be expanding over time. Although women also earned less than men in 2009 (according to the 2009 NBHS), the gender difference then was less than in 2014/15. In 2009, the female-to-male earnings ratio was 0.66, meaning that female workers earned 34 percent less than male workers (compared to 45 percent in 2014/15). This statistic does not take into account differences in experience, skill, occupation, education, hours worked, or whether it is full-time or part-time work. The gender pay gap is still larger in rural areas (0.54 in 2009 compared to 0.53 in 2014/15) than in urban areas (0.70 in 2009 compared to 0.58 in 2014/15). While some of the gender earnings differences may be explained by other factors, a significant part of it is attributable to gender inequality. Table 4: Earnings summary by gender (SDG per month) Male Female Female-to-male Ratio Mean 1230 682 0.55 Median 900 587 0.65 10th percentile 300 140 0.47 90th percentile 2000 1200 0.60 Urban 1536 886 0.58 25 Rural 1062 560 0.53 Source: NHBPS 2014/15. Differences in returns to characteristics between male and female wage workers in terms of education, industry, location, and age explain about 96 percent of the gender gap in monthly earnings. The paper uses the Oaxaca-Blinder decomposition (Table A.4, Appendix) which divides the wage gap between men and women into three parts: endowment effect (reflects the mean increase in women’s wages if they had the same characteristics as men), coefficient effect (reflects the change in women’s wages when applying the men’s coefficients to the women’s characteristics), and interaction effect (measures the simultaneous effect of differences in endowments and coefficients). Determinants used in the analysis are education, industry, location, and age. According to Table A.4, Appendix, the mean log wage is 6.75 for men and 6.20 for women, yielding a wage gap of 0.55 (73.5 percent higher wages for male workers). The endowment effect explains about 5.4 percent of this difference, while 95.5 percent is explained by the coefficient effect and negative 0.9 percent by the interaction effect. The large coefficient effect is mainly captured in the regression intercept, possibly showing gender- based discrimination in the labor market (World Bank 2018b). The coefficient effect explains that men benefit from more favorable return to characteristics, except for higher educational attainment where women benefit more. In addition, according to the endowment effect, male wage workers are more advantaged due to their overrepresentation in industries which that are linked with higher wage premiums (Table A.4, Appendix). 3.4.3. Nonfarm household enterprises Female-run household enterprises are more likely to be in services, in urban areas, and with only one- quarter formally registered. According to NHBPS 2014/15, both Sudanese female- and male-run household enterprises are more likely to be in services (84 percent and 97 percent, respectively), with male-run household enterprises being more likely to be in the industry sector and to be registered as well (compared to female-run enterprises). Furthermore, female-run businesses are concentrated in urban areas and nearly two out of five are from poor households (Table 5).13 Table 5: Males versus females households’ enterprises by sector, area, and poverty (%) Employment Sector Male-run female-run Industry 16 3 Services 84 97 Registered 46 25 Share in urban areas 44 65 Share in poor households 26 37 Source: NHBPS 2014/15. Sudan scores lower than the average of Sub-Saharan African countries across all gender indicators in the 2014 Enterprise Survey. Only 3 percent of firms have a female top manager compared to 16 percent in Sub-Saharan Africa, and 2 percent of firms in Sudan have a majority of female ownership, compared to 13 percent in Sub-Saharan Africa. Moreover, only 14 percent of full-time workers, 12 percent of non- 13 Survey questions were limited to head of households only and no follow-up questions on profits and workers 26 production workers, and 9 percent of production workers are female (Figure 20a). On the other hand, manufacturing and large firms (more than 100 workers) have the highest share of female participation in ownership (15 percent and 18 percent, respectively). Retail and medium-size (20–99 workers) firms have a higher share of female workers (Table A.7, Appendix). Firms with female top managers report customs and trade regulations, tax administration, political instability, and access to finance as their top four business environment obstacles, according to the 2014 Enterprise Survey. In addition, firms with male top managers reported the same obstacles but with a lower percentage of incidence. Moreover, female top managers are more likely to cite labor regulations, transportation, and business licensing and permits as obstacles, while male top managers are more likely to cite tax rates, corruption, and practices of the informal sector (Table 20b). Figure 20: Enterprise survey indicators a) Country comparison on gender b) Biggest 10 obstacles indicators: by top indicators manager's gender Permanent full-time workers that are female Customs and trade regulations 28% (%) Tax administration 22% Permanent full-time non- production workers that Political instability 19% are female (%)* Access to finance 10% Permanent full-time production workers that Labor regulations 6% are female (%)* Transportation 5% Firms with female participation in ownership Business licensing and permits 4% Tax rates 3% Firms with a female top manager Corruption 3% Practices of the informal… 0% Firms with majority female ownership 0% 5% 10% 15% 20% 25% 30% 0% 10% 20% 30% 40% Male top manager Female top manager Sub-Saharan Africa Sudan Source: Enterprise Survey 2014/15. Note: *This indicator is computed using data from manufacturing firms only. 3.4.4. Agriculture Two-fifths of Sudanese households use or own agricultural land, forest land, or pastureland, with no difference between male-headed and female-headed households.14 About 39.4 percent and 38.7 percent of male- and female-headed households, respectively, used or owned land for agriculture activities according to the 2014/15 NHBPS. 27 The most common crop grown by households engaging in crop farming differ between male-headed and female-headed households. Male-headed households are more likely to grow sorghum (50 percent) followed by millet (32 percent), sesame (16 percent), and groundnut (15 percent), while female-headed households are more engaged in growing millet (45 percent) and sorghum (36 percent) followed by sesame and groundnut (16 percent each) (Figure 21). On the other hand, poor households (female and male headed) have similar trends for the common crops grown; however, poor households are engaging more in growing millet compared to non-poor. About 43 percent of poor male-headed households grow millet compared to 26 percent for non-poor ones, and 52 percent of poor female-headed households grow millet compared to 40 percent for non-poor ones (Figure A.1, Appendix). Figure 21: Main crop grown by household head gender 60% 50% 40% 30% 20% 10% 0% Sorghum Millet Sesame Groundnut Vegetables Fruits Other crops Male head Female head Source: Own calculation based on NHBPS 2014/15. Note: Other crops include wheat, yellow maize, cotton, roselle, sunflower, Egyptian beans, beans, lentils, Arabic gum, and another crops. Generally, the usage of agricultural inputs is low among crop farming households, with male heads more likely to use a range of inputs than female heads. Nearly three out of five male heads use agricultural inputs, compared to two out of five female heads (Figure A.3, Appendix). Moreover, male heads spend nearly four times more on input use annually compared to females (SDG 4,112 and SDG 1,054, respectively). Labor costs represent the biggest inputs expense for males, while for females it is fuel and lubricants (Table 6). Looking at the four main crops grown by households (millet, sorghum, groundnut, and sesame), both male and female-headed households’ input usage is higher for groundnut and sorghum and is the lowest for millet (Appendix, Figure A.2). 28 Figure 22: Input usage by household head gender Table 6: Average yearly input expenses in SDG by of household head gender Male- Female- Farm repair 11% 16% headed headed Labour cost 9% 14% Labor cost 2,502 910 Fuel and lubricants 2,162 1,252 Pesticides/Fertilizer 5% 11% Farm repair 1,714 598 Machine/equipmen… 10% 9% Other 969 353 Other 4% Pesticides/fertilizer 793 516 4% Machine/equipment 702 299 Fuel and lubricants 2% 1% repairs Average total input 4,112 1,054 0% 5% 10% 15% 20% expenses Male-headed Female-headed Source: Own calculation based on NHBPS 2014/15. According to NHBPS 2014/15, female-headed households achieve lower yields than male-headed households for all crops except millet (Figure 23a). As observed in World Bank (2019b), crops with the smallest plot size had the highest yield production, which is the case for the groundnut yield, followed by sorghum, millet, and sesame for both households. Except for millet, male-headed households have higher yields, with 50, 23, and 11 percentage higher production for groundnuts, sesame, and sorghum, respectively, compared to female-headed households. Female-headed households achieve a 13 percent higher yield for millet than male-headed households. Groundnut, which has the largest gender yield gap favoring men, is also the crop that has the highest proportion of output sold on the market by sesame. Smaller proportions of sorghum and millet production are sold on the market as they are generally food crops that are grown more for household consumption (Figure 23 a, b). Figure 23: Farm size, yield, and proportion sold by main crop grown: male-headed versus female-headed households a) Yield by crop grown b) Proportion of total crop produce 5.00 sold 4.00 80% 3.00 60% 2.00 40% 1.00 20% 0.00 0% Groundnut Sorghum Millet Seasam Groundnut Sesame Sorghum Millet Male headed Female headed Male-headed Female-headed Source: Own calculation based on NHBPS 2014/15. Input usage has a different effect on yield, depending on the gender of the household head. Regression results of the determinates of yield show fertilizer/pesticide input use is positively related to sorghum and groundnut yields for male heads. For female heads, only millet yield was positively related to the use of fertilizers (significant at 5 percent), and although other crops are not statistically significant, they showed 29 a negative relation. Machine use is positively related to male heads’ yields for sorghum, millet, and groundnuts. Acquiring a non-agricultural loan is negatively related to sorghum and millet yields for male heads and sesame yields for female heads. Having more household members is positively related to sorghum and groundnut yield for male heads. On the other hand, having a school degree is positively related to millet yield for female heads. For groundnut yields, having a higher degree is positively related for females and negatively related for males. Moreover, a negative relationship is observed for all crops between plot area and yields, for both male and female heads (Table A.6, Appendix). On average, agriculture accounts for 25 percent of total income of households headed by males and 23 percent for those headed by females (Table 7). Male heads’ yearly agricultural income was twice as that of female heads, while the average yearly income from non-agricultural activities for female heads was higher than that for males, as half of this income comes from domestic and international remittances (Table A.8, Appendix). On the other hand, on average, male heads made three times more crop revenues than female heads (Table 7). Groundnut makes up the highest proportion of total revenues for both households, followed by sorghum and sesame for male heads and vegetables and sesame for female heads (Figure A.4, Appendix). Table 7: Yearly income and revenue in SDG by household head gender Male-headed Female-headed Agricultural yearly income 3,551 1,916 Non-agricultural yearly income 5,822 7,013 Agricultural income percentage from total income 24 23 Total crop revenue 39,837 13,120 Source: Own calculation based on NHBPS 2014/15. Nearly two-thirds of agrarian male-headed households and half of female-headed households had access to some form of credit, mainly through family loans. Male heads are more likely than their female counterpart to use loans for agricultural purposes, while the female heads are more likely than their male counterparts to use loans for consumption (Figure 24a,c). With the low percentage of owning a bank account, all credits were accessed through a family loan, with only 1 percent having accessed it through banks and less than 0.5 percent from microfinance loans. Very few agrarian households owned a bank account, as only 6 percent male-headed households had a bank account compared to only 2 percent of female-headed households. The majority of households acquired loan for consumption needs—88 percent for female heads and 78 percent for male heads. Only 11 percent of male heads and 5 percent of female heads acquired loans for agriculture needs (Figure A.5, Appendix). For those not obtaining any kind of loans, the most common reason for both households was the lack of need for credit, and nearly one- third were afraid of debt (Figure 24b). 30 Figure 24: Loan purpose and reasons for not acquiring loans: male-headed versus female-headed households a) Loan purpose b) reasons for not attaining loans 60% 9% No need 3% Debt Fear 40% Inadequate collatoral Expensive 47% 44% Expectation of Refusal 20% Other No lender 0% Refused Loan Male-headed Female-headed Non-agriculture loan Agriculture loan Debt -10% 10% 30% 50% Female-headed Male-headed c) Reasons for acquiring loan 80% 70% 60% 50% 40% 30% 20% 10% 0% Other Working capital and Religious, wedding, burial Buy other equipment Consumption needs Farm inputs Buy heavy equipment Purchase and improvement On-lending Buy agricultural land Buy animals Other agricultural costs Land and/or building Consumer durables Other business expenses purchase of inputs equipment of dwelling Agriculture Non-Agriculture Consumption Other Male-headed Female-headed Source: Own calculation based on NHBPS 2014/1515. 3.5. Food security According to the 2014/15 NHBPS, 6 percent of Sudanese households were food insecure. The FAO defined food security as: “A situation that exists when all people, at all times, have physical, social and economic access to sufficient, safe and nutritious food that meets their dietary needs and food preferences for an active and healthy life” (FAO et al. 2019). Food insecurity is then the lack of access to food availability, economic and physical access to food, food utilization, and stability over time (FAO et al. 2019). Recent data on Sudan only reflected dietary questions at the household level and did not ask further food security questions at the individual level. 15 Other loans include loans accessed through bank and government agencies, Non-governmental organizations (NGO), microfinance institutions, and employer or landlords 31 Female heads have a higher prevalence of food insecurity compared to male heads (7.2 percent and 6.4 percent, respectively, p < 0.01) (Figure 25a). Female heads’ incidence of food insecurity is high in rural areas, agrarian households, and in the Darfur region (Figure 25). Nearly, three-quarters of food insecure households reside in rural areas, with food insecure female-headed households even more concentrated in rural eras (82 percent) (Figure 25a). Interestingly, food insecurity is more common among agrarian households and, again, this is even more the case for female-headed households, among whom 75 percent of those classed as food insecure work in agriculture(Figure 25c). Except for East Darfur state, all other Darfur states are among the most food insecure, with an even higher incidence among female- headed households. Male-headed households have higher trends than female-headed households in all states, except the abovementioned Darfur states, in addition to Algadarif state (Figure 25c). Figure 25: Food security among households, by area, employment sector, and states a) Overall b) Area 100% 6% 6.4% 7.2% 100% 80% 80% 60% 60% 74% 72% 82% 94% 94% 93% 40% 40% 20% 20% 26% 28% 18% 0% 0% Sudan Male-headed Female-headed Sudan Male-headed Female-headed Urban Rural Food secure Food insecure c) Employment sector 100% d) States 7% 8% 3% 20% 80% 15% 60% 66% 65% 75% 10% 40% 5% 20% 0% West Darfur East Darfur Sinnar River Nile South Darfur White Nile South Kordufan Al-Gadarif Kassala Central Darfur North Kordufan Khartoum North Darfur Red Sea West Kordufan Northern Blue Nile Al-Gezira 27% 28% 21% 0% Sudan Male-headed Female-headed Service Agriculture Industry Male-headed Female-headed Source: NHBPS 2014/15. The majority of the help received by food insecure households comes from food aid organizations, zakat, and family support (Figure 26). Two-thirds of food insecure male-headed households receive help, mainly from food aid organizations (30 percent), zakat (26 percent), and family support (18 percent). 80 percent of female-headed households receive help, largely from the same groups: 37 percent from food 32 aid organizations, followed by family support (28 percent) and zakat (14 percent). Government support represented less than 10 percent of the help received for both households (Figure 26). Figure 26: Type of help received among food insecure households 40% 37% 31%30% 28% 30% 24%26% 20%18% 20% 14% 12%13% 6% 8% 8% 7% 8% 10% 5% 5% 0% Food Aid Zakat Center Non Hh. Members Other NGO Gov't Benefits Other Groups Sudan Male-headed Female-headed Source: NHBPS 2014/15. Among female-headed households, living in rural areas, being unemployed, and working in agricultural activities increase the probability of being food insecure. Employment status appears to be more important for female than for male-headed households in determining the chance of being food insecure, while the opposite is true for education which does not appear to offer the same protection to female- headed households as to their male counterparts. Using multivariate analysis, we estimated the probability of households being food insecure for female- and male-headed households, conditional on different sociodemographic variables. Living in rural areas increases the odds of being food insecure by 76 percent for female-headed households and decreases it by 15 percent for male-headed households (compared to those living in urban areas). Male heads who have completed primary school, secondary, or post-secondary and above are less likely to be food insecure by 39 percent, 51 percent, and 53 percent, respectively, compared to those with no education. Educated female heads, on the other hand, showed no significant difference being food insecure compared to those with no education. Looking at employment status, unemployed female heads have higher odds of being food insecure, with no significant effect for male heads, relative to households with an employed head. For both heads, being an agrarian household is linked with higher probability to be food insecure (27 percent for male-headed and 49 percent for female-headed households). Household size increases the odds of being food insecure. Lastly, receiving help increases the odds of being food insecure for male-headed households but not for female-headed households. Table 8: Food insecurity determinates by household head gender: odds ratios Male-headed Female-headed Area Rural 0.849* 1.761** (0.06) (0.35) Education Some/completed primary 0.615*** 0.826 (0.04) (0.22) Secondary 0.491*** 0.873 (0.08) (0.38) Post-secondary and above 0.470*** — (0.09) (.) 33 Male-headed Female-headed Employment Unemployed 1.313 3.379** (0.28) (1.41) Out of labor force 1.135 1.612 (0.23) (0.78) Sector Agriculture 1.273*** 1.490* (0.08) (0.26) Industries 0.788* 0.96 (0.08) (0.25) Household size Household size 1.236*** 1.229*** (0.01) (0.04) Receiving help Help received 1.219** 0.784 (0.08) (0.12) Observations 7,520 949 Note: Odds ratios after probit estimation (see Table A.9, Appendix). Reference categories: Area = urban; Education = primary; Employment = being employed; Sector = services; Receiving help = no helped received. Standard errors in parentheses, *p < 0.05, **p < 0.01, ***p < 0.001. 3.6. Shocks exposure and coping strategies The incidence of shocks in Sudan is much higher in rural areas and in poorer households and among agriculture-related activities. In 2014, households were asked about the shocks that severely affected them during the previous five years from a list of preidentified categories of shocks. These include floods/drought, crop disease, livestock loss, severe illness, damage to household dwelling, water shortage, and death of a household member. In rural areas, nearly half of the households have been exposed to at least one shock, compared to 36 percent in rural areas. Households living in the bottom quintiles report a higher prevalence of shocks, but the prevalence is still high among the middle and highest quintiles. Households working on crop farming and animal husbandry have the highest occurrence of shocks, compared to other livelihoods (World Bank 2019a). Households headed by females are more likely to report experiencing shocks compared to male-headed households. Almost half of female-headed households report experiencing a shock. Both female- and male-headed households experience a higher prevalence of shocks in rural areas, poorer consumption quintiles, and among those engaged in an agricultural livelihood. More than half of female-headed households and 49 percent of male-headed households in rural areas are exposed to at least one shock, compared to 42 percent and 35 percent in urban areas, respectively. Except for among the poorest quintile, female-headed households experience a higher prevalence of shocks than their male counterparts (Figure 27). Female- and male-headed households engaged in crop farming and animal husbandry are more likely to be exposed to shocks compared to other livelihoods. More than 60 percent of female- and male- headed households working on crop farming experienced at least one shock, followed by animal husbandry, with 47 percent and 50 percent prevalence among female- and male-headed households, respectively. However, engaging in wage and salaried employment showed the largest gender gap; female 34 heads wage and salaried workers are 28 percent more likely to be exposed to shocks compared to their male heads’ counterparts. This can be linked directly that Sudanese women wages and salaries are in the lower end compared to males. Figure 27: Prevalence of shocks across female- and male-headed households, by area, quintile, and livelihood 70% 60% 50% 40% 30% 20% 10% 0% Rich Other Urban Pension Aid Poorest Richest Crop farming Poor Animal husbandry Total Remittances Transfers from Rural Wages and Owned business Middle Property income household… salaries enterprise Sudan Area Consumption Quintile Livelihood Female headed Male headed Source: Own calculations based on NHBPS 2014/15. The majority of female- and male-headed households are affected by one shock, with agricultural shocks being the most common. There is little difference between the number of shocks reported by female- and male-headed households, with 60 percent reporting being affected by one shock (in the last five years from 2014), and nearly one-quarter affected by two shocks. Livestock loss and flood are the two main types of shocks reported by both households. Other than agriculture-related shocks, female- and male-headed households experience different types of shocks. Female-headed households are more likely to be exposed to health-related shocks, such as severe illness and death of a household head and/or member (which can be the reason the household is headed by a female), whereas male-headed households’ experience of shocks is more likely to be related to illness and dwelling damage (Figure 28 a, b). 35 Figure 28: Number of shocks and specific types of shocks reported: female- and male-headed households a) Number of shocks reported b) Type of shocks reported 1 60% Livestock Loss 60% Flood 25% Severe Illness/Accident 2 24% Death of Hh. Head 10% Death of a Hh. Member 3 11% Crop Disease 3% Dwelling… 4 3% Fire 1% Robbery/burglary/ass… 5 1% Drought More than Severe Water Shortage 0% 5 1% Other Shocks 0% 20% 40% 60% 80% 0% 5% 10% 15% 20% Female headed Male headed Female headed Male headed Source: Own calculations based on NHBPS 2014/15. Robbery and health shocks have the highest impact on female heads. On average, the estimated cost of shocks to female-headed households was highest for robbery (SDG 21,000), the death of a household head (13,000), and livestock loss (8,000). On the other hand, the cost of shocks to male-headed households were highest for severe illness (SDG 12,000) and livestock loss (SDG 11,000) (Figure 29a). Receiving help is the main coping strategy for both female- and male-headed households. Nearly one- quarter of female- and male-headed households affected by shocks in 2014 reported receiving help from different bodies to cope with the shocks: religious groups, local and international nongovernmental organizations, the government, and family and friends. The second most common coping strategy is selling assets, using savings, and renting the farm. The main gender differences observed in reported coping strategies is that female-headed households are more likely to reduce household expenses or consumption, while male-headed households are more likely to borrow (Figure 29b). 36 Figure 29: Effects and coping strategies: female- and male-headed households a) Average effect of shocks in SDG b) Coping strategy Robbery/burglary/assault Received Help 24% Death of Hh. Head Sell Assets/Use… 23% Livestock loss Death of a Hh. Member Other 16% Severe Illness/Accident Reduced Expenses or… 8% Fire Flood No Coping Strategy 8% Dwelling… Increase Labor Supply 7% Drought Crop Disease Borrowed Money 10% Other Shocks Kids Migrate or… 4% Severe Water Shortage - 10,000 20,000 0% 10% 20% 30% Female headed Male headed Female headed Male headed Source: Own calculations based on NHBPS 2014/15. 3.7. Voice and agency 3.7.1. Women’s mobility Women in Sudan are legally restricted in a way that affects their mobility and decision making. According to the Passport and Immigration Act of 1994, women need the written approval of their male guardian to travel. In addition, the Muslim Personal Status Act of 1991, Article 75 (b), states that married women cannot leave the matrimonial home without the permission of their husbands. While women’s representation in parliament has steadily increased from around 10 percent of seats in the early 2000’s to 31 percent today, the overall legal and policy environment is still relatively restrictive. According to the WBL 2020 report, Sudan has scored 0 on the indicators of mobility, workplace and marriage. Sudanese women can’t choose to live the same way as Sudanese men, can’t get a job in the same way can’t be the head of the household the same way as Sudanese men (World Bank 2020). Married women cannot choose where to live nor work outside the home in the same way as married men. Article 51 of the Personal Status Law for Muslims designates men as the family breadwinners and the legislation also notes that women must obey their husbands (Tønnessen and Kjøstvedt 2010). Laws around marriage and divorce are also unequal. Marriage is regulated by the 1991 Personal Status Law for Muslims; the minimum age for marriage is when both parties have reached puberty. Both parties must consent for a marriage to be valid; however, women must seek permission from a male guardian to enter marriage. The rights of women to initiate divorce vary according to different types of laws in Sudan. Under Islamic family law, women are able to file for divorce under specific circumstances, including if the husband fails a financial obligation to support her, if a husband has multiple wives and a woman can prove that her husband does not treat them equally, if the husband has a defect that was not disclosed before marriage, if the husband is impotent, if he is abroad for more than one year, or if the husband is sentenced to prison for more than two years. On the other hand, men are able to divorce wives out-of-court by simply declaring “I divorce you.” Improvements for women’s rights in divorce have been made for Muslim women since the 37 codification of Islamic family law in 1991: women are no longer legally required to return to a marriage under the ‘house of obedience’ principle (Tønnessen and Kjøstvedt 2010). Under customary law, it is typically more difficult to obtain a divorce due to complications surrounding women’s dowry; because a dowry is the property of the wife’s family, losing it in the event of a divorce could have severe economic consequences for the family, the thought of which often prevents women from seeking divorce. For Christian women, divorce regulations can vary according to their denomination, region, and tribe (Tønnessen 2007). According to the labor code, women cannot work from 10 p.m. to 6 a.m. (night shifts) except for some health and social services jobs. Also, women are not able to work in the same industries as men and are not allowed to work in jobs deemed hazardous, arduous, or morally inappropriate16 . 3.7.2. Gender-based violence Sudan is one of the few countries that is not a party to the United Nations Convention on the Elimination of All Forms of Discrimination against Women. The convention provides the basis for realizing equality between women and men through ensuring women's equal access to, inter alia, education, health, and employment. Sudan scored zero out of 100 in protecting women from violence, according to the WBL indicator that considers laws on domestic violence and sexual harassment in education and employment. All of these obstacles prevent Sudanese women from having full economic participation. Sudan’s Criminal Law of 1991 governs consequences for acts of violence against women (Tønnessen 2012). In 2009, the law was amended to include sexual violence17 and in 2015 the Sudan Penal Code amendment included for the first time a new element criminalizing sexual harassment (Article 151). However, domestic violence is not criminalized and existing rape laws do not specifically address spousal rape. While it has previously been reported that women who were determined to have committed moral crimes, such as wearing trousers in public, have been subject to corporal punishment such as flogging, recent legal reforms have repealed a public order law that dictated how women could dress and behave in public18 (SIHA 2017; Sudan Tribune 2013). On February 22, 2015, the Sudan Penal Code of 1991 was amended to expand and clarify the definition of rape that meets international standards (Article 149). The amendment distinctly separates the crime of rape from the crime of zina (sex outside of marriage), thus reducing the risk of women being accused of zina when they report rape. These factors may make it more likely that women who have been raped will report the crime and take the perpetrator to court. However, judges in prosecutions for rape can require the sexual act to have been witnessed by multiple men; in addition, male testimonies are typically more heavily weighted or believed than female testimonies. If a woman is unable to prove that she did not 38 consent, she is vulnerable to being charged with zina because she has confessed to sexual activity outside of marriage. Customary law dictates that if a rape victim’s family agrees, rapists are able to avoid punishment by marrying their victim (Tønnessen and Roald 2007). If a woman fails to prove her case in court, she may be tried for adultery.19 Laws do not cover spousal rape. Sudan has one of the highest rates of FGM prevalence, with nearly 87 percent of women (ages 15–49) having been circumcised (MICS 2014), though the government recently criminalized the practice. The prevalence is higher in rural areas, among women with less education and in the richest quintiles – the latter statistic owing to the fact that FGM is more prevalent in the northern state where household welfare is higher. The prevalence of FGM is lower when the mother has higher education, resides in urban areas, and is from the top quintile (Figure 30). The practice has a dramatic impact on maternal and reproductive health indicators, as it contributes to infections, obstructed labor, and fistulas. UNICEF and the United Nations Population Fund found that 87 percent of girls and women between 15 and 49 years had undergone FGM, but the prevalence varied depending on geography (United States Department of State 2015). The 2014 MICS data show that this drops to 32 percent for daughters ages 0–14 years, suggesting a decline in the practice. While 41 percent of women ages 15–49 years still state that FGM should be continued, there are indications that support for FGM is diminishing: for girls and women between 15 and 19 years, 37 percent supported FGM compared to 73 percent in 2006 (United States Department of State 2015). The government has responded to these issues in several ways. The Ministry of Welfare and Social Security established a Unit for Combating Violence against Women and a Woman Center for Human Rights. Further, the National Council for Child Welfare has created campaigns against FGM, which has gained wide international support (Ali Siddig and Hassan 2016). In May 2020, the government criminalized the practice of FGM and made it punishable with three years imprisonment and a fine.20 Figure 30: FGM across women and daughters by education, quintile, and area a) FGM among daughters by mothers b) FGM prevelnace among women and history 50% daughters by women/mother education Mother not circumcised Mother circumcised 0% None Primary Secondary Higher 0% 20% 40% 60% 80% 100% Women Daughter Daughter circumcised Daughter not circumcised 39 c) FGM prevelnace among women and d) FGM prevelnace among women and daughters by quintile 100% daughters by area 0% 0% Poorest Second Middle Fourth Richest Urban Rural Women Daughter Women Daughter Source: Own calculations based on MICS 2014. More than one-third of women ages 15–49 justified husbands beating their wives for at least one of five possible reasons, according to MICS 2014. These five reasons reflect gender norms, including around expected behaviors related to taking care of the children, making food, and leaving the house without permission. Women from rural areas are by far the most likely to justify domestic violence, though acceptance of wife beating is also somewhat higher than average among those in the lowest wealth quintiles and those with less education (Figure 31). Figure 31: Attitudes toward domestic violence among women a) Women's attitude toward b) Women agreeing with husband domestic domestic violence violence by area, quintile, and education Diagree with 80% 60% husband 40% 34% beating his wife 20% 0% Poorest Urban Fourth Rural Richest Primary Secondary Higher Middle None Second 66% Agree with husband beating his wife Area Quintile Education c) Various circumstances justifying domestic violence If she neglects the children If she goes out with out telling husband If she argues with husband If she refuses sex with husband If she burns the food 0% 5% 10% 15% 20% 25% 30% Source: Own calculations based on MICS 2014. 3.7.3. Voting and women’s rights The majority of Sudanese men and women are reticent about women’s rights in many areas. The Arab Barometer fifth wave was conducted in 2018/19, which collected public opinion data on different issues, including women’s rights and their roles. In Sudan, public majorities believe that men are better political leaders, with low acceptance from the male side on women being head of state, but there was high acceptance for women’s quota in political offices by both genders. In the private sphere, a high proportion of men and women believe that the husband should have the final say in family decisions (80 percent males, 67 percent females). Very few Sudanese accept women entitled to an equal share of inheritance (Figure 32). However, more women than men believe that they should have equal rights on divorce (37 40 percent males, 54 percent females) and that they should be allowed to travel alone (25 percent males, 50 percent women). In addition, both genders largely support equal education access, and only 35 percent males and 23 percent females think university education is more important for males. Furthermore, more Sudanese males than females vote in elections. Nearly 60 percent and 52 percent of Sudanese males voted in the last parliamentary and local elections in Sudan, respectively, compared to 44 percent and 39 percent for females (Figure 33), respectively. Figure 32: Sudanese perception on women’s rights: male versus female Men are better at political leadership Acceptance of women's quota Husband have the final say in family decisions Woman can be head of state Equal right on divorce decision Permission for woman to travel alone University education is more important for males Equal inheritance 0% 20% 40% 60% 80% Male Female Source: Arab Barometer 2018/19. Note: The percentages show those who strongly agree or agree with the statements above. Figure 33: Gender differences in voting 60% 50% 40% 30% 20% 10% 0% Parliamentary Elections Local Elections Male Female Source: Arab Barometer 2018/19. 41 4. POTENTIAL IMPACT OF COVID-19 ON GENDER INEQUALITY IN SUDAN The coronavirus disease 2019 (COVID-19) pandemic is still in the early stages of its spread across Sub- Saharan Africa compared to most other regions in the world. However, as of June 16, 2020, there were already 7,435 reported cases (468 deaths) in Sudan and the capital has already been placed on a lockdown. Given the high levels of poverty, low access to basic health and sanitation infrastructure, and high insecurity, limiting the spread of the virus is likely to prove even more difficult than it has been across other countries. While impacts from the pandemic will be severe for the population as a whole, there are likely to be differential and greater impacts on women and girls, including across health, education, and economic outcomes. First, women may be more exposed to the virus because of their vulnerability as frontline health workers, their exposure to sick patients during facility-assisted childbirths, their disproportionate role of caring for sick household members, and their low agency. In terms of direct impacts, women’s likely greater role of caring for sick family members and their higher representation among health care workers mean that they are likely to be more exposed to COVID-19. Boniol et al. (2009) find that across all of Africa, women represent 65 percent of nurses. Evidence indicates, for example, that during the Ebola outbreak of 2014– 2016, women’s greater role as household caregivers and frontline health care workers led to higher infections rates (Davies and Bennett 2016). There is also early evidence of the disproportionate impact on female health care workers during the current COVID-19 crisis. Spain, for example, has recorded more than twice as many infections among female workers than male health care workers.21 Women are also exposed to specific risks as health care patients, with antenatal care, postnatal care, and delivery of babies potentially bringing them into close contact with COVID-19 patients. Finally, women’s low agency may also delay or restrict their access to health care if they contract COVID-19. While we do not have data for Sudan on who makes decisions regarding women’s own health care, Sudan’s performance across a range of other indicators (see section 3.7) suggests that women’s overall agency is low. In terms of access to basic services, the negative impacts of school closures or reduced funding for non- COVID health services will likely disproportionately affect women, especially adolescent girls, given this is a time when young women make key decisions on their education, and aspirations for the future.22 The closure of schools may be the end of schooling for some girls. During the 2014–2016 Ebola outbreak, for example, many girls in Sierra Leone did not return to school once the education system reopened (Bandiera et al., 2019). The pandemic may expose women to higher GBV risks, both inside and outside the household. A recent study from the Center for Global Development identifies nine channels through which pandemics may contribute to an increase in GBV: “(1) economic insecurity and poverty-related stress, (2) quarantines and social isolation, (3) disaster and conflict-related unrest and instability, (4) exposure to exploitative relationships due to changing demographics, (5) reduced health service availability and access to first 42 responders, (6) inability of women to temporarily escape abusive partners, (7) virus-specific sources of violence, (8) exposure to violence and coercion in response efforts, and (9) violence perpetrated against health care workers” (Peterman et al. 2020). These aspects of GBV risk are likely to be even more severe in an FCV context such as Sudan where existing levels of conflict, violence, and poverty are high and women’s agency is low. In terms of economic opportunities, there are likely to be disproportionate impacts on women. We do not have any data on Sudan yet, but a survey of four countries in the Asia and the Pacific region (Bangladesh, Pakistan, Maldives, and the Philippines) indicates that a higher percentage of women than men have already seen decreases in their incomes across a range of sources including family businesses, farming or fishing, paid work, and remittances.23 Given Sudanese women’s lower incomes, lower agricultural yields/higher food insecurity, and lower ability to come up with emergency funds, they may be less able than men to cope with additional shocks brought about by COVID-19. Moreover, women may be less able to benefit from any social protection measures aimed at smoothing the negative impacts of the virus. For example, we know that women-owned firms in Sudan are less likely to be registered than their male counterparts, which may reduce their access to basic formal social protection. A higher proportion of women (65 percent) than men (45 percent) are also estimated to be in vulnerable employment (includes contributing family workers and own-account workers), potentially putting them outside the reach of some formal social protection programs. Given women’s greater share of childcare tasks as well as a greater share of the responsibility of caring for sick household members, the pandemic is likely to further reduce women’s ability to engage in paid work. Finally, recent research focused on the United States suggests that the social distancing measures related to COVID-19 will have large impacts on employment in the areas of the service sector, such as restaurants and travel and tourism, where employment tends to be dominated by women.24 This may be less of an issue in Sudan, where women’s level of labor force participation is very low right across the economy, though it will be important to monitor the differential impact of employment losses on women and men. 43 5. CONCLUSION AND POLICY OPTIONS The analysis presented in this report points to a range of possible policy responses that could be used to close gender gaps across a wide variety of key outcomes. The following paragraphs summarize a few of these options; however, given the broad scope of the analysis presented here, it is not intended as an exhaustive list. Targeting the most vulnerable and helping women cope with shocks Given the large regional and wealth-related disparities in the size of gender gaps in Sudan, policies across all of the areas discussed should target the poorest households and female-headed households in lagging rural areas, especially those headed by divorced women. This means that social protection programs could have a key role to play in delivering interventions across multiple constraints faced by women and girls as well as in developing targeting methodologies and tools, such as national registries, that can be used by projects across all sectors to ensure that they reach those most in need. Social protection interventions will be especially important for helping women manage the specific shocks they face and reduce the likelihood of households resorting to harmful coping strategies, such as reducing consumption and pulling children out of school. In some cases, such as conditional cash transfers, evidence suggests that impacts will likely be larger for girls/women and so help close gender gaps. However, in other cases it will be important to include specific consideration for women and girls in the design of programs to account for the different constraints and shocks that women and girls face. For example, to ensure that women are able to benefit from public works opportunities, interventions can consider providing childcare options, allowing greater flexibility in working hours and the types of tasks allocated to women, or exempting women with childcare responsibilities from labor requirements and giving straight cash instead. Given the wide range of severe constraints faced by the most vulnerable women and girls, social protection interventions may also consider using a graduation approach to simultaneously address multiple constraints. This graduation approach has been taken by the Supporting Women’s Livelihoods component of the Girls' Education and Women's Empowerment and Livelihood Project (GEWEL) in Zambia. This project is supporting the poorest women with a livelihood package that includes a life and business skills training, a productive grant of US$225, and savings groups and coaching support to start/expand a livelihoods activity. Policies to address gender gaps and constraints in education and skills To improve outcomes and close gender gaps in the education sector, interventions that address households’ financial constraints are a promising option as evidence suggests they can improve household investments in the least favored children, such as girls. Additionally, as fees are the major constraints for both girls and boys, interventions that address household financial constraints can play a role in raising access to education overall. Evidence from across the Sub-Saharan Africa region suggests that conditional cash transfer programs are particularly effective at improving school enrollment for the least favored children in a household, including girls (Akresh et al. 2013), and that they can have knock-on effects on other related outcomes, such as early childbearing and marriage (Baird et al. 2010). In this regard, other programs that have a conditionality element (such as school feeding programs) may be a promising option for policy makers looking to target girls’ schooling. Given the large number of girls who have already dropped out of school and are particularly at risk of entering a vicious cycle of early marriage, childbearing, and economic inactivity, education and skills 44 programs that can also reach these vulnerable girls could be especially valuable. Programs that deliver vocational and life skills training through girls’ clubs that act as safe spaces have been found to be effective in addressing early childbearing and marriage and can be designed to reach those who are still in school as well as those who have already dropped out. Such a model has been used by BRAC in Uganda, where it had promising impacts on sexual and reproductive outcomes as well as economic empowerment outcomes. The program called ELA used girl-only clubs which both to deliver the training and as a safe space for adolescent girls to socialize. ELA training was held outside of regular school hours, to allow both girls still in school and those who dropped out to attend. Overall, girls in the ELA program were 26 percent less likely to have a child, 58 percent less likely to be married or cohabiting, 25 percent more likely to report always using a condom during sexual intercourse, 44 percent less likely to have had sex against their will over the previous 12 months, 72 percent more likely to be engaged in income-generating activities, and reported self-employment earnings three times higher, compared to the original average (Bandiera et al. 2018). In South Sudan, ELA established 100 community-level girls’ clubs in four states of South Sudan targeting girls ages 15 to 24. ELA had positive impacts on a range of labor market and financial outcomes for girls who were not affected by the conflict (Buehren et al, 2017). The program increased the probability of being engaged in income generation for adolescents in treatment areas that were not exposed to the conflict by almost 10 percentage points which was largely driven by nonfarm self- employment. The program also significantly increased the likelihood that girls (not affected by the conflict) had any savings and controlled some cash on their own. The impact of the program on girls’ social empowerment and the control over their own bodies, however, is ambiguous. Attention should also be paid to better understanding the factors around boys’ secondary dropout among wealthier quintiles and in urban areas. Ensuring that youth have sufficient economic opportunities will be an important aspect of capturing any demographic dividend and building long-term stability and peace in the country. Policies to reduce gender gaps in economic opportunities Policies to reduce gender gaps in economic opportunities could focus on building on recent legal reforms to improve women’s access to assets, supporting programs that target social norms to increase the chance that legal changes are reflected in changes in attitudes and behaviors on the ground, and designing programs that support entrepreneurship and agriculture keeping in mind the specific constraints women face. There appears to be some appetite in government for improving women’s status under the law, considering the recent reforms around banning FGM and repealing the public order law. Building on these efforts to identify and enact legal reforms that specifically address women’s access to assets will be important to improve women’s economic opportunities. Such reforms are especially urgent in Sudan given the finding of WBL that Sudan continues to be one of the countries with the most differences in legal status of women and men, with only Yemen and the West Bank and Gaza achieving a lower score on the 2020 edition of the WBL Index. Reforms to equalize statutory law around inheritance and divorce should be considered. Such efforts would make it easier for women to have secure ownership over assets (rather than ownership through a husband or other male relative) and would be especially important given the high rate of poverty among divorced women. Reforms to divorce law, for example, could include those that provide for a more equitable division of property in the case of marriage dissolution, including based on nonmonetary contributions. 45 Given the influence of customary law and practices, it will be necessary to go beyond statutory legal reforms with interventions that can influence pervasive social norms. For example, in terms of women’s secure land tenure, it would be important to work closely with the local elders and chiefs who have the authority to handle land disputes. It would also be important to work with husbands, so that they are more likely to be allies in their wives’ economic empowerment. Here we have evidence that small financial and informational nudges can encourage men, even in contexts where relatively conservative social norms prevail, to include their wives on land titles (see below). Policies focused on improving women’s secure land tenure could be important not only for women’s agricultural productivity by increasing incentives to make productive investments in land (for example, Goldstein and Udry 2008) but also for their ability to switch into nonagricultural activities (as they no longer need to guard their land as closely) and for their financial inclusion, given that land is an important source of collateral for loans (for example, Besley and Ghatak 2010). A body of evidence from several countries across the region suggests that relatively low-cost interventions to formalize customary land rights, while paying specific attention to women’s rights, can be effective. Land formalization programs have been found to offer particular benefits to widows, who are often victims of land grabbing when their spouse dies. In Benin, for example, widows who benefited from the PFR process to document and formalize customary land rights were 13 percentage points more likely to remain in their home village compared to a control group of widows. Overall, the program also closed the gender gap in fallowing, an important soil fertility investment (Goldstein et al. 2014). However, such programs need to be designed carefully to ensure that they do not further entrench existing inequalities in land tenure. In Rwanda, for example, a land tenure regularization program enforced compulsory joint registration for all married couples, with larger positive impacts on women’s land investments and a shift toward off-farm activities (Ali et al. 2014). The impact evaluation of the initial pilot highlighted the dangers of excluding some women from the benefits—by requiring a marriage certificate, the pilot unintentionally weakened the tenure security of women who were not in formal marriages. This weakness was subsequently corrected before the program was scaled up nationally. Another way to ensure that women are able to get more secure land tenure is to provide small nudges to influence customary practices and encourage husbands to include their wives’ names on land titles. In Uganda, relatively low-cost information and conditional incentives increased demand for co-titling of spouses without dampening overall demand for land titles. The CEDP offered fully subsidized freehold land titles to households as well as (a) provided information about the benefits of joint titling or (b) made the titles conditional on including the wife’s name on the title. Both of these options increased demand for co-titling, though adding a condition is particularly effective, with the probability of co-titling increasing by 50 percent under the condition, relative to a 25 percent increase with gender information (Cherchi et al. 2019). Policies to promote women’s entrepreneurship could include those focused on developing their skills. However, emerging evidence from the region suggests that who is targeted and what skills are taught are two critical parameters that existing interventions often get wrong. As women across the continent often go into self-employment out of necessity rather than choice, maximizing the impact of training programs may depend to some extent on targeting those women who are growth oriented and share more of the characteristics of opportunity rather than necessity entrepreneurs. One approach that has had some success in targeting higher potential entrepreneurs is the use of business plan competitions (McKenzie 2015). Additionally, evidence from Tanzania suggests that targeting more experienced entrepreneurs may be important, with a business incubator showing a larger positive impact on the revenues of more 46 experienced women entrepreneurs compared to those with fewer years of experience (Bardasi et al. 2017). In terms of the types of skills that training programs should focus on, emerging evidence suggests that noncognitive skills could be more important than traditional management skills. In Togo, for example, an impact evaluation found that a training that focused on developing an entrepreneurial mindset (that is, noncognitive skills) was more effective than a traditional business training in increasing women entrepreneurs’ performance. With a grounding in psychology, the training encouraged small business owners to anticipate problems and plan ways to overcome them, emphasizing skills such as personal initiative. Entrepreneurs who took the training experienced a 30 percent increase in their profits, with a 40 percent increase among women entrepreneurs—this compares to a statistically insignificant 5 percent increase for those taking a traditional business training. In addition to the boosted profits, women who received the enhanced training also introduced more new products in their businesses and increased investment (Campos et al. 2017). Increasing women’s access to business capital means not only increasing the amount of credit available to them but also ensuring that they are able to exert control over the finance they access. The fact that women’s borrowing appears to be skewed more than men’s toward borrowing for human capital investment versus business investment may indicate that women face greater social pressure to invest in household human capital, and therefore, it is important to have policies that allow self-employed women to achieve more balance between household and business priorities. This may include secure savings products that allow women to more easily separate business and household finances. For example, in Kenya, providing savings accounts to women market vendors led to a more than 45 percent increase in business investment, while there was no impact of providing such accounts to male motorbike drivers (Dupas and Robinson 2013). Another promising approach to increasing women’s financial inclusion is the use of alternatives to traditional collateral. In Ethiopia, for example, the World Bank is piloting the use of psychometric testing to replace the need for land as business loan collateral. The psychometric tests have been developed to predict the likelihood that a loan applicant will be able to repay a loan. Early results show that loan customers who scored at a high threshold on the psychometric test are seven times more likely to repay compared to lower performing customers (Alibhai et al. 2018). The use of psychometric tests is significant because it gives lenders another source of information on women’s loan applicants and so may reduce any perceived higher risk of lending to women. Encouraging women, both as entrepreneurs and as wage workers, to enter new sectors may be particularly important given Sudanese women’s current concentration in few sectors, such as education and domestic work for private households. On this issue, evidence from across the region on women who cross over into traditionally male-dominated sectors may be relevant. The evidence from Ethiopia and Uganda suggests that women entrepreneurs who operate in male-dominated sectors earn the same as their male counterparts and much more than women who stick to traditionally female sectors. Factors that appear to be associated with women operating businesses in nontraditional sectors include access to information on differential profits across sectors and exposure to male-dominated sectors through professional experience or role models, especially male family members (Alibhai et al. 2017; Campos et al. 2015). However, it would be important that any such policies also paid attention to supporting women who do cross over into male-dominated sectors to help them overcome the various risks and constraints they are likely to face once they are working in the sector, such as increased risks of GBV. In the agriculture sector, large differences in agricultural productivity and incomes between female- headed and male-headed households suggest that interventions to address underlying constraints to 47 women in agriculture could have large welfare impacts. Interventions could focus on the deficits highlighted in this paper, including female-headed households access to productive inputs (pesticides) and labor.25 Interventions to address these constraints could include providing women with financing to hire farm labor, tasking extension agents with helping women find farm labor, supporting community- based childcare so that women have more time to engage in and supervise farm labor, and providing women with financing or price discounts to improve their access to nonlabor inputs such as pesticides and fertilizers. For example, in Zambia, Seidenfeld, Handa, and Tembo (2013) find that giving cash transfers to households with children under five years old in Zambia increased their spending on hired labor by four times. We also have strong evidence on the positive impacts of interventions to improve women’s land tenure security, including land formalization and incentives to promote co-titling of husbands and wives (see above). Finally, women’s overall economic participation across all sectors could be supported by a focus on social norms. Given that the majority of women (85 percent) who are not participating in the labor market cite being a homemaker/housewife as the main reason, there is a need to address social norms around women’s and men’s roles. There is emerging evidence that couples’ gender norms discussion groups can shift behaviors and attitudes. In Rwanda, for example, the Bandebereho couples’ intervention engaged men and their partners in small, participatory discussion groups that addressed issues related to gender and power relations, fatherhood, couples’ communication and decision making, domestic violence, caregiving, child development, and male engagement in reproductive and maternal health. An impact evaluation of the intervention found that it significantly increased men’s involvement in domestic and childcare tasks and led to greater shared decision making among couples (Doyle et al. 2018). There is also evidence that such interventions can change men’s attitudes to their wives’ economic activities. In Uganda, Blattman et al. (2013) find that including male partners in discussions on gender relations in a cash transfer and business skills training intervention aimed at women improved couples’ communications and led to a small increase in men’s support for their partners’ businesses. Policies to support women’s voice and agency Given women’s low status in Sudan, as evidenced by legal inequalities, acceptance of GBV, FGM, and attitudes to women’s role in household and public decision making, efforts to support women’s voice and agency are critical and can support progress across all other areas. Such efforts could focus on shifting restrictive social norms, which would increase the likelihood that recent high-level policy reforms regarding women’s rights (such as the banning of FGM and the repeal of the public order law) would be matched by changes in practices on the ground. Social norms interventions could be used to target a wide range of attitudes and behaviors, including those related to the division of decision making and of household tasks between husbands and wives and women’s ability to work outside the house as well as practices related to inheritance and asset ownership, GBV, and human capital investments, including timing of marriage and childbearing. Potential mechanisms for influencing these behaviors and attitudes include working with authoritative institutions to build support for new norms, such as by mandating that land title certificates include names of both the husband and wife; using mass media and edutainment to convey messages, such as through soap operas (La Ferrara, Chong, and Duryea 2012); and using discussion 48 groups and community dialogue (for example, Doyle et al. 2018). In cases where social norms are too persistent to be tackled directly, interventions could focus instead on appealing to existing norms that can help achieve a desired outcome rather than trying to change the problematic norm directly, such as by appealing to health concerns to address early childbearing. Emerging evidence from across the region suggests that programs targeting gender-related social norms by engaging men or couples may be a promising channel for addressing issues around women’s agency, including GBV and the sharing of domestic tasks. In Rwanda, for example, the Bandebereho couples’ intervention engaged men and their partners in small, participatory discussion groups that addressed issues related to gender and power relations, fatherhood, couples’ communication and decision making, domestic violence, caregiving, child development, and male engagement in reproductive and maternal health. An impact evaluation of the intervention found that it significantly reduced women’s experience of physical and sexual violence; increased the use of contraceptives, skilled birth attendance, and men’s involvement in domestic and childcare tasks; and led to greater shared decision making among couples (Doyle et al. 2018). Such efforts to engage men may be particularly important in activities aimed at empowering women, to ensure that such work does not lead to an increase in intra-household tensions and violence if men view it as an attempt to undermine their traditional role. In Burundi, Iyengar and Ferrari (2011) find that adding a gender norms discussion series for couples to a Village Saving and Loan Associations intervention led to improved attitudes toward GBV, higher female participation in household decision making, and improved negotiation skills between men and women partners. Similarly, in Uganda, Blattman et al. (2013) find that including male partners in discussions on gender relations in a cash transfer and business skills training intervention aimed at women improved couples’ communications and led to a small increase in men’s support for their partners’ businesses. 49 REFERENCES Akresh, R., De Walque, D., and Kazianga, H. 2013. “Cash Transfers and Child Schooling: Evidence from a Randomized Evaluation of the Role of Conditionality”, World Bank Policy Research Working Paper: Impact Evaluation Series #82, World Bank, Washington, D.C. Ali, D.A., Deininger, K. and Goldstein, M., 2014. Environmental and gender impacts of land tenure regularization in Africa: Pilot evidence from Rwanda. Journal of Development Economics, 110, pp.262-275. Alibhai, Salman; Buehren, Niklas; Papineni, Sreelakshmi; Pierotti, Rachael Susan. 2017. Crossovers: female entrepreneurs who enter male sectors -- evidence from Ethiopia (English). Policy Research working paper; no. WPS 8065. Washington, D.C. : World Bank Group. Alibhai, Salman; Buehren, Niklas; Coleman, Rachel; Goldstein, Markus; and Strobbe, Francesco. 2018. Disruptive Finance: Using Psychometrics to Overcome Collateral Constraints in Ethiopia. World Bank: Washington, DC. https://openknowledge.worldbank.org/handle/10986/29746. Ali Siddig, El-Fatih, and Fareed M. A. Hassan. 2016. The Sudan Interim Poverty Reduction Strategy Paper Status Report: A joint World Bank Group and Sudan’s Ministry of Finance and Economic Planning Assessment (English). Washington, DC: World Bank Group. African Development Bank. 2015. Empowering African Women: An Agenda for Action. Africa Gender Equality Index 2015. Azargoshasb, Azita. 2016. Supporting Women's Empowerment and Gender Equality in Fragile States: Guinea, Lebanon and Sudan - Research Series Summary. UN Women. Baird, Sarah, Ephraim Chirwa, Craig McIntosh, and Berk Ozler. (2010). “The short-term impacts of a schooling conditional cash transfer program on the sexual behavior of young women,” Health Economics 19(Suppl. 1): 55–68. Bandiera, O., N. Buehren, M. P. Goldstein, I. Rasul, and A. Smurra. 2019. The Economic Lives of Young Women in the Time of Ebola: Lessons from an Empowerment Program. World Bank. Bandiera, O., Buehren, N., Burgess, R., Goldstein, M., Gulesci, S., Rasul, I., & Sulaiman, M. (2018). Women’s empowerment in action: evidence from a randomized control trial in Africa. World Bank. Besley, T. and Ghatak, M., 2010. Property rights and economic development. In Handbook of development economics (Vol. 5, pp. 4525-4595). Elsevier.Blattman, C., Green, E., Annan, J. and Jamison, J., 2013. Building women's economic and social empowerment through enterprise: an experimental assessment of the women's income generating support program in Uganda. Boniol, M., M. McIsaac, L. Xu, T. Wuliji, K. Diallo, and J. Campbell. 2019. “Gender Equity in the Health Workforce: Analysis of 104 Countries.” Working Paper 1, World Health Organization, Geneva. Boudet, Muñoz, Ana María, Patti Petesch, Carolyn Turk, and Angélica Thumala. 2013. “On Norms and Agency: Conversations About Gender Equality with Women and Men In 20 Countries.” Washington, DC: World Bank. 50 Buehren, N., Chakravarty, S., Goldstein, M., Slavchevska, V., & Sulaiman, M. (2017). Adolescent Girls’ Empowerment in Conflict-Affected Settings: Experimental Evidence from South Sudan. Campos, F., Goldstein, M., McGorman, L., Boudet, A.M.M. and Pimhidzai, O., 2015. Breaking the metal ceiling: female entrepreneurs who succeed in male-dominated sectors. The World Bank. Campos, F., Frese, M., Goldstein, M., Iacovone, L., Johnson, H.C., McKenzie, D. and Mensmann, M., 2017. Teaching personal initiative beats traditional training in boosting small business in West Africa. Science, 357(6357), pp.1287-1290. CBS (Central Bureau of Statistics) and UNICEF (United Nations Children’s Fund). 2016. Multiple Indicator Cluster Survey 2014 of Sudan, Final Report. Khartoum, Sudan. Cherchi, Ludovica, Markus Goldstein, James Habyarimana, Joao Montalvao, Michael O’Sullivan, Chris Udry, and Ariel Gruver. 2019. “Empowering Women through Equal Land Rights: Experimental Evidence from Rural Uganda.” Gender Innovation Lab Policy Brief 33, World Bank, Washington, DC. Davies, S. E., and B. Bennett. 2016. “A Gendered Human Rights Analysis of Ebola and Zika: Locating Gender in Global Health Emergencies.” Int Aff 92:1041–60. Dinkelman, T. 2011. “The Effects of Rural Electrification on Employment: New Evidence from South Africa.” American Economic Review 101 (7): 3078–3108. doi:10.1257/aer.101.7.3078. Dupas, P. and Robinson, J., 2013. Savings constraints and microenterprise development: Evidence from a field experiment in Kenya. American Economic Journal: Applied Economics, 5(1), pp.163-92. Doyle, K., Levtov, R.G., Barker, G., Bastian, G.G., Bingenheimer, J.B., Kazimbaya, S., Nzabonimpa, A., Pulerwitz, J., Sayinzoga, F., Sharma, V. and Shattuck, D., 2018. Gender-transformative Bandebereho couples’ intervention to promote male engagement in reproductive and maternal health and violence prevention in Rwanda: findings from a randomized controlled trial. PLoS One, 13(4).FAO (Food and Agriculture Organization of the United Nations), IFAD (International Fund for Agricultural Development), UNICEF (United Nations Children’s Fund), WFP (World Food Programme), and WHO (World Health Organization). 2019. The State of Food Security and Nutrition in the World 2019: Safeguarding against Economic Slowdowns and Downturns. Rome, FAO. La Ferrara, E., Chong, A. and Duryea, S., 2012. Soap operas and fertility: Evidence from Brazil. American Economic Journal: Applied Economics, 4(4), pp.1-31. Goldstein, M. and Udry, C., 2008. The profits of power: Land rights and agricultural investment in Ghana. Journal of political Economy, 116(6), pp.981-1022. Housing Rules. 2014. Land Department. http://khrland.gov.sd/housing-rules.php. Government of Sudan. Iyengar, R. and Ferrari, G., 2011. Discussion sessions coupled with microfinancing may enhance the role of women in household decision-making in Burundi. Cambridge, MA: National Bureau of Economic Research. Munoz Boudet et al .2018. Gender Differences in Poverty and Household Composition through the Lifecycle: A Global Perspective 51 Peterman, A., A. Potts, M. O’Donnell, K. Thompson, N. Shah, S. Oertelt-Prigione, and N. van Gelder. 2020. “Pandemics and Violence against Women and Children.” Center for Global Development Working Paper. Seidenfeld, D., S. Handa, and G. Tembo. 2013. Social Cash Transfer Scheme: 24-Month Impact Report for the Child Grant Programme. American Institutes for Research. SIHA (Strategic Initiative for Women in the Horn of Africa). 2009. Beyond Trousers: The Public Order Regime and the Human Rights of Women and Girls in Sudan. SIHA (Strategic Initiative for Women in the Horn of Africa). 2015. Third-Class Citizens: Women and Citizenship in Sudan - A Paper on Women’s Struggle for Equal Citizenship in Sudan. Kampala, Uganda: SIHA Network. SIHA. 2017. The Criminalization of Women in Sudan. SIHA. 2018. A Grim Portrait of Sexual Violence in Conflict Zones in The Horn of Africa (Sudan, South Sudan, and Somalia). SIHA. 2019. SIHA Network Gender Briefing-Sudan. Sochas, L., A. A. Channon, and S. Nam. 2017. “Counting Indirect Crisis-related Deaths in the Context of a Low-resilience Health System: The Case of Maternal and Neonatal Health during the Ebola Epidemic in Sierra Leone.” Health Policy Plan 32 (suppl 3): iii32–39. Sudan Tribune. 2013. Calls for Sudan to End Flogging of Women. https://www.sudantribune.com/spip.php?article48751 Thomas, Kathrin. 2019. Women’s Rights in Middle East and North Africa - Arab Barometer Wave V. Princeton University. Tønnessen, L. 2007. “Gendered Citizenship in Sudan: Competing Perceptions of Women’s Civil Rights within the Family Laws among Northern and Southern Elites in Khartoum.” CMI Working Paper. http://bora.cmi.no/dspace/bitstream/10202/27/1/Working%20paper%20WP%202007-4.pdf. Tønnessen, L. 2012. From Impunity to Prosecution? Sexual Violence in Sudan beyond Darfur. Norwegian Peacebuilding Resource Centre (NOREF) Report. Tønnessen, L., and A. Roald. 2007. CMI Report: Discrimination in the Name of Religious Freedom: The Rights of Women and Non-Muslims after the Comprehensive Peace Agreement in Sudan. http://www.cmi.no/publications/file/2704-discrimination-in-the-name-of-religious- freedom.pdf. Tønnessen, L., and Hilde Granås Kjøstvedt. 2010. The Politics of Women’s Representation in Sudan: Debating Women’s Rights in Islam from the Elites to the Grassroots. Chr Michelsen Institute (CMI). United States Department of State. 2018. Country Reports on Human Rights Practices for 2015. https://www.state.gov/reports/2018-country-reports-on-human-rights-practices/sudan/. 52 United Nations Development Programme (UNDP). 2012. Somalia Human Development Report 2012: Empowering youth for peace and development. UNDP Somalia. Wodon, Q. (2016). Early childhood development in the context of the family: the case of child marriage. Journal of Human Development and Capabilities, 17(4), 590-598. World Bank. 2014. Enterprise Survey: Sudan Country Profile 2014. International Finance Cooperation, World Bank Group. World Bank. 2018a. Education, Health and Poverty in Sudan. Washington, DC: World Bank. World Bank. 2018b. Kenya Poverty and Gender Assessment 2015/16 - A Decade of Progress and the Challenges Ahead. Washington, DC: World Bank. World Bank. 2019a. Shocks and Household Welfare in Sudan. Washington, DC: World Bank. World Bank. 2019b. Agricultural Productivity and Poverty in Rural Sudan. Washington, DC: World Bank. World Bank. 2019c. Doing Business 2019: Training for Reform-Sudan Economy Profile. Washington, DC: World Bank. World Bank. 2019d. Poverty and Inequality in Sudan: 2009–2014. Washington, DC: World Bank. World Bank. 2019e. The Little Data Book on Gender. Washington, DC: World Bank. World Bank Group. 2018. Women, Business and the Law 2018. Washington, DC: World Bank. World Bank. 2020. Women, Business and the Law 2020. Women, Business and the Law. Washington, DC: World Bank. doi:10.1596/978-1-4648-1532-4. License: Creative Commons Attribution CC BY 3.0 IGO 53 APPENDIX Table A.1: Primary and secondary GPI by states (%) Primary Secondary River Nile 104 98 South Darfor 103 92 North Darfor 102 121 Khartoum 102 103 Blue Nile 101 154 Northern 100 158 White Nile 99 128 Gezira 99 101 Red Sea 97 115 South Kordofan 96 140 North Kordofan 95 79 Sinnar 95 159 Gadarif 93 122 West Darfor 92 63 Central Darfor 89 81 Kassala 89 102 East Darfor 86 128 West Kordofan 84 72 Source: MICS 2014/15. Table A.2: Determinants of labor force participation by gender, probit estimations (coefficients) 2009 2014 Male Female Male Female 25–34 0.972*** 0.294*** 0.868*** 0.220*** (0.050) (0.045) (0.050) (0.038) 35–44 1.103*** 0.332*** 0.959*** 0.334*** (0.090) (0.051) (0.077) (0.042) 45–54 0.611*** 0.278*** 0.560*** 0.320*** (0.103) (0.062) (0.087) (0.048) 55–64 −0.311** 0.0265 −0.260** 0.0383 (0.098) (0.079) (0.082) (0.063) Rural −0.0167 0.173*** 0.134*** 0.276*** (0.041) (0.037) (0.038) (0.030) (own) Some/completed primary −0.568*** −0.433*** −0.435*** −0.223*** (0.050) (0.049) (0.048) (0.038) (own) Secondary −0.894*** −0.229*** −0.371*** −0.0287 (0.058) (0.057) (0.061) (0.050) (own) Post-secondary and above −0.490*** 0.644*** 0.000323 0.794*** (0.091) (0.077) (0.081) (0.058) (head) Some/completed primary 0.054 −0.159*** −0.0536 −0.0438 (0.050) (0.043) (0.048) (0.034) (head) Secondary 0.106 −0.108 −0.143* −0.182*** (0.071) (0.060) (0.069) (0.055) (head) Post-secondary and above −0.214* −0.0903 −0.293*** −0.0974 (0.090) (0.081) (0.083) (0.061) Married 0.888*** −0.128** 0.862*** −0.240*** (0.075) (0.047) (0.077) (0.037) Widowed −0.274 0.0247 0.365 −0.263*** 54 2009 2014 Male Female Male Female (0.209) (0.093) (0.204) (0.075) Divorced 0.202 0.310*** 0.105 0.173* (0.250) (0.089) (0.190) (0.076) Have children = 1 0.622*** 1.263*** 0.400*** 0.871*** (0.088) (0.073) (0.079) (0.053) Poor = 1 −0.0713 0.235*** −0.0801* 0.292*** (0.040) (0.032) (0.038) (0.028) Constant 0.312*** −1.005*** 0.335*** −0.759*** (0.055) (0.053) (0.051) (0.044) Observations 11,845 12,245 14,379 1,5606 Pseudo R-squared 0.368 0.109 0.281 0.075 Note: Standard errors in parentheses *p < 0.05, **p < 0.01, ***p < 0.001. Table A.3: Earnings summary by industry, gender, and place of residence (SDG per month) Urban Rural Total Employment sector Male Female Male Female Male Female Agriculture Mean 1,305 562 960 501 989 505 Median 900 450 700 400 700 400 10th percentile 300 150 210 100 240 100 90th percentile 2,000 1,000 1,650 1000 1,800 1,000 Manufacturing/construction Mean 1,459 851 1,376 716 1,419 825 Median 1,000 700 900 500 1,000 700 10th percentile 300 320 350 88 320 120 90th percentile 2,800 1,300 2,100 1,200 2,400 1,300 Services Mean 1,587 925 1,123 724 1,358 857 Median 1,000 760 900 600 900 700 10th percentile 400 250 380 180 400 200 90th percentile 2,500 1,500 2,000 1,200 2,050 1,500 Table A.4: Oaxaca-Blinder decomposition of gender gaps in monthly earnings Differential Coef. Linearized std. err. t P>t [95% conf. interval] % of total difference Prediction_1 6.750818 0.010065 670.7 0.000 6.731089 6.770547 Prediction_2 6.200205 0.01934 320.59 0.000 6.162296 6.238114 Difference 0.5506126 0.021803 25.25 0.000 0.5078774 0.5933479 Endowments 0.0296419 5.4 0.009777 3.03 0.002 0.0104774 0.0488064 Coefficients 0.5257046 95.5 0.020489 25.66 0.000 0.4855451 0.5658641 Interaction −0.0047339 −0.9 0.005818 −0.81 0.416 −0.0161384 0.0066706 % of total endowmen ts effect Endowments Education 0.0014214 5 0.005497 0.26 0.796 −0.0093525 0.0121952 Industry 0.0146243 49 0.004038 3.62 0.000 0.0067096 0.022539 55 Differential Coef. Linearized std. err. t P>t [95% conf. interval] % of total difference Location −0.0045575 −15 0.002374 −1.92 0.055 −0.0092115 0.0000965 Age 0.0181538 61 0.004134 4.39 0.000 0.0100516 0.026256 % of total coefficients effect Coefficients Education −0.1370417 −26 0.038552 −3.55 0.000 −0.2126074 −0.061476 Industry −0.0498547 −9 0.053389 −0.93 0.35 −0.1545024 0.054793 Location 0.0634583 12 0.076606 0.83 0.407 −0.0866975 0.2136141 Age −0.0373285 −7 0.066865 −0.56 0.577 −0.1683902 0.0937331 _cons 0.6864713 131 0.136454 5.03 0.000 0.419009 0.9539336 % of total interaction effect Interaction Education −0.0005567 12 0.002158 −0.26 0.796 −0.004786 0.0036727 Industry −0.0032839 69 0.003563 −0.92 0.357 −0.0102677 0.0036999 Location 0.0010529 −22 0.00136 0.77 0.439 −0.0016134 0.0037192 Age −0.0019462 41 0.003498 −0.56 0.578 −0.0088035 0.004911 Table A.5: Monthly earnings by gender: regression results Male Female Education Some/Completed Primary 0.0622** 0.114* (0.023) (0.057) Secondary 0.183*** 0.291*** (0.032) (0.070) Post-Secondary and above 0.342*** 0.587*** (0.040) (0.052) Industry Manufacturing/Construction 0.223*** 0.167 (0.035) (0.104) Services 0.180*** 0.222*** (0.024) (0.046) Location Rural −0.118*** −0.172*** (0.023) (0.042) Age 25–34 0.231*** 0.0496 (0.032) (0.060) 35–44 0.301*** 0.170** (0.032) (0.058) 45–54 0.365*** 0.342*** (0.035) (0.062) 55–64 0.354*** 0.241** (0.039) (0.083) Constant 6.364*** 5.889*** 56 Male Female (0.037) (0.066) Observations 10,175 3,370 Note: Standard errors in parentheses *p < 0.05, **p < 0.01, ***p < 0.001 Table A.6: Yield determinants: regression results by head of household gender Sorghum yield Millet yield Groundnut yield Sesame yield Male Female Female Male Female Male Female Male head head head head head head head head Poor −0.127 0.0978 0.0000755 0.115 −0.338*** 0.0648 0.079 −0.569 −0.069 −0.229 −0.073 −0.134 −0.1 −0.216 −0.119 −0.66 Agricultural loan 0.0457 0.808 −0.319 0.272 −0.277 0.975** 0.208 1.039 −0.104 −0.52 −0.17 −0.278 −0.213 −0.334 −0.151 −0.677 Nonagricultural −0.188** 0.0982 −0.200** −0.121 −0.147 −0.0516 −0.0942 −0.720* loan −0.062 −0.195 −0.063 −0.115 −0.089 −0.211 −0.113 −0.346 Input use Fertilizer/pestici 0.281*** −0.323 0.206 0.731** 0.439* −0.0521 0.121 −1.926* de −0.082 −0.294 −0.141 −0.276 −0.206 −0.289 −0.144 −0.722 Labor 0.0515 0.028 0.114 0.215 0.127 0.319 0.220* −0.396 −0.063 −0.219 -0.078 −0.154 -0.12 −0.373 −0.11 −0.617 Machine 0.131* 0.0858 0.346*** 0.254 0.216* −0.288 0.197 0.898 −0.064 −0.243 −0.069 −0.129 -0.109 −0.214 −0.125 −0.56 Farm repair 0.239*** 0.223 0.117 −0.145 −0.320*** 0.221 0.167 −0.394 −0.064 −0.18 −0.078 −0.142 −0.096 −0.241 −0.099 −0.631 Fuel 0.250* −0.0776 −0.00783 −0.157 0.328 0.224 0.277 0.0915 −0.12 −0.423 −0.242 −0.264 −0.312 −0.62 −0.174 −0.911 Other Inputs −0.0767 0.421* −0.447*** 0.049 0.455*** −0.0597 0.314* 0.095 −0.086 −0.164 −0.095 −0.19 −0.088 −0.223 −0.135 −0.702 Irrigated 0.192* −0.104 0.182 0.603* 0.472* −0.559 0.261 0.0713 −0.092 −0.25 −0.113 −0.286 −0.224 −0.288 -0.242 −0.34 Land status (owned omitted) Rented 0.0738 −0.104 0.198* 0.0947 −0.0404 0.347 0.00455 0.118 −0.061 −0.199 −0.08 −0.111 −0.093 −0.188 −0.132 −0.415 Partially owned 0.0759 1.113** 0.197 −0.168 0.144 0 −0.313 * −0.142 −0.249 −0.19 −0.184 −0.323 (.) −0.405 Communal 0.113 0.28 −0.0807 0.288 0.0217 0.711** −0.233 -0.498 −0.107 −0.244 −0.113 −0.208 −0.159 −0.236 −0.16 -0.683 Area (plot size) −0.0320* −0.0465 −0.0578*** −0.0853 −0.0599** −0.0902* −0.0312* −0.0462* ** *** * ** ** ** −0.003 −0.027 −0.006 −0.012 −0.009 −0.023 −0.004 −0.011 No. of household 0.0349** −0.0258 0.0272 −0.0362 0.0718*** 0.0824 -0.0054 0.143 members −0.013 −0.047 −0.015 −0.026 −0.021 −0.044 −0.022 −0.098 Some/complete −0.0795 −0.682 0.0195 0.434* −0.104 0.588 0.15 −0.585 d primary −0.075 −0.474 −0.071 −0.174 −0.094 −0.36 −0.118 −0.452 Secondary 0.106 0.0857 0.756* −0.235 0.301 0.0934 57 Sorghum yield Millet yield Groundnut yield Sesame yield Male Female Female Male Female Male Female Male head head head head head head head head −0.134 −0.143 −0.344 −0.259 −0.408 −0.205 Post-secondary 0.382** −0.0248 0.945** −0.495* 1.215*** 0.354 and above * −0.143 −0.171 −0.182 −0.199 −0.241 −0.266 Constant 0.357*** 0.53 0.256* 0.613** 1.166*** 0.41 −0.414** −0.276 * −0.095 −0.338 −0.108 -0.15 −0.161 −0.297 −0.16 −0.658 Observations 1,860 194 1,196 340 542 125 413 38 R-squared 0.21 0.257 0.324 0.395 0.308 0.338 0.265 0.595 Source: Own calculations using NHBPS 2014/15. Note: All input use variables are greater than 100 SDG. Household weights and standard errors are robust. Standard errors are in parenthesis ***p < 0.01, **p < 0.05, *p < 0.1. Table A.7: Enterprise Survey gender indicators: by type and size of firms (%) Permanent Permanent Permanent Firms with Firms with Firms with full-time full-time non- full-time female majority a female production production workers who participation female top workers who workers who are female in ownership ownership manager are female (%) are female (%) (%) * * Type Manufacturing 15 7 3 9 9 12 Services 7 2 3 14 — — Wholesale 8 2 4 14 — — Retail 7 2 2 15 — — Size Small (5–19) 4 2 3 14 7 6 Medium (20–99) 13 3 5 15 10 18 Large (100+) 18 7 4 4 9 7 Table A.8: Agriculture and nonagriculture yearly income breakdown (a) Agriculture yearly income breakdown (b) Nonagriculture yearly income breakdown Male- Female- Male-headed Female-headed headed headed Industrial activities 2,626 1,408 Crop 2,235 1,267 Rented estates 111 134 Animal 788 316 Transfer support 555 3,604 Other agriculture 529 332 Other 2,530 1,867 Agriculture yearly 3,551 1,916 nonagriculture income Nonagriculture 5,822 7,013 yearly income Table A.9: Determinants of food insecurity by household head gender, probit estimations (coefficients) Male-headed Female-headed Area Rural −0.163* 0.566** (0.07) (0.20) Education 58 Male-headed Female-headed Some/completed primary −0.486*** −0.191 (0.07) (0.26) Secondary −0.711*** −0.136 (0.16) (0.44) Post-secondary and above −0.755*** 0 (0.19) (.) Employment status Unemployed 0.272 1.217** (0.21) (0.42) Out of labor force 0.127 0.478 (0.20) (0.48) Employment sector Agriculture 0.242*** 0.399* (0.06) (0.18) Industries −0.239* −0.0407 (0.10) (0.26) Household size Household size 0.212*** 0.207*** (0.01) (0.03) Receiving help Help received 0.198** −0.244 (0.07) (0.15) Constant −2.813*** −3.078*** (0.10) (0.27) Observations 7520 949 Note: Standard errors in parentheses, *p < 0.05, **p < 0.01, ***p < 0.001 Figure A.1: Main crops grown by poor/non-poor: male-headed and female-headed households a) Male-headed households b) Female-headed households 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Non poor Male-headed Poor male headed Non poor female-headed Poor female-headed Source: Own calculation based on NHBPS 2014/15 Note: Other crops include wheat, yellow maize, cotton, roselle, sunflower, Egyptian beans, beans, lentils, Arabic gum, and another crops. 59 Figure A.2: Input use by crop grown: male-headed versus female-headed households b) Groundnut a) Sorghum Pesticides/Fertilizer Farm repair Labour cost Machine/equipment repairs Pesticides/Fertilizer Labour cost Machine/equipment repairs Farm repair Other Fuel and lubricants Other 0% 20% 40% 60% Fuel and lubricants Male Female 0% Male Female c) Millet d) Sesame Machine/equipment repairs Labour cost Farm repair Labour cost Farm repair Machine/equipment repairs Pesticides/Fertilizer Other Other Pesticides/Fertilizer Fuel and lubricants Fuel and lubricants 0% 10% 20% 30% 0% 10% 20% 30% 40% 50% Male Female Male Female Source: Own calculation based on NHBPS 2014/15. Figure A.3: Difference in input use between female-headed and male-headed households 10% 5% 0% Fuel and lubricants Farm repair Machine/equipment repairs -5% -10% -15% -20% -25% Groundnut Sorghum Millet Seasam Source: Own calculation based on NHBPS 2014/15. 60 Figure A.4: Crop revenue proportion from total revenue: male-headed versus female-headed households 30% 27% 25% 26% 18% 18% 20% 13% 15% 12% 13% 8% 10% 10% 5% 7% 4% 0% Groundnut Sorghum Seasam Vegtables Millet Fruits Others Male-headed Female-headed Source: Own calculation based on NHBPS 2014/15. Figure A.5: Reasons for acquiring loans by household head gender 80% 70% 60% 50% 40% 30% 20% 10% 0% Other Buy animals Farm inputs Other agricultural costs Consumption needs Buy heavy equipment improvement of dwelling Consumer durables On-lending Buy other equipment Working capital and Land and/or building Other business expenses Buy agricultural land Religious, wedding, burial purchase of inputs equipment Purchase and Male-headed Female-headed Source: Own calculation based on NHBPS 2014/15. 61