76509 The Kyrgyz Republic: Poverty Profile and Overview of Living Conditions June 28, 2011 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank FISCAL YEAR January 1-December 31 CURRENCY EQUIVALENTS (Exchange Rate Effective as of May 13, 2011) Currency Unit = Kyrgyz som (KGS) US$1.00 = Kyrgyz som 46.5 Weights and Measures Metric System ABBREVIATION AND ACRONYMS GDP gross domestic product GNI gross national income KIHS Kyrgyz Integrated Household Survey MDG Millennium Development Goal NSC National Statistical Committee of the Kyrgyz Republic Vice President: Philippe H. Le Houerou Country Director: Motoo Konishi Sector Director Yvonne Tsikata Sector Manager: Benu Bidani Country Manager Alexander Kremer Task Team Leader: Sarosh Sattar ii TABLE OF CONTENTS THE KYRGYZ REPUBLIC: POVERTY PROFILE AND OVERVIEW OF LIVING CONDITIONS ................................................................................................................................ 1  A.  Development Indicators ....................................................................................................... 1  B.  The Profile of the Poor ......................................................................................................... 4  C.  Conclusion ......................................................................................................................... 19  List of Tables Table 1: Progress on Millennium Development Goals, Kyrgyz Republic ..................................... 3  Table 2: Composition of Poor and Nonpoor Households, by Number of Children, 2008, % ........ 9  Table 3: Average Persons per Household, by Age-Group ............................................................ 10  Table 4: Share of Individuals by Age-Group Who Are Poor or Nonpoor .................................... 11  Table 5: Employed Household Heads, by Type of Employment, percent ................................... 14  Table 6: Education Level of Household Heads, by Gender and Consumption Quintile .............. 15  Table 7: Education Level of Household Heads, by Gender and Urban-Rural Location .............. 16  Table 8: Education Level of Adults Aged 25 and Older, by Wealth Group and Sex ................... 16  Table 9: Children Attending School, by Poverty Status, Sex, and Age-Group ............................ 17  List of Figures Figure 1: Regional Comparison of Key Economic and Social Indicators ...................................... 2  Figure 2: The Urban-Rural Distribution of Poverty ....................................................................... 6  Figure 3: The Poverty Gap .............................................................................................................. 7  Figure 4: The Distribution of Poverty, by Oblast ........................................................................... 8  Figure 5: Poverty Distribution across Mountains and Plains.......................................................... 8  Figure 6: Household Size, by Poor or Nonpoor Category .............................................................. 9  Figure 7: Household Dependency Ratio, by Poor or Nonpoor Category ..................................... 10  Figure 8: Percentage of Poor and Nonpoor Households, by Age-Group ..................................... 11  Figure 9: Poverty and the Gender Status of the Household Head ................................................ 12  Figure 10: Domestic Migration among Household Heads, .......................................................... 13  Figure 11: Employment Status of Household Head, by Location, percent .................................. 14  Figure 12: Children of Basic School Age ..................................................................................... 18  List of Boxes Box 1: Measuring Poverty in the Kyrgyz Republic ........................................................................ 5  iii ACKNOWLEDGMENTS This report is a product of cooperative efforts between the National Statistical Committee of the Kyrgyz Republic and the World Bank. It reflects the continued cooperation between the two institutions on better policy analysis of poverty and related issues in the Kyrgyz Republic. The report would not have been possible without the cooperation of Orozmat Abdykalykov (Chairman of the National Statistical Committee of the Kyrgyz Republic) and Galina Samohleb (Head of the Survey Department). The World Bank’s task team members were Sarosh Sattar (team leader), Aibek Baibagysh uulu, Sanjukta Mukherjee, Yeva Gulnazaryan, and Sasun Tsirunyan. Administrative support was provided by Asel Almanbetova and Helena Makarenko. The sector manager is Benu Bidani. iv THE KYRGYZ REPUBLIC: POVERTY PROFILE AND OVERVIEW OF LIVING CONDITIONS Introduction 1. This report has been prepared by the World Bank based on data of the Kyrgyz Integrated Household Survey (KIHS) for 2008, which was collected by the National Statistical Committee of the Kyrgyz Republic (NSC) and made publicly available in late 2009. The report reflects the ongoing efforts by the World Bank to monitor and assess the poverty situation in its member countries. The report builds on the previous World Bank publication, in 2007, entitled “Kyrgyz Republic: Poverty Assessment� and provides a brief overview of poverty in the Kyrgyz Republic in 2008, the beginning of the global financial crisis. 2. The objective of the note is to offer a snapshot of living conditions and poverty in the Kyrgyz Republic in 2008. The note takes an analytical approach and supplies a quantitative assessment of the population’s welfare. The note covers the following components: an international comparison at the aggregate level of living conditions in the Kyrgyz Republic, an estimate of the level of poverty, and a description of the characteristics of the poor. 3. The outline of the report is as follows. In the first part, we give an account of the progress toward the achievement of the Millennium Development Goals (MDGs) and an international comparison of key social and economic indicators of the Kyrgyz Republic relative to selected countries in Europe and Central Asia, as well as low-income countries. The second part focuses on the main features that characterize the poor in the Kyrgyz Republic. A. Development Indicators 4. The Kyrgyz Republic is a low-income country with a gross national income (GNI) per capita of US$790 in 2008 (Atlas method) and US$870 in 2009. By income level, it ranks as the second lowest in the Europe and Central Asia region after Tajikistan. The average GNI per capita in the Europe and Central Asia region is US$6,880. However, the Kyrgyz Republic is well above the average income per capita of low-income countries, US$467. 5. In terms of selected social indicators, the Kyrgyz Republic scores among the lowest in the Europe and Central Asia region, but above the average among low-income countries (see figure 1). This may be explained by the better initial levels of social indicators in the country (since independence in 1991). However, because income per capita is growing slowly, the Kyrgyz Republic has limited fiscal space to increase expenditures on key social and physical infrastructure that could enhance the quality of the living conditions among the population. 1 Figure 1: Regional Comparison of Key Economic and Social Indicators a. GNI per capita, Atlas method, US$, current b. Primary school enrollment rate, net, % 16,000  100  90  89  94  92  84  87  14,000  90  80  12,000  70  10,000  60  8,000  50  6,000  40  4,000  30  20  2,000  10  � � KGZ CA CIS  WB EU�10 ECA KGZ CA CIS  WB EU�10 ECA Other Other c. Labor force participation rates, by sex, d. Maternal mortality ratio, % of female population aged 15–64 modeled estimate, per 100,000 live births 90  90  80  80  70  70  60  60  50  40  50  30  40  20  30  10  � 20  KGZ CA CIS  WB EU�10 ECA 10  Other � KGZ CA CIS  WB EU�10 ECA Female Male Other Source: World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data- catalog/world-development-indicators (2009 or latest available data). Note: Values of country aggregates are averages. CA = Central Asia (Kazakhstan, Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan); CIS other (Commonwealth of Independent States, other) = includes Armenia, Azerbaijan, Belarus, Georgia, Moldova, Russian Federation, and Ukraine. ECA = average for all developing countries of Europe and Central Asia. EU10 = the new member states of the European Union (Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia). KGZ = Kyrgyz Republic. WB = western Balkans (Albania, Bosnia and Herzegovina, Croatia, the former Yugoslav Republic of Macedonia, Montenegro, and Serbia). 2 Millennium Development Goals 6. The eight MDGs were established in 2000 to encourage development through improvements in economic and social conditions. Targets were set for 2015, and progress is regularly monitored by individual countries, as well as international organizations. Table 1 shows a selection of targets and the progress toward achieving them in the Kyrgyz Republic and selected country groups. The goals are (1) eradicate extreme poverty and hunger; (2) achieve universal primary education; (3) promote gender equality and empower women; (4) reduce child mortality rates; (5) improve maternal health; (6) combat HIV/AIDS, malaria, and other diseases; (7) ensure environmental sustainability; and (8) develop a global partnership for development. Table 1: Progress on Millennium Development Goals, Kyrgyz Republic Indicator Kyrgyz Republic LIC ECA 1990–99 2000–08 2000–08 2000–08 Hunger and poverty Prevalence of undernourishment (% of population) 15 14 33 7 Malnutrition prevalence, height for age (% of children under 5) 33 18 49 — Malnutrition prevalence, weight for age (% of children under 5) 8 3 33 — Poverty headcount ratio at US$1.25 a day (PPP) (% of population) 25 19 — 4 Schooling School enrollment, primary (% net) 90 85 71 91 Gender Share of women employed in the nonagricultural sector (% of total 49 48 — 47 nonagricultural employment) Health Maternal mortality ratio (modeled estimate, per 100,000 live births) 88 80 653 38 Maternal mortality ratio (national estimate, per 100,000 live births) — 69 — — Mortality rate, infant (per 1,000 live births) 58 37 82 24 Mortality rate, under 5 (per 1,000) 68 42 128 28 Incidence of tuberculosis (per 100,000 people) 144 157 310 95 Prevalence of HIV, total (% of population ages 15–49) 0 0 3 0 Environment Carbon dioxide emissions (metric tons per capita) 1 1 0 7 Improved sanitation facilities (% of population with access) 93 93 34 89 Improved water source (% of population with access) 78 86 62 95 Connectedness Internet users (per 100 people) 0 8 1 12 Mobile cellular subscriptions (per 100 people) 0 17 6 49 Telephone lines (per 100 people) 8 8 1 23 Other GNI per capita, Atlas method (current US$) 388 439 327 3444 Life expectancy at birth, total (years) 67 68 56 68 Gross capital formation (% of GDP) 18 19 21 22 Trade (% of GDP) 82 102 54 66 Source: World Development Indicators (database), World Bank, Washington, DC, http://data.worldbank.org/data- catalog/world-development-indicators (2009 or latest available data). Note: The table shows period averages. GDP = gross domestic product. LIC = low-income countries. ECA = the developing countries of Europe and Central Asia. PPP = purchasing power parity. — = not available. 7. The Kyrgyz Republic shows mixed progress in achieving the MDGs. Progress related to eradicating extreme poverty and hunger, ensuring environmental sustainability, and developing a global partnership for development appears to be on track to meet the goals. A national measure 3 of extreme poverty fell to 3 percent in 2008, well below the target level. At the same time, health- and education-related objectives are lagging, and it is not likely that the established targets will be reached. 8. The lack of progress in the health-related MDGs is a concern. There are three health MDGs: (1) reduce child mortality rates, (2) improve maternal health, and (3) combat HIV/AIDS, malaria, and other diseases. Despite economic growth and a dramatic decrease in poverty, there has been only a slow decline in the infant mortality rates over the period 2005–08, from 37 to 33 per 1,000 live births.1 The estimated maternal mortality ratio, though low relative to the average among low-income countries, increased from 78 to 81 deaths per 100,000 live births between 2005 and 2008. Statistics of the United Nations Children’s Fund, in fact, show that the annual percentage change in the estimated maternal mortality ratio between 1990 and 2008 was 0.3, that is, the indicator was increasing. The incidence of tuberculoses is high, at 159 per 100,000 people. In addition, though the number of detected cases of HIV is low, it is rising over time. 9. Primary education enrollment rates are lagging. World Bank data show that primary school enrollment rate (%, net) in the Kyrgyz Republic has been declining, that is, the indicator leveled at 88 in 1999, but, in 2008, it fell to 83. The trend indicates that it would not be possible to ensure that 100 percent of boys and girls would be able to complete the full course of primary schooling. The underperformance of the educational sector is likely related to the lack of comprehensive reforms, despite the fact that the Kyrgyz Republic allocates a considerable share of its budget to education. 10. The Kyrgyz Republic has not yet achieved its target in gender. The goal is to eliminate the disparities in access to education and to empower women in terms of the more active involvement of women in the political and economic life of society. The female to male ratio in primary school enrollment has been declining in recent years; it was at 99.4 percent in 2008, which is close to the target value. However, the female labor force participation rate is declining and far from the target level. Thus, the share of women in the total labor force was 45 percent in 1998, but declined to 42 percent in 2008. In addition, recent data of the United Nations Development Programme indicate that women’s salaries are only 67 percent of men’s salaries, and this level has been fairly constant over the last several years. B. The Profile of the Poor 11. This section discusses the level of poverty in the Kyrgyz Republic based on a national poverty line established by the NSC. The poverty line is the estimated Kyrgyz som value of a minimum consumption basket computed using the consumption patterns of food and nonfood items among low-income groups (box 1). The absolute (or total) poverty rate estimates the share 1 The data here are taken from statistics of the United Nations Children’s Fund, http://www.childinfo.org/. 4 of the population that cannot cover standards in both the need for food and nonfood needs, while the extreme poverty line measures the share of the population that is unable to meet even minimum daily food requirements. A new poverty line was established in 2008 by the NSC at a time when there were significant changes in relative prices in the economy. Box 1: Measuring Poverty in the Kyrgyz Republic The estimation of poverty is based on a standard World Bank method, which follows a basic needs poverty line using the consumption approach. The number of the poor is equal to the number of the members of a household whose per capita consumption is below the estimated poverty line. Normally, the estimation of a poverty line involves several stages. First, a consumption aggregate based on the food (including home produced food) and nonfood expenditures of a household are computed, including the imputed or computed user value of durables. Second, the reference group is identified, and this group’s consumption pattern is used as a basis of measurement. For the Kyrgyz Republic, the third, fourth, and fifth consumption deciles were used. The next step is to allocate 2,100 calories per day per person among the most important food items according to the consumption shares of these items in the reference group. The expenditure level that is required to reach the threshold of 2,100 calories per day per person is then applied as the food poverty line threshold to estimate the incidence of extreme poverty. The share of nonfood consumption is subsequently estimated for the group that is located just above the food poverty line. This share is used to estimate the allowance for nonfood consumption. Finally, the two calculations (for food and nonfood consumption) are used to estimate the total poverty line based on consumption. The straightforward implication of these calculations is that all those groups for which per capita consumption falls below the total poverty line are poor (and those whose per capita consumption falls below the food poverty line are extreme poor). In the Kyrgyz Republic in 2008, the NSC updated the poverty line in light of dramatic price changes and ensuing shifts in the consumption pattern of households. This means that the measurement of the national poverty line is not consistent across time, namely, before and after 2008. 12. The estimates of poverty are based upon the KIHS. This survey has involved interviews among a total of 5,016 households per year and has collected information on a total of about 19,060 persons. The survey is conducted quarterly on a rolling basis. Households remain in the sample for a period of four years, on average. The sample size is sufficiently large to allow for robust estimates of poverty at the national, urban-rural, and oblast (province) levels. The survey collects information on household and individual characteristics, expenditures (food and nonfood), income, assets, living conditions, and labor activities. 5 Number and geographical distribution of the poor 13. Almost one in three persons was living in poverty in the Kyrgyz Republic in 2008. The proportion of the population living below the poverty line was 31.7 percent. This translates into an estimated 1.7 million people in the Kyrgyz Republic who were living in poverty. Furthermore, an estimated 3.1 percent of the population was considered to be living in extreme poverty and unable to meet basic food requirements of 2,100 calories per day.2 14. The majority of the poor live in rural areas. The population of the Kyrgyz Republic remains largely rural; 64 percent of the population resides in rural areas (figure 2). This has a direct bearing on the nature of poverty in the Kyrgyz Republic. In 2008, 74 percent of the country’s poor were rural residents, equal to 1.2 million people, while the rest of the poor (24 percent) resided in urban areas, totaling 430,000 people. The incidence (or concentration) of poverty was also higher in rural areas than in urban areas at 37 and 23 percent, respectively. Figure 2: The Urban-Rural Distribution of Poverty a. By type b. By population Rural Urban Rural 7.7 Extreme� Poverty Urban 64 3.2 74 36.8 Absolute  36 Poverty 26 22.6 Distribution of the Poor Distribution of Population Source: Estimates based on KIHS 2008 data. Note: The distribution of the poor is the share of the poor residing in either the urban sector or the rural sector. 15. Poverty is relatively shallow in the Kyrgyz Republic. Estimates of the average gap between the actual consumption of the poor and the poverty line as a proportion of the latter (the poverty gap) indicate that poverty is not especially deep. In 2008, the total poverty gap was 7.5 percent (figure 3). In theory, this implies that the transfer of som 1,374 per year to an average poor individual would allow the individual to exit from the poor category. Similar to the case of 2 These poverty lines are based on consumption per capita (rather than per household) and are not adjusted for age or gender. 6 the headcount poverty rates, the poverty gap is more pronounced in rural areas. The depth of poverty is two times higher in rural areas than in urban areas: 9.0 percent in rural areas versus 4.5 percent in urban areas. Figure 3: The Poverty Gap 9.2 7.5 4.5 National Total Poverty Gap  Poverty Gap, Rural (%) Poverty Gap, Urban (%)  (%) Source: Estimates based on KIHS 2008 data. Note: The numbers in the figure are shown as a percent of the poverty line. 16. Oblast level poverty rates vary greatly. The total poverty rates range between 15 and 52 percent of the population, while the extreme poverty rates vary between 2 and 17 percent. There are five oblasts in which 38 percent or more of the population is living in poverty. The incidence of poverty is highest in Issyk-Kul, where the headcount reaches 52 percent of the oblast population, and the lowest is in Bishkek and Chui, where the headcounts are at 15 and 14 percent of the oblast population, respectively. 17. Half of the poor live in the two most populous oblasts of Osh and Jalal-Abad. The total share of the Kyrgyz Republic’s population living in Osh and Jalal-Abad is 44 percent, while 54 percent of the poor live in these two southern oblasts (figure 4). Though Issyk-Kul has the highest concentration of poverty, only an estimated 14 percent of the poor live there because the population size is comparatively modest. The oblast with the smallest share of the poor is Talas, which, because of its small size, has only 4 percent of the country’s poor. 18. The mountainous terrain of the Kyrgyz Republic increases the likelihood of poverty. The incidence of poverty is greatest in the mountains and least on the plains (figure 5). Only 13 percent of the population lives in mountainous areas; however, more than half of the population in those areas is poor. In contrast, 60 percent of all the poor live on the plains, which is explained by the fact that three quarters (74 percent) of the population of the republic resides on the plains. These observations underline the geographical dimension of poverty in the Kyrgyz Republic and also likely points to the relationship between poverty and the remoteness of the areas in which the poor live. 7 Figure 4: The Distribution of Poverty, by Oblast a. By type b. By population Absolute Poverty Extreme Poverty Distribution of the Poor Distribution of Population 30 52 25 43 43 24 40 38 19 16 15 14 21 17 16 15 8 8 8 7 7 12 10 5 6 5 4 4 5 5 2 2 Bishkek Abad Osh Chui Issyk� Naryn Batken Talas Jalal� Bishkek Naryn Issyk� Chui Batken Abad Osh Talas Jalal� kul kul Source: Estimates based on KIHS 2008 data. Figure 5: Poverty Distribution across Mountains and Plains a. By type b. By population Extreme  5 Poverty 74 Plain zones Plain zones 26 Absolute  60 Poverty Distribution of  Moderate  Moderate  9 13 Population mountainous  mountainous zones 46 19 zones Distribution of  the Poor 9 High  High mountainous  13 mountainous  zones 52 21 zones Source: Estimates based on KIHS 2008 data. 8 Demographic characteristics 19. Poor households are, on average, larger than nonpoor households. At the national level, the average household size is 3.8 persons, but, among the poor, the average household size is 5.1 persons, compared with only 3.5 persons per nonpoor household (figure 6 and table 2)3. The difference is more pronounced in rural areas (5.3 members for the poor compared with 3.8 members for the nonpoor) than in urban areas (4.5 for the poor, 3.0 for the nonpoor). Half of all poor households in the Kyrgyz Republic have two to three children each, whereas approximately 47 percent of nonpoor households have no children. This trend can be seen in urban and rural areas, too, though the difference is more striking in the sense that more than half (53 percent) of rural poor households have two to three children, while over half (54 percent) of urban nonpoor households have no children. Figure 6: Household Size, by Poor or Nonpoor Category 5.1 5.4 4.3 4.5 3.9 3.8 3.5 3.3 3.1 Poor Average Poor Average Poor Average Non�poor Non�poor Non�poor National Rural Urban Source: Estimates based on KIHS 2008 data. Table 2: Composition of Poor and Nonpoor Households, by Number of Children, 2008, % Nation Rural Urban Household type Poor Nonpoor Average Poor Nonpoor Average Poor Nonpoor Average Without children 16 47 39 14 41 33 22 54 48 One child 22 26 25 18 28 25 30 24 25 Two to three children 51 25 31 53 28 35 44 21 25 More than three children 12 2 5 15 3 7 5 1 2 Source: Estimates based on KIHS 2008 data. 3 This report uses per capita consumption as its welfare benchmark. This approach does not take into account demographic composition and economies of scale issues. As a result this approach may bias poverty among larger households and youth. 9 20. The difference in size between poor and nonpoor households is largely accounted for by differences in the number of children living in the household. The average number of persons per household aged over 64 is the same across all households irrespective of whether the household is poor or nonpoor, rural or urban. Figure 7 confirms that the dependency ratios of poor households are considerably higher (0.95) than those of nonpoor households (0.55); logically, one of the main reasons for the poverty of the former is the relatively high number of dependents (mainly children) per member of the household of working age.4 In table 3, the relationship is confirmed: the average number of children below the age of 15 is lower among households in the nonpoor category, even while the average number of persons aged 45 to 64 is greater in these same households, though the average number of persons aged over 64 is consistent across all groups. Figure 7: Household Dependency Ratio, by Poor or Nonpoor Category 1.01 0.95 0.79 0.73 0.65 0.61 0.65 0.55 0.49 Poor All Poor All Poor All Non�poor Non�poor Non�poor National Rural Urban Source: Estimates based on KIHS 2008 data. Table 3: Average Persons per Household, by Age-Group number Nation Rural Urban Age Poor Nonpoor All Poor Nonpoor All Poor Nonpoor All Under 5 0.6 0.3 0.3 0.7 0.3 0.4 0.5 0.2 0.3 5 to 14 1.4 0.7 0.8 1.5 0.8 1.0 1.1 0.6 0.6 15 to 29 1.2 0.9 1.0 1.3 1.1 1.1 1.1 0.8 0.9 30 to 44 1.0 0.6 0.7 1.0 0.6 0.7 1.1 0.6 0.7 45 to 64 0.7 0.8 0.7 0.7 0.9 0.8 0.6 0.6 0.6 65 and older 0.2 0.2 0.2 0.2 0.3 0.2 0.2 0.2 0.2 Source: Estimates based on KIHS 2008 data. 4 The dependency ratio = (sum of persons aged over 64 and under age 15) / (number of persons aged 15 to 64). 10 21. The age composition of poverty is skewed toward children and younger members of households. In terms of the distribution of children and the elderly among the poor, 11 percent of the poor are children under the age of 5 years; 27 percent of the poor are children 5 to 14 years of age; and 4 percent of the poor are people over the age of 64 (table 4 and figure 8). These facts highlight the importance of the demographic composition of households in poverty analysis. The presence of children seems to be one of the defining characteristics of poor households; this feature is more pronounced in rural areas. Table 4: Share of Individuals by Age-Group Who Are Poor or Nonpoor percent Nation Rural Urban Age Poor Nonpoor All Poor Nonpoor All Poor Nonpoor All Under 5 12 7 9 12 8 9 10 7 8 5 to 14 27 19 22 28 20 23 24 18 19 15 to 29 24 27 26 24 27 26 25 26 26 30 to 44 20 18 19 18 16 17 23 21 22 45 to 64 13 22 19 13 22 19 13 21 19 65 and older 5 7 6 4 7 6 5 7 7 Source: Estimates based on KIHS 2008 data. Figure 8: Percentage of Poor and Nonpoor Households, by Age-Group a. Nation b. Rural c. Urban National Poor National Non�poor National All Rural Poor Rural Non�poor Rural All Urban Poor Urban Non�poor Urban All 28 27 27 27 26 26 2626 25 24 24 24 23 23 22 22 22 2122 21 20 20 19 19 19 19 19 1819 18 18 17 16 13 13 13 12 12 10 9 9 8 8 7 7 77 7 7 6 6 5 5 4 Under 5 5 to 14 15 to 29 30 to 44 45 to 64 65 and  Under 5 5 to 14 15 to 29 30 to 44 45 to 64 65 and  Under 5 5 to 14 15 to 29 30 to 44 45 to 64 65 and  older older older Source: Estimates based on KIHS 2008 data. 11 Gender characteristics 22. Gender seems to play a mixed role in defining poverty in the Kyrgyz Republic. More individuals in absolute poverty (33 percent) live in man-headed households rather than woman- headed households (29 percent) (figure 9). Meanwhile, a higher proportion of individuals in extreme poverty live in woman-headed households (8 percent), and a lower proportion live in man-headed households (5 percent). Figure 9: Poverty and the Gender Status of the Household Head Male headed household Female headed household 33 Absolute  Poverty 29 5 Extreme  Poverty 8 Source: Estimates based on KIHS 2008 data. Domestic migration status 23. Migration status has a significant bearing on poverty outcomes. The relationship between the level of per capita consumption and the internal migration status of the household head is positive. The share of household heads who have migrated within the country is larger among the more well of category. Moreover, the more well off household heads stayed longer in the new places of migration (figure 10). Only 16 percent of the household heads in the lower quintiles migrated, whereas more than 33 percent in the fourth quintile and 46 percent in the fifth quintile changed the place of residence. By oblast, 70 percent of household heads presently residing in Bishkek, the capital, were born elsewhere, and nearly 64 percent of household heads residing in Chui were born elsewhere. After migrating, the less affluent household heads in the first quintile stayed, on average, 26 years in the new locations, while the household heads in the fifth quintile, that is, the more affluent, lived for 31 years in the new places of residence. More detailed data on the urban-rural aspect of migration and in terms of oblasts reveal that 47 percent of urban heads of households migrated from elsewhere and 70 percent of household heads residing in Bishkek were born elsewhere. This provides an initial explanation for the positive relationship between migration status and per capita consumption. Urban areas and, specifically, Bishkek city, which 12 are the main centers attracting domestic migrants, might provide better employment and better wage opportunities, thus allowing for better income prospects and higher per capita consumption. So, domestic migration is one of the most important coping mechanisms of the poor in rural areas. Figure 10: Domestic Migration among Household Heads, by Quintiles of Per Capita Consumption a. By percentage b. By years of residence in migration 46% 30 33% 28 28 27% 27 22% 26 25 16% 16% 1 2 3 4 5 1 2 3 4 5 All Quintiles of PC Consumption All Quintiles of PC Consumption Source: Estimates based on KIHS 2008 data. Employment status 24. The poor and nonpoor do not differ greatly in terms of their place and status of employment. The most striking observation based on the KIHS 2008 data arises from the attempt to link the poor with employment status. At the aggregate level, 71 percent of poor household heads were employed, while only 66 percent of nonpoor household heads had work (figure 11). This type of relationship is also apparent in rural areas: 74 percent of poor household heads have jobs, while relatively lower percent of nonpoor household heads are employed, i.e., only 65 percent. It appears that at national and rural levels the poor household heads have higher employment rate compared to nonpoor household heads. The usual relationship only weakly holds in urban areas: 64 percent of poor household heads have jobs versus 69 percent of non- poor household heads with employment. The nonconventional fact observed at national and rural levels merits an explanation. It may be that the employed category includes a large share of the underemployed, which has more features in common with unemployment than with employment. In addition, household heads are more typically elderly in the Kyrgyz Republic. These people may generally report their status as employed even if they are only marginally involved in cultivation in gardens or small plots of land (especially in rural areas), and this may 13 tend to exaggerate the number of the employed heads among poor households. The KIHS data does not collect sufficient detailed information on labor markets to allow for a definitive explanation of the employment and poverty relationship.5 Figure 11: Employment Status of Household Head, by Location, percent a. Nation b. Rural c. Urban Unemployed Employed Unemployed Employed Unemployed Employed 71 74 67 68 67 69 68 65 64 33 36 36 29 32 33 31 32 26 poor non poor Total  poor non poor Total  poor non poor Total  National Rural Urban Source: Estimates based on KIHS 2008 data. 25. As with employment status, the type/place of employment seems to be irrelevant for defining the linkage to the poor. In rural areas, poor and nonpoor heads of household occupy mostly jobs in enterprises, collective farms, and peasant farms (table 5). In urban areas, the poor and nonpoor are largely involved in enterprises and wage work for private individuals. The reason there is no direct correlation between the type or place of employment and poverty status in urban and rural areas may be related to the fact that the data aggregate some nuances among the types of employment. This may, in turn, mask important details about the positions, salaries, and sectors of employment, which are important in defining the poor. However, at the aggregate (national) level, the poor, relative to the nonpoor, are overrepresented in the peasant farm category and underrepresented in the private individual and enterprise or organization categories. Table 5: Employed Household Heads, by Type of Employment, percent Nation Rural Urban Type or place of employment Nonpoor Poor Total Nonpoor Poor Total Nonpoor Poor Total Enterprise, organization, collective farm, agricultural cooperative 43 35 41 36 32 35 52 45 51 (Peasant) farm 12 25 15 21 32 25 2 3 2 Individual commercial activities 1 0 1 — — — 1 1 1 Individual basis 21 15 20 21 14 19 22 20 21 Wage work for private individuals 19 20 19 15 16 15 23 29 24 Other 4 5 4 7 6 7 1 1 1 Source: Estimates based on KIHS 2008 data. Note: — = not available. 5 The labor market module of the KIHS 2008 was not shared with the World Bank. 14 Educational characteristics 26. As the data show, the educational attainment of the household head is a significant factor defining income status. Almost half the poor (45 percent) live in households in which the household heads are illiterate, and only 13 percent live in households in which the household heads have attained higher education. This confirms the positive conventional correlation between education and income levels. However, the statistics do not allow one to determine the casualty of the relationship. Human capital plays an important role in determining whether an individual is nonpoor, and, thus, educational level represents another characteristic of poverty, that is, low educational attainment. Furthermore, the educational differences highlight the urban- rural divide. In urban areas, man- and woman-headed households are more likely to have university or professional education, whereas the heads of rural households typically do not continue beyond secondary school. 27. Table 6 shows the educational distribution of household heads by sex and quintiles of per capita consumption. One may observe that men and women household heads in the fourth and fifth quintiles, that is, the more affluent households, account for a large share of the individuals who have obtained university degrees. Thus, 23.9 percent of men and 33.7 percent of women household heads in the fifth quintile have university degrees, compared with only 8.2 percent of men and 10.4 percent of women household heads in the first quintile. In the case of complete secondary education, one sees that the trend works in the opposite direction, that is, close to half of men and women household heads in the first, second, and third quintiles have completed secondary education, as opposed to only a third of men and women household heads in the fourth and fifth quintiles. It appears that, in the no education category, women household heads are more likely than men heads of household to fall into the low consumption quintiles (first, second, and third): the poor groups. The situation is similar across the urban-rural divide (table 7). Table 6: Education Level of Household Heads, by Gender and Consumption Quintile percent 1 quintile 2 quintile 3 quintile 4 quintile 5 quintile Education level Wome Wome Wome Wome Wome Men Men Men Men Men n n n n n University 8 10 13 7 13 8 25 20 24 34 Secondary professional 2 0 1 1 0 1 2 1 3 3 education Primary professional/technical 32 20 28 21 37 16 26 32 33 27 education Complete secondary (grades 51 40 47 48 41 50 37 29 32 23 10–11) Basic secondary (grades 5–9) 3 7 6 3 4 7 5 7 5 5 Basic primary (grades 1–4) 3 13 6 10 3 9 5 8 3 7 No education 0 9 0 10 1 10 1 3 0 2 Source: Estimates based on KIHS 2008 data. 15 Table 7: Education Level of Household Heads, by Gender and Urban-Rural Location percent Urban Rural All Education level Male Female Male Female Male Female University 28 32 11 9 18 20 Secondary professional education 3 3 1 0 2 2 Primary professional/technical education 28 26 33 22 31 24 Complete secondary (grades 10–11) 34 28 44 42 41 34 Basic secondary (grades 5–9) 4 4 5 7 5 6 Basic primary (grades 1–4) 3 5 5 13 4 9 No education 1 3 0 8 1 6 Source: Estimates based on KIHS 2008 data. 28. Poor households are associated with low educational attainment among all household members above 25 years of age. The poor seem generally to stop their education at the completion of secondary school (table 8). This relationship holds in equal degree among both men and women, except in the no education category, where women represent the higher share compared with men. Table 8: Education Level of Adults Aged 25 and Older, by Wealth Group and Sex percent Extreme poor Poor Nonpoor All Education level Men Women Men Women Men Women Men Women University 1 6 9 10 21 23 17 20 Secondary professional education 2 0 2 1 2 2 2 1 Primary professional/technical 18 17 26 23 29 25 28 25 education Complete secondary (grades 10–11) 48 59 50 53 38 38 42 42 Basic secondary (grades 5–9) 21 8 8 5 6 5 6 5 Basic primary (grades 1–4) 10 7 5 5 4 4 4 5 No education 0 3 0 4 1 3 1 3 Source: Estimates based on KIHS 2008 data. 29. There is a gender difference in school attendance within the extremely poor household group. In the extreme poverty category, the school attendance of girls aged 7–10 years is significantly lower (45 percent) than the school attendance of boys (74 percent) in the same age cohort (table 9). This is true irrespective of the urban-rural divide. Across all age-groups in the extreme poverty category, the school attendance of girls is lagging behind the school attendance of boys. In the poor category, the attendance of 17- to 19-year-old girls in age-appropriate schooling (66 percent) is lagging behind that of their male peers (74 percent). Based on this, one might assume that poor households are less likely to send girls to school, which implies the existence of gender inequalities in school attendance. In the nonpoor category, the attendance of girls in school is roughly equal to that of boys up to 17 years of age, but, between the ages of 17 and 24, a higher percentage of girls attend school. 16 Table 9: Children Attending School, by Poverty Status, Sex, and Age-Group percent Extreme poor Poor Nonpoor All Age Male Female Male Female Male Female Male Female Nation 7–10 74 45 64 65 68 66 66 65 11–13 95 87 98 98 100 100 99 99 14–16 99 93 95 99 99 99 98 99 17–19 45 36 74 66 83 89 80 82 20–24 3 5 8 12 34 42 28 35 Urban 7–10 70 48 73 63 69 69 70 68 11–13 70 100 94 99 100 99 98 99 14–16 100 83 96 96 100 99 99 98 17–19 83 74 68 82 75 86 74 85 20–24 3 16 22 22 48 50 43 44 Rural 7–10 76 44 60 65 67 63 64 64 11–13 100 82 99 98 100 100 100 99 14–16 99 98 95 100 98 99 97 99 17–19 42 23 75 60 88 92 83 80 20–24 3 2 3 6 25 37 20 30 Source: Estimates based on KIHS 2008 data. Note: The figure shows school of all types: grades 1–11, special/professional/technical secondary school, and university. 30. The urban-rural differential in terms of school attendance among extremely poor children above the age of 17 is significant and calls for policy attention. An estimated 83 percent of boys aged 17–19 years in the extremely poor group attend school in urban areas, compared with only 42 percent of the corresponding boys in rural areas. Similarly, 74 percent of girls aged 17–19 years in the extremely poor group attend school in urban areas compared with only 23 percent of the corresponding girls in rural areas. This suggests that, among the extremely poor population in rural areas, there are impediments to continued attendance in schools of higher education either because of a lack of access or because of perceived lower returns to higher education in rural areas relative to urban areas. 31. As the statistics on net and gross enrollment in basic school show, enrolment rates in the extreme poor category are considerably higher among boys than among girls. This gender difference is driven by the differential enrolment rates between girls and boys in rural areas in the extreme poor category (figure 12). For instance, in extremely poor households in rural areas, 94 percent of boys aged 7–14 are enrolled in school, compared with only 58 percent of girls. In urban areas, there is no significant difference in the net or gross enrolment rates among boys and girls across welfare groups. The net and gross enrolment rates in basic school among boys and among girls are roughly equal for the poor and nonpoor groups, irrespective of urban or rural location. 17 Figure 12: Children of Basic School Age Who Are in the Proper Grade for Age and Total Students in Basic School a. Nation % of children aged 7�14 attending school Any children in BASIC school (grades 1�8) 105 107 103 104 100 99 100 89 85 81 81 82 84 84 83 62 Male Female Male Female Male Female Male Female Extreme poor Poor Non�poor All b. Urban areas % of children aged 7�14 attending school Any children in BASIC school (grades 1�8) 108 105 105 106 99 103 83 85 85 84 85 84 85 79 71 74 Male Female Male Female Male Female Male Female Extreme poor Poor Non�poor All 18 c. Rural areas % of children aged 7�14 attending school Any children in BASIC school (grades 1�8) 106 105 108 99 103 103 98 94 86 84 83 82 80 80 81 58 Male Female Male Female Male Female Male Female Extreme poor Poor Non�poor All Source: Estimates based on KIHS 2008 data. Note: Basic school corresponds to grades 1–8. Proper grade for age indicates net enrollment. Total students in school corresponds to gross enrollment. C. Conclusion 32. From this report, one may conclude that the poor individual is more likely to reside in remote rural areas, perhaps living in a more populous oblast, such as Osh. The poor are typically young individuals living in a household in which the dependency ratio is relatively higher than the ratio among nonpoor households. Poor households more usually have five or more members, including many younger children. There is an equal chance for the household head in a poor household to be a man or a woman; however, extremely poor households are more likely to be headed by women. The lower the educational attainment of the household head, the more likely the household falls into the poor category. The poor have most likely not migrated within the country or, if they have migrated, have stayed only a few years in the new place of residence, which is likely to be an urban area. Surprisingly, employment status has no direct correlation with being poor, which might be a reflection of the low-income levels of those who are employed. In contrast, the educational attainment of the individual plays a substantial role in defining poverty status; the link is significant and positive. One may also observe that the extreme poor have difficulty in attending or completing school, especially in rural areas. In the extreme poor category, the urban-rural divide is exacerbated by gender inequalities in school attendance. All these findings call out for policy attention to prevent a widening of the gap between rural and urban residents and between men and women. 19