Social Protection & Jobs Global Practice October 2018 Europe and Central Asia Region Uzbek Population Risk and vulnerability assessment Table of contents Acknowledgments........................................................................................................ iii Executive summary ..................................................................................................... iv 1 Introduction ..................................................................................................... 1 1.1 Conceptual framework and methodology ............................................. 1 1.2 Data sources ....................................................................................... 2 2 Identification of most relevant risks and shocks ............................................... 5 2.1 Risks associated with early childhood and school age ......................... 6 2.2 Risks associated with young adults and working-age people ............... 7 2.3 Risks for the elderly ........................................................................... 11 2.4 Shocks and risks across the lifecycle ................................................. 11 3 Who is vulnerable and why? ......................................................................... 16 4 Coping strategies .......................................................................................... 18 4.1 Migration ............................................................................................ 19 5 Policies to reduce vulnerability ...................................................................... 21 6 References.................................................................................................... 23 List of tables and figures Table 1: Economic activity by age group and sex .............................................. 9 Table 2: Catastrophic health expenditure by population subgroup ................... 14 Figure 1: Risks across the lifecycle ..................................................................... 2 Figure 2: Geographical coverage of qualitative study .......................................... 3 Figure 3: Population age distribution by quintile .................................................. 5 Figure 4: Percentage of households (HHs) with children by quintile .................... 6 Figure 5: Percentage of children aged 3–6 attending preschool by quintile ......... 7 Figure 6: Percentage of young adults (20–24) attending educational institutions by quintile 8 Figure 7: Activity status among adults aged 25–34 by quintile ............................ 9 Figure 8: Type of main job among adults aged 25–34 by quintile...................... 10 Figure 9: Issues to be tackled with immediate priority ....................................... 11 Figure 10: Disability and need for health assistance by age group...................... 12 i Figure 11: Impoverishing health expenditure by quintile ..................................... 13 Figure 12: Sources of income by quintile ............................................................ 18 Figure 13: Percentage of households with at least one member who migrated abroad by quintile, 2013 and 2018 .............................................................................. 19 Figure 14: Percentage of population 15+ working abroad by age group, disaggregated by quintile and by sex .......................................................................... 20 Figure 15: Percentage who said their family situation now is worse than two years ago 21 List of Abbreviations CALISS Central Asia’s Labor and Skills Survey FGD Focus group discussion Khokimiyat Local authority administration KII Key informant interview L2CU Listening to the Citizens of Uzbekistan Mahalla Local self-government community NGO Nongovernmental organization OOP Out-of-pocket PPP Purchasing power parity ii Acknowledgments This report was prepared by a team comprising Ludovico Carraro (Consultant, the World Bank), Maddalena Honorati (Senior Economist, the World Bank) and Manuel Salazar (Lead Social Protection Specialist, the World Bank) under the guidance of Cem Mete (Practice Manager) and Hideki Mori (Country Manager). Trinidad Berenice Saavedra Facusse (Consultant, the World Bank) provided excellent data analysis support. Kakhramon Yusupov (Consultant, the World Bank) coordinated the qualitative data collection undertaken by Al Mar Consulting and provided inputs throughout the process. Ruslan Yemtsov (Lead Economist, the World Bank) and Phillippe Leite (Senior Social Protection Specialist) served as peer reviewers and offered valuable insights to the paper. William Hutchins Seitz provided excellent comments. The study was made possible thanks to the Rapid Social Response multi-donor Trust Fund grant “Supporting Uzbekistan’s Inclusive Economic reform through Strengthening its Social Protection System”. iii Executive summary This report presents the results of a diagnostic study on the main risks households face in Uzbekistan and the strategies they adopt to cope with them. This is relevant to understand whether and how social protection is reducing vulnerability and identify policy areas of possible improvement. The study is part of the collaboration between the Government of Uzbekistan and the World Bank in improving the effectiveness of social protection. To identify and assess the relevance of different risks, the study used two main data sources: the Listening to the Citizens of Uzbekistan (L2CU) household survey and a qualitative study that conducted interviews and focus group discussions (FGDs) in four different regions of the country. Both data sources are very recent as they were collected between May and September 2018. The analysis looked at risks across five main phases of the lifecycle – early childhood, school age, youth, adults in working age, and adults in old age – and wherever possible, disaggregated the analysis by sex. Furthermore, the study considered whether people in the bottom and top of the income distribution are affected differently by such risks. Among early childhood the most prominent risk identified was access to preschool education: only one-quarter of children aged three to six attend preschool, with very significant differences in the bottom and top quintiles. On the contrary, attendance is very high for general education, but higher education is almost exclusively pursued among people in the richest quintile. Young adults struggle to enter the labor market, especially if they live in relatively poor households and, while the employment rate increases for older working ages, rates remain significantly lower in the bottom quintiles. Furthermore, lower employment rates in the bottom part of the distribution are accompanied by occupations that are more likely to be precarious (seasonal, temporary, and informal jobs). Very significant differences arise in the main activities of working-age people by sex, with women having a much smaller rate of employment, often being “homemakers” looking after children. Both the quantitative and the qualitative study revealed that one main concern of people is the lack of jobs as well as low salaries, especially among the poorest and the beneficiaries of social assistance. Overall, the elderly appear to be well protected by social protection transfers, but both disability prevalence and need for medical assistance increase with age, resulting in high health expenditure. Therefore, elderly persons with low pensions are particularly vulnerable to health-related needs, and health-related risks are repeatedly quoted in all FGDs as a significant worry among the population. Furthermore, the analysis clearly showed how households in receipt of social assistance, or where at least one member is a person with disabilities, tend to face catastrophic health expenditure (defined as reducing essential expenses or representing more than 40 percent of their expenditure after subtracting essential expenses) and often forgo necessary medical treatments. Other risks that affect people across the lifecycle are macroeconomic shocks and natural disasters. Eighty percent of people reported being affected by the high inflation in the country and this percentage increased among beneficiaries of social assistance. Finally, in some regions people are worried by changes in the weather patterns (hotter iv summers and colder winters) and environmental problems affecting the Aral Sea. These changes affect livelihoods, especially those related to agriculture. People adopt both short- and long-term coping strategies against the above risks. The main short-term strategies include asking for support from relatives and friends, applying for social assistance, taking up temporary jobs or public works (even if at low wages) or migrating abroad for work, selling assets, or asking employers for advance payments. Long-term strategies consist of investing in children’s education and relying on their incomes when they grow up (sometimes through work abroad), or making small investments that can support new or improved self-employment activities. Such small investments have been promoted through microloans provided at preferential rates or are possible if households receive significant remittances. Although migration has reduced, it is still a very significant phenomenon, which is of particular importance for relatively poorer households: one out of four households in the bottom quintile has at least one household member working abroad, and remittances are an important source of income for such households. This shows how migration is an important mechanism to address lack of jobs and cope with vulnerability. Migrants tend to be young and male, but it is increasingly felt that migration to Russia is becoming riskier, with people worrying about migrants’ safety and well-being. Based on this analysis and people’s own recommendations during the FGDs, the policy areas that deserve specific attention to reduce vulnerability are the following: • Improve job opportunities, especially in rural areas, through new investments in enterprises and infrastructure, or through better loan schemes at preferential rates for low-income households. • Improve social protection measures by simplifying the administration process, ensuring that amounts of support keep pace with inflation, and by addressing challenges of high health expenditure among vulnerable groups. • Increase the availability of kindergartens and ensure that access is free of charge for low-income families. • Improve the way higher level education meets the demands of the labor market, including the quality of education at vocational and tertiary level. Since the government is planning further reforms that may generate macroeconomic instability, ensuring that social assistance is responsive and adequate is extremely critical. v 1 Introduction As part of the collaboration between the Government of Uzbekistan and the World Bank in improving the effectiveness of social protection, the Bank conducted a diagnostic study on the main risks households face and the main strategies adopted to reduce vulnerability and cope with such risks. The main objective of the study is to determine the extent to which social protection is addressing such risks and in which areas gaps remain. This could inform areas of potential further work and collaboration between the Government of Uzbekistan and the World Bank. 1.1 Conceptual framework and methodology The study intends to answer a number of key questions: What risks do people face and how do these affect poverty and vulnerability? Who is considered vulnerable in Uzbekistan? What strategies do people adopt to reduce and mitigate such risks? How can the government intervene to reduce vulnerability and poverty, particularly in relation to social protection policies? To answer these questions the study uses two main sources of data: a nationally representative household survey (Listening to the Citizens of Uzbekistan, L2CU) and a qualitative study with focus group discussions (FGDs) and in-depth interviews across different regions of Uzbekistan. The household survey was used to analyze some of the main risks and determine the extent to which people and households across the income distribution are affected by such risks. The qualitative study investigated people’s understanding of poverty and vulnerability, and discussed why people are vulnerable, the main coping strategies adopted by households, and how government interventions could help reduce vulnerability. The approach used to understand the relevance of different risks is based on a common framework that starts with a diagnostic assessment of the main risks and vulnerabilities faced by the population. Such assessment follows the lifecycle, which is divided into five main phases: early childhood, school age, young adulthood, working age, and old age. Each stage has its own risks, which whenever possible are also analyzed by sex, and takes into account the changing needs of people and their age-specific risks of failing to progress and participate in society. The conceptual framework shown in Figure 1 highlights the importance of shocks that could occur at any time during the lifecycle and affect either a single person or household (such as an illness) or a whole region/country (such as natural disasters or macroeconomic instability). Finally, the figure highlights the importance of disability and chronic illnesses, which could be acquired at any stage of the lifecycle and present new risks. This conceptual framework helps in providing a systematic approach to identify and assess the risks that can affect the Uzbek population. The assessment starts by providing the results of the analysis of the household survey and determines not only the overall relevance of a certain risk, but also its incidence across the welfare distribution by grouping people in five population quintiles, ranked from poorest to richest in terms of per capita consumption expenditure.1 The quantitative survey does not have information for 1Consumption expenditure was also adjusted to take into account price differences across the country. The per capita correction implicitly assumes that each member has the same needs and that there are no economies of size within the household. To determine the sensitivity of the results to this fundamental 1 all risks, but its analysis is supported and complemented by the qualitative study, which also inquired about risks affecting the population and collected information on why such risks are relevant. Figure 1: Risks across the lifecycle 1.2 Data sources Two main data sources were used for this assessment: a nationwide country representative household survey and a qualitative assessment. The quantitative survey, “Listening to the Citizens of Uzbekistan” (L2CU), interviewed just over 4,000 households across the country between May and August 2018, covering around 200 mahallas across the country in both urban and rural areas and across the 14 regions of Uzbekistan (12 provinces, Karakalpakstan, and Tashkent City). The survey also focused on beneficiaries of social benefits and purposively oversampled them to be able to provide reliable estimates for this specific subgroup of the population. The questionnaire collected information on a broad number of topics: household composition, migration, education, health, employment, banking and savings, social benefits, household income, consumption expenditure, dwelling characteristics, and people’s opinions of the current quality of infrastructure facilities, the performance of local administrative organizations, and the priority of different types of interventions. assumption, the analysis was also conducted using a common set of equivalence scales used in Eastern Europe and Central Asia, also known as the old OECD (Organisation for Economic Co-operation and Development) equivalence scales. These assume that the first adult in the household counts as 1.0, and each other adult member as 0.7, whereas children younger than 15 years count as 0.5. While using this different adjustment inevitably changes the estimates, the fundamental results and trends by quintile provided in this report remain the same, and so appear to be robust to this different assumption. 2 Information collected in the survey enables calculation of both household income and household consumption expenditure, which can be used as welfare indicators once adjusted by household size and composition as well as differences in prices. In this analysis the preferred welfare measure is per capita consumption expenditure, which was used to group the population in quintiles, each containing 20 percent of the population, and ordered from poorest to richest. The qualitative study was also conducted between May and September 2018, and involved a mixture of FGDs and key informant interviews (KIIs). The study did not aim to be country representative, but was conducted in various parts of the country to ensure the coverage of different contexts and situations. More specifically, the study was conducted in four different regions: Tashkent city, Fergana, Khashkadarya, and Karakalpakstan (Figure 2 shows the specific locations included in the study). In all regions but Tashkent city, interviews occurred within four different mahallas: one in urban areas and three in rural areas. In each mahalla two FGDs were conducted, one with beneficiaries of social transfers and another with nonbeneficiaries that were relatively poor households (either ex- beneficiaries, or people who applied for some form of support but were rejected). Discussions covered people’s understanding of poverty and vulnerability, strategies to cope with risk and uncertainty, and ways in which the government can reduce vulnerability. KIIs were conducted with the district Khokimiyat, Employment Support Center, and Pension Fund as well as with mahalla representatives. Interviews covered existing government interventions with a specific focus on social protection programs and possible areas of improvement. Figure 2: Geographical coverage of qualitative study The remainder of the report is organized as follows. The next section identifies risks across the lifecycle, whenever possible disaggregating by sex, and determines the significance of different risks. In the same section results from the qualitative study 3 providing evidence on the risks analyzed are reported. The third section summarizes the evidence gathered through the qualitative study on who the vulnerable are and why they are vulnerable. The fourth section looks at the strategies for coping with risks and mitigating their impact. This is done providing evidence from both the qualitative and the quantitative study. The last section summarizes areas where policies and programs appear to be insufficient and identifies possible government interventions that could reduce vulnerability in Uzbekistan. 4 2 Identification of most relevant risks and shocks Before analyzing and presenting the incidence of different risks across the lifecycle, it is useful to look at the overall age distribution of the population in each quintile. This enables identification of whether some age groups are relatively poorer than others. Figure 3 shows the result of this analysis. Each bar shows the percentage of population of a certain age in the respective quintile. For example, 19 percent of people in the bottom quintile are below the age of 6. The sum of all age percentages within each quintile is equal to 100. The figures illustrate that young people (aged below 7 and between 7 and 14) are more likely to be found among the relatively poor, and the same applies for those aged between 25 and 34, who are likely to be parents of young children. Therefore, some evidence exists that overall in Uzbekistan young families with children are those relatively more likely to be poor, compared to young adults and older people. Figure 3: Population age distribution by quintile Source: Analysis of L2CU 2018. Note: In each quintile, the sum of percentages of different age groups is equal to 100. The above finding is confirmed by looking at the percentage of households with children across quintiles (Figure 4). While in the bottom quintile the percentage of households with at least one child younger than 7 is 65 percent, the percentage decreases to just 35 percent among the top quintile, and similar differences exist for households with at least one child aged 7–14. Therefore, in general it can be seen that households with children are more likely to be poor. The following analysis of risks thus needs to identify the underlying issues faced by such households. Finally, it should be noted that evidence of challenges faced by households with children also emerged during the FGDs. Not only were young families with many children identified as a vulnerable group, but in various occasions it was also described how 5 households move out of a situation of poverty and vulnerability when children grow up, get married, and move out of the home. Figure 4: Percentage of households (HHs) with children by quintile Source: Analysis of L2CU 2018. 2.1 Risks associated with early childhood and school age It is widely recognized that the early years of life are crucial for the development of children, and inadequate care and nutrition can have very long-term negative impacts. Uzbekistan historically had significant problems of undernutrition (stunting and wasting), but these have declined considerably over the years. For example, the prevalence of stunting among children under five decreased from 40 percent to 20 percent between 1996 and 2006 (Development Initiatives 2017); this prevalence is expected to have decreased further, but more recent estimates are not available (the next Multiple Indicators Cluster Survey is expected to take place in 2019). The challenge instead seems to come from certain micronutrient deficiencies and anemia. Uzbekistan also has good coverage of maternal health care and good access to ante- natal and post-natal care. Indeed, infant and under-five mortality rates have also declined steadily. However, one area that appears particularly underdeveloped is the possibility to enroll in preschool education. The L2CU survey revealed that only about one-quarter of children aged 3–6 are attending preschool, and significant differences exist across quintiles. In the bottom quintile only 17 percent attend preschool whereas the percentage increases to 33 percent among those in the top quintile. No significant differences exist between boys and girls. 6 Figure 5: Percentage of children aged 3–6 attending preschool by quintile Source: Analysis of L2CU 2018. The lack of kindergartens was also reported in many FGDs. How this impacts the possibility of many mothers to look for work was emphasized as well. In our mahalla there is one old kindergarten build in 1970, therefore the building is old and not suitable for operation. This old kindergarten is very small and therefore our children are not admitted to this kindergarten due to overload. We have no kindergarten in our mahalla. If there was, many women could work. Source: Focus group discussions, Qualitative Study 2018. Among children aged 7–14, school attendance increases very significantly. Overall, 95 percent of children do attend an educational institution, and while there are no gender differences in attendance, relatively poor children in the bottom quintile are a bit less likely to attend school than children in the top quintiles (92 percent and 97 percent, respectively). Child labor is not considered a problem in Uzbekistan and early marriage is reported to have declined over the years. In 2006, just 7 percent of women aged 20–24 had married before the age of 18. 2.2 Risks associated with young adults and working-age people The main risks among young adults concern the ability to enter the labor market and, connected to this, to acquire the right level of training and skills. Previous studies (World Bank 2016; Ajwad et al. 2014) have shown how the chances of finding a well-remunerated job are linked to the level of education achieved; very significant improvement is seen when comparing people with a university or tertiary education to those who only achieve a general secondary education. Moreover, the L2CU 2018 survey data confirm that higher education is primarily pursued by people in the top quintile. This can be seen by the percentage of people aged 20–24 who are still 7 attending an educational institution. While the percentage still in education is very low in the first two quintiles (and also among beneficiaries of social assistance2), almost one out of every four young adults in the top quintile is in education (Figure 6). Based on the FGDs education is one of the means to improve people’s living standards, though it was also emphasized that many poor households cannot afford it. In the 2018 L2CU survey, 47 percent of participants agreed with the statement “I am worried about being able to give to my children a good education”; this percentage increased to 54 percent and 61 percent, respectively, among households in the bottom quintile and those receiving social assistance allowances. Figure 6: Percentage of young adults (20–24) attending educational institutions by quintile Source: Analysis of L2CU 2018. Note: SA = social assistance. Unfortunately, the overall relatively low education achievement in higher education is matched by reports that employers have difficulty finding suitably qualified staff with higher education (World Bank 2016, 33). Looking more broadly at the working-age population (15–64) and people’s activity status, very significant differences emerge by age and sex. As already anticipated, while about 70 percent of early youth (aged 15–19) are in education, the percentage drops when considering the group aged 20–24, which also has the highest percentage of people not working. The percentage of people in employment then increases up to the age of 54, whereas between age 55 and 64, over one-half of people are already retired. Looking at differences by sex, the percentage of women employed is much smaller than that of men. Instead, women are more likely to be either “full-time homemakers” or simply not in employment (Table 1). 2Beneficiaries of social assistance include those receiving any of the low-income family allowances, old-age allowance, disability allowance, allowance for loss of the breadwinner, or other in-kind social assistance. 8 Table 1: Economic activity by age group and sex Activity 15-19 20-24 25-34 35-54 55-64 M F T M F T M F T M F T M F T Employee 6 4 5 31 15 23 47 24 36 49 32 40 28 5 16 Other type of job 5 3 4 16 5 11 21 6 13 22 10 16 14 5 10 Not working 16 12 14 37 25 31 27 22 24 20 14 17 13 3 8 In school/ training 69 70 70 10 7 8 1 0 0 0 0 0 0 0 0 Retired 0 0 0 0 1 0 0 0 0 1 2 2 32 74 55 Full time homemaker 2 9 6 5 46 25 2 45 24 2 36 19 4 7 5 Disease / disability 2 1 1 1 2 1 2 2 2 5 6 6 9 5 7 Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 Source: Analysis of L2CU 2018. Note: M = men; F = women; T = total. Numbers refers to percentages. Focusing on the group aged 25–34, Figure 7 shows the main activity status by quintile. In the bottom quintile the percentage of people in work is significantly less than in the top quintile. Furthermore, looking at indirect measures of job quality, not only are people in the bottom quintile less likely to be working, but their jobs are also more precarious (i.e., seasonal, occasional, or temporary), increasing the chance that they will lose their job and income source (Figure 8). Similar results are also obtained for the group aged 35–54. Figure 7: Activity status among adults aged 25–34 by quintile (percentages) Source: Analysis of L2CU 2018. Note: Sum across quintile = 100. 9 Figure 8: Type of main job among adults aged 25–34 by quintile (percentages) Source: Analysis of L2CU 2018. Note: Sum across quintile = 100. While unemployment rates are relatively low (since they require an active search of work), many adults are inactive and discouraged about finding a job, so they simply give up looking for work. The 2013 CALISS (Central Asia’s Labor and Skills Survey) tested cognitive and noncognitive skills among the employed, the inactive, and the discouraged. It found evidence that both inactive and discouraged people have lower cognitive and noncognitive skills than those employed, and that among the discouraged, clear disadvantages also arise in terms of “decision making” (Ajwad et al. 2014). Finally, to confirm that lack of jobs as well as relatively low salaries is one of the major sources of vulnerability, recent findings from the L2CU survey are informative. The survey collected people’s opinion on the priority of improving the supply of jobs and the low salaries and wages in their own community. The priority was ranked from 1 (not an immediate priority) to 4 (an issue of immediate priority). Figure 9 shows the percentage of people reporting whether a certain issue is of immediate priority (i.e., ranked as 4). The majority of people reported the lack of jobs for both men and women as an issue of maximum priority. These statistics were computed for all households, for those in the bottom quintile, and for recipients of social assistance. While percentages are already high for the overall population and not very different among the bottom quintile, beneficiaries of social assistance almost unanimously identified these issues as immediate priorities. Although lack of business opportunities and entrepreneurship is less important among the overall population, it is again of highest priority among social assistance beneficiaries. 10 Figure 9: Issues to be tackled with immediate priority Source: Analysis of L2CU 2018. Note: SA = social assistance. Numbers refer to the percentage of people reporting whether a certain issue is of immediate priority. Nevertheless, during the FGDs people also spoke about the positive effects of the recent reforms and the opportunities provided through microloans given at preferential rates. This came out quite strongly in Fergana, where many felt that the new climate of openness favored more opportunities for trade and, in particular, the development of horticulture. The role of public works was also mentioned in various FGDs; some participants had directly benefited from them, and reported that they had been of help for a few months. However, it was also felt that public works faced more challenges in rural areas, where opportunities are relatively few and only seasonal. 2.3 Risks for the elderly Among the elderly, the main risks are related to the gradual inability to work, increasing medical needs, and lack of care and support from the family. The inability to work is a problem if a person was unable to save during his/her lifetime, does not have a contributory pension, or the amount received is very low. In Uzbekistan, people of pension age have almost universal coverage of social transfers, however, and family support is generally strong, so it is very rare for elderly people to be left alone and without care. On the contrary, there is evidence that the pension received by elderly members of the household often also supports the needs of younger family members, and pensions are an important source of income across all quintiles. 2.4 Shocks and risks across the lifecycle Disability and health shocks are usually critical aspects that can affect people’s vulnerability. Figure 10 shows both the prevalence of disability3 and the use of medical 3 A person is considered disabled if s/he reports a lot of difficulty or inability in any of the following main functions: seeing (even if wearing glasses), hearing (even if using a hearing aid), walking, remembering, self- care activities such as dressing or washing, or communicating. 11 assistance (in the last 30 days before the interview) by age group. In general, both disability and use of medical assistance increase with age, although use of medical assistance is also particularly high for children under seven. No significant differences arise when the data are disaggregated by sex; and no trend in disability emerges between quintiles. Instead, use of health care tends to be considerably higher among the top quintiles. Figure 10: Disability and need for health assistance by age group (percentage) Source: Analysis of L2CU 2018. Nevertheless, both disability and health care provision require significant extra expenditure, which for some households can be impoverishing and represents a very significant source of vulnerability. To look at this in more detail, the size and share of out- of-pocket (OOP) health expenditure over total consumption expenditure is analyzed using the methodology proposed by the WHO European Regional Office (Thomson et al. 2016). The WHO approach argues that it is important to take into account that for households with consumption expenditure below a basic needs line, health expenditure is catastrophic independent of the relative percentage of budget expenditure. Moreover, the standard definition of capacity to pay, based on the level of expenditure excluding food expenses, does not take into account other essential needs that are equally important. Finally, it is important to consider households that do not spend anything on health. While it is not possible to know whether this is because people forego health care or simply do not need it, it is still useful to analyze changes in these percentages across quintiles. Based on this different approach, the international poverty line of US$3.2 in purchasing power parity (PPP) is identified as a normative basic needs expenditure. The population is then divided into five mutually exclusive groups: • households that do not incur any OOP health expenditure; • households that are not at risk of impoverishment because of health expenditure (their level of expenditure after OOP health expenditure is more than 1.2 times the basic needs line); 12 • households that are at risk of impoverishment because their health expenditure (after OOP) is close to the basic needs line (between 1.0 and 1.2 times); • households that are impoverished because of health expenditure, i.e., they have an expenditure below the basic needs line (after OOP); and • households that are further impoverished because they have a level of expenditure below the basic needs line and OOP health expenditure further reduces their level of expenditure on other essential items. The results of this categorization are reported in Figure 11 by quintile. Overall, 56 percent of households do not report any health expenditure, but for those that have expenditure this comes at a risk of impoverishment or further impoverishment for 7 percent of households. In the lowest quintile, 72 percent of households are without OOP health expenditure, a figure that reduces to 46 percent in the top quintile. Considering that just about 10 percent of people have an overall consumption expenditure below the international line of US$3.2 PPP, it is significant that in the second quintile there are also people who are impoverished because of health expenditure. Figure 11: Share of households at risk of impoverishment because of health expenditure by quintile Source: Analysis of L2CU 2018. It is useful to consider different measures of catastrophic health expenditure. The literature provides different ways of defining catastrophic health expenditure. Health expenditure is considered catastrophic when it represents more than 10 percent or 25 percent of the overall household expenditure, or 40 percent of nonfood consumption expenditure. However, based on the above classification an alternative approach is to consider catastrophic health expenditure when it represents more than 40 percent of consumption expenditure after subtracting a normative amount that captures essential expenses (in this case, the US$3.2 PPP per day). Furthermore, under this approach households also experience catastrophic expenses whenever they are impoverished or further impoverished by health expenditure. The effect can easily be seen in Table 2. While for the more traditional measure (health expenditure is higher than 25 percent of the total consumption expenditure) the percentage of households incurring catastrophic health expenditure increases moving toward higher consumption quintiles, looking at expenditure higher than 40 percent of the capacity to pay, poorer households are more 13 likely to incur catastrophic health expenditure. Finally, Table 2 shows the percentage of households where at least one member in the 30 days preceding the interview had to forgo health treatment because it was considered too expensive; once again, this is more likely to happen in the bottom quintile (8 percent) than in the top one (3 percent). The above statistics were also computed for different population subgroups: recipients of social assistance; households with at least one disabled member; and households with at least one elderly member (65+). All such households do incur health expenses more than the average household, and households in receipt of social assistance or with at least one disabled member have higher-than-usual catastrophic health expenditure as well as a significantly higher percentage of cases where health treatment was forgone due to cost. Table 2: Catastrophic health expenditure by population subgroup % of hhs % of hhs % of health exp. % not seeking with health Average with health over capacity to healthcare exp. health exp. exp. > 25% pay > 40% because of cost Consumption quintiles Bottom 20 28.1 3.4 3.8 22.6 8.0 Q2 37.9 5.1 7.0 14.8 5.6 Q3 44.5 7.3 10.6 13.2 5.0 Q4 52.5 9.1 15.5 14.2 6.6 Top 20 54.5 13.0 20.4 17.0 3.3 Goups of interest Recipient of SA 55.2 10.2 16.3 27.3 11.4 With at least 1 disabled 66.9 16.1 24.9 33.0 14.1 With at least 1 elderly (65+) 53.6 10.3 15.3 20.6 5.7 Overall 44.3 7.9 12.0 16.2 5.6 Source: Analysis of L2CU 2018. Note: SA = social assistance; HH = household. Expenditure linked to health care needs and disability was also strongly emphasized in FGDs as a fundamental source of vulnerability affecting people because they result in inability to work, either for the person involved or for the caregiver. Therefore, all FGDs emphasized how “good health” is a critical aspect enabling people to reduce their vulnerability. The quotes below show how health is often a concern and how large expenses are dealt with through some ad hoc arrangements and solutions, which unfortunately are not always guaranteed. “In winter people are afraid of children being sick with flu as it will be necessary to go to pharmacy for medicines.” “A person asked for 40 mln UZS for the medical treatment in Tashkent. Then, special bank account was opened where benefactors made their donations. Khokimiyat also donated 1,2 mln UZS”. “If a person needs specific medical treatment, mahalla tries to find specialists in the nearest areas or in Tashkent. They may help with transportation of a patient as well.” Source: Focus group discussions and key informant interviews, Qualitative Study 2018. 14 Besides health, people seem to be affected by Uzbekistan’s changing macroeconomic conditions, particularly its relatively high inflation. Indeed, 80 percent of households agreed with the statement that “prices for the things I buy are rising too quickly” , and among recipients of social assistance allowances, this percentage increases to 86 percent. This should be put in a context where food consumption represents about two- thirds of the whole consumption expenditure for those in the bottom quintile and where the cost of housing and utilities is the second most important subaggregate, representing 12 percent of the whole consumption expenditure. The budget shares of both food and utilities decline in the better-off quintiles, suggesting that these are necessary goods whose price increases are likely to affect other important expenditures, such as on health or education. Finally, some regions of the country are badly affected by environmental problems and the likely effects of climate change. In the Republic of Karakalpakstan’s FGDs, specific challenges were mentioned regarding environmental problems in the Aral Sea, such as “salt storms,” drought, and lack of water, all affecting agricultural crops and livelihoods. In both Karakalpakstan and Khaskadarya people reported increasingly extreme weather conditions, with hotter summers and colder winters affecting livestock and horticulture. Finally, natural disasters and calamities have emerged as a result of soil erosion and mudslides, particularly in Kashkadarya and Samarkand regions. 15 3 Who is vulnerable and why? The previous section identified the main risks faced by people in Uzbekistan and while it is intuitive that risks affect and influence poverty and vulnerability, the qualitative study, especially through the FGDs, also directly asked questions aimed at understanding what it means to be poor in Uzbekistan and the causes of poverty. FGDs tended to classify households in their communities into three groups: high-, middle- , and low-income households. But in relatively better-off regions, namely Tashkent and Fergana, people identified five different categories, including at the two extremes an extra category: destitute and very rich households. Poor households are those with one or more of the following characteristics: • working-age members are jobless and rely on casual and temporary jobs; • household members have a disability or chronic illness, especially when the disability involves constant care, whereby in addition to the costs incurred for the disability a household member is also limited in his/her opportunity to find work; • households have many dependents and especially young children, precluding mothers from finding work; • households do not have land/gardens and livestock and so need to purchase their food in the market and also lack the opportunity to earn extra income; • households have a single earner and a relatively low salary or pension in relation to household needs; • households live in small accommodations without sufficient living space or without access to water or gas supply (specifically mentioned in Karakalpakstan, where one location visited had such problems). Moreover, low-income households were described as those that struggle to pay utilities bills on time and must compromise on their needs – they definitely cannot afford to buy a car or good home appliances, and cannot afford to buy good clothes or support children’s education as they would like. However, problems of hunger and malnutrition were not mentioned as a significant issue. People also singled out the “undeserving poor,” who are unemployed because they are alcoholics or drug addicts or simply labelled as ”lazy.” This category was especially mentioned during interviews with KIIs, but in the majority of cases the “undeserving poor” was considered to be a relatively small group. On the contrary, high-income and rich households are identified as those that have good jobs, own cars and good houses, and have political and social connections. Moreover, such households have good assets, such as large and fertile land plots or successful businesses. Middle-income households are defined as those households that are able to get by without asking for help, have regular jobs or their children are already capable of working, do not have debt, own their dwelling, and can afford to buy some home appliances. Everywhere, good health was considered a precondition to enable people to succeed in their lives. Similarly, obtaining a good education emerged as one of the key factors that could enable people to improve their circumstances. 16 Poor households were considered vulnerable, but it was also recognized that middle- income households could fall into poverty because of risks and shocks. On the contrary, better-off households were seen as usually having the means to overcome such problems. In all FGDs the main cause of poverty and vulnerability was considered the absence of work. This could be the result of inadequate education and skills, lack of good health, or lack of assets (land/livestock/business activity), or in general the few available job opportunities. This was especially the case in rural areas. Moreover, people distinguished between long-term loss of job due to chronic health conditions/disability and situations where the impossibility of working was likely to be temporary due to short-term illness or pregnancy/childbirth and looking after young children. Another specific element of vulnerability arises in areas where there is no work in the winter months. This occurs in rural areas and although people try to save some summer earnings for the winter months, unexpected setbacks often emerge Moreover, a long period of unemployment is in itself considered a cause of vulnerability because it tends to establish a vicious circle whereby people are forced to accept jobs in poor conditions and with low pay. Households are also considered vulnerable when they are unable to protect themselves against different type of shocks, such as high inflation, the loss of their job, or climatic challenges resulting in extreme weather (very hot summers and cold winters). 17 4 Coping strategies The FGDs highlighted both short- and long-term strategies to deal with risks. Short-term coping strategies include: borrowing from relatives or friends; selling assets; and searching for temporary jobs (seasonal and public works or travelling for earnings). In some cases employed people reported asking their employers to pay part of their wages in advance. Moreover, especially in rural areas, low-income families and the unemployed apply to the mahalla committee for support and for social allowances. On the contrary, in some urban areas and especially in Tashkent, the mahalla committee is not considered a body that can provide assistance; people prefer to go directly to district-level institutions (such as the Pension Fund or the Employment Support Center) or to ask for support from nongovernmental organizations (NGOs), churches, and other charitable institutions. However, outside Tashkent it was difficult to find any evidence of NGOs providing direct assistance. Mid- and long-term strategies include investing in children and relying on their earnings when they grow up, especially through work abroad. Remittances and loans at preferential rates were considered to provide the opportunity to invest in small businesses and to acquire useful assets (such as sewing machines, small equipment and tools, or livestock) or to improve the profitability of their agricultural activities (for example, through construction of greenhouses). Looking at income sources by quintile reveals the relevant importance of the different coping strategies (Figure 12). Income from wages represents the most important source of income across all quintiles, but the second most important source is contributory government transfers (primarily pensions). However, notably for the bottom quintile, very relevant sources of income are also noncontributory government transfers (different types of social assistance) and remittances from abroad, confirming the findings of the qualitative study. Figure 12: Composition of income by source and quintile Source: Analysis of L2CU 2018. 18 4.1 Migration In Uzbekistan 16 percent of households were found to have at least one member who migrated to work abroad, but the incidence is almost one out of four in the bottom quintile (Figure 13). These very high percentages certainly reflect the inability of people to find work. However, when such figures are compared with similar estimates in 2013 (based on the CALISS), the phenomenon appears to have halved – in 2013 the percentage of households with at least one migrant was 33 percent. However, such estimates must be taken with caution since it is unclear whether the two surveys are fully comparable. Other sources do seem to corroborate such findings. For example, World Bank estimates based on the balance of payments show that personal remittances declined from 11.6 percent of gross domestic product (GDP) to 3.7 percent of GDP in 2016. Moreover, though estimates are not up to date, net migration rates seem to be declining in recent years, but remain negative. Various FGDs reported that migration is a fundamental coping strategy, but while migration was said to have declined in Fergana, migration was very widespread in Karakalpakstan and Kashkadarya, especially in rural areas. However, people reported that migration to Russia has become more risky and complex. Figure 13: Percentage of households with at least one member who migrated abroad by quintile, 2013 and 2018 Source: Analysis of L2CU 2018 and CALISS 2013. The very large majority of migrants are men and are relatively young; the bulk of them are between 20 and 34 years old (Figure 14). Finally, while the percentage of such migrants is significantly higher among the bottom quintile than in better-off groups, the level of migration among recipients of social assistance is lower than the national average. This could suggest that for various reasons migration among households receiving social assistance is more difficult. 19 Figure 14: Percentage of population 15+ working abroad by age group, disaggregated by quintile and by sex Source: Analysis of L2CU 2018. 20 5 Policies to reduce vulnerability Reduction of vulnerability comes with general improvement in a country’s economic conditions, and evidence suggests that Uzbekistan’s situation has improved in recent years. In all FGDs people reported that circumstances are greatly improved compared to 10 years ago, due to investment in infrastructure (roads, electricity, water). But when people were asked to compare the current situation to that of three years ago, the response was mixed. Some FGDs considered the situation to have improved, but others considered it stable or even more difficult, mentioning problems due to high inflation, environmental changes, and difficulties in accessing social assistance. In the L2CU survey, only 36 percent of people agreed with the statement “My family’s financial situation is worse now than it was two years ago.” This means that the majority saw an improvement. However, higher percentages of some subgroups of the population agreed with this statement. In particular, 52 percent within the bottom quintile agreed, as did 59 percent of social assistance recipients. Such households might have suffered specific shocks that worsened their conditions and it is symptomatic of the need to better protect such groups (Figure 15). Figure 15: Percentage who said their family situation now is worse than two years ago Source: Analysis of L2CU 2018. Note: SA = social assistance. During the FGDs people suggested possible interventions to reduce vulnerability. These included the following: • Improve job opportunities, especially in rural areas, through: new investment in enterprises and infrastructure; better microloan schemes at preferential rates for low- income households, accompanied by some informal training and coaching once loans have been taken out; and job fairs where people can learn from the experiences of successful entrepreneurs. • Improve social protection measures by simplifying the administration process, ensuring that amounts of support keep pace with inflation; account for specific challenges faced by people due to environmental and climatic changes; and support people with low incomes when they face high health expenditure, especially certain groups in receipt of social assistance or persons with disabilities. • Increase the availability of kindergartens and ensure that access is free of charge for low-income families. 21 • Increase the quality of education at vocational and tertiary level by: improving the way higher level education meets the demands of the labor market; increasing the percentage of students in high level degrees; and addressing the current constraint of lack of teachers with a higher degree. All of these interventions could address the risks seen affecting the Uzbek population. In particular, as the government is planning further reforms that may generate macroeconomic instability, the capacity to ensure that social assistance is responsive and able to reach people in need is very critical. 22 6 References Ajwad, M. I., et al. 2014. The Skills Road : Skills for Employability in Uzbekistan. Washington, D.C.: World Bank. Development Initiatives. 2017. Global Nutrition Report 2017: Nourishing the SDGs. Bristol: Development initiatives. Thomson, S., T. Evetovits, J. Cylus, and M. Jakab. 2016 (September). “Monitoring Financial Protection to Assess Progress Towards Universal Health Coverage in Europe.” Public Health Panorama 2(3): 357–366. World Bank. 2016. Systematic Country Diagnostic for Uzbekistan. Washington, D.C.: World Bank. 23