Report No. 20285-TJ Republic of Tajikistan Poverty Assessment June 29, 2000 Human Development Sector Unit Europe and Central Asia Region Document of the World Bank Republic of Tajikistan POVERTY ASSESSMENT Acronyms ABA American Bar Association ADB Asian Development Bank AKDN Aga Khan Development Network CCP Cash Compensation Program EPI Expanded Program of Immunization ESAF Enhanced Structural Adjustment Facility FAP Feldscher Accuscherske Punkt FPSP Farm Privatization and Support P'roject FSU Former Soviet Union GAA German Agro-Action GAI State Automobile Inspectorate GBAO Gorno-Badakhshan Autonomous Oblast GDP Gross Domestic Product GNP Gross National Product IDPs Internally Displaced People ILO International Labor Organization IMF International Monetary Fund LDFA Leninsky District Farner's Association ME Ministry of Education MF Ministry of Finance MH Ministry of Health MSP Ministry of Social Protection MVD Ministry of Intemal Affairs NGO Non-Government Organization ORA Orphans, Refugees and Aid PPP Purchasing Power Parity. PTA Parent-Teacher Association RRS Regions of Republican Subordination SAC Structural Adjustment Credit SCF/US Save The Children (US) SSA State Statistical Agency SPF Social Protection Fund SVA Rural Outpatient Clinic TASIF Tajikistan Social Investment Fund TLSS Tajikistan Living Standards Sunrey TR Tajikistan Ruble UNDP United Nations Development Program WFP World Food Program Vice President: Johannes Linn Country Director: Kiyoshi Kodera Sector Manager: Michal Rutkowski -Task Manager: Michael Mills Republic of Tajikistan POVERTY ASSESSMENT Contents Preface .....................................................1 Summary and Conclusion .....................................................2 Introduction .................................................... 14 Chapter 1: A Profile of Poverty .................................................... 16 A. Measuring Poverty in Tajikistan .................................................... 16 B. The Level and Depth of Poverty .................................................... 17 C. Identifying the Poorest .................................................... 20 D. Coping Strategies .................................................... 31 E. Food Security .................................................... 31 F. The Development of a Pro-Poor Country Development Strategy ............................... 33 Chapter II. Employment and Poverty .................................................... 35 A. The Main Features of the Labor Market in Tajikistan ................................................. 35 B. Changes in the Labor Market 1991-1998 .................................................... 42 C. Labor, Wages and Poverty .................................................... 44 D. The Main Labor Market Issues .................................................... 47 E. The Development of a Pro-Poor Employment Strategy .............................................. 47 Chapter III. Agriculture and Poverty .................................................... 52 A. The Main Features of Agriculture in Tajikistan .................................................... 52 B. The Profile of Rural Poverty .................................................... 54 C. Land Reform .................................................... 58 D. Cotton .................................................... 63 E. Irrigation .................................................... 64 F. Finance and Credit .................................................... 67 G. The Development of a Pro-Poor Agricultural Strategy ............................................... 68 Chapter IV: Education and Poverty .................................................... 72 A. The Main Features of the Education Sector .................................................... 72 B. Public and Private Spending on Education .................................................... 74 C. School Enrolments and Poverty .................................................... 79 D. Poverty and Reasons for Absence at School .................................................... 83 E. Technical and Higher Education .................................................... 85 F. The Development of a Pro-Poor Education Strategy ................................................... 87 Chapter V: Health and Poverty .................................................... 91 A. Health and Nutritional Status .................................................... 91 B. Main Features of the Health Sector .................................................... 95 C. Financing of Health Care .................................................... 96 D. Access to Health Care .................................................... 99 E. Utilisation of Health Care Services .................................................... 100 - i - F. The Restructuring of the Health Services ................................................................ 102 G. The Development of a Pro-Poor Health Strategy ...................................................... 103 Chapter VI: Social Protection and Poverty ................................................................ 106 A. The Main Features of the Social Protection System .................................................. 106 B. The Current Systems of Cash Benefits ................................................................ 108 C. An Assessment of the Current Cash Benefit Systems ............................................... 110 D. Options for the Reforn of Social Assistance ............................................................. 114 E. The Development of a Reforrned System of Social Protection ................................. 118 Chapter VII: Governance and Poverty .............................................. 121 A. The Political and Economic Context to Governance ................................................. 121 B. The Impact of Improper Governance on Poverty ...................................................... 122 C. Ongoing Efforts to Improve Governance ................................................................ 126 D. "Top-Down" Measures to Address Outstanding Problems ....................................... 128 E. "Bottom-Up" Measures to Address Outstanding Problems ....................................... 130 F. The Development of an Improved Governance Strategy ........................................... 133 Chapter VIII: Looking Forward ................................................................ 135 Appendix 1: The Tajikistan Living Starmdards Survey Methodology ................................... 141 Appendix 2: The Sensitivity of the Poverty Profile to Alternative Assumptions of Equivalence Scales ................................................................ 147 List of Tables and Figures Summary Table: Poverty Measures in Tajikistan ................................................................. 3 Table 1: Poverty and Human Development in USSR (1989) .................................................. 14 Table 2: Summary Macro-Economic Indicators, 1991-1999 .................................................. 15 Table 3: Poverty Measures in Tajikistan ................................................................ 18 Table 4: Distribution of Income and Expenditure in Selected Countries ................................ 19 Table 5: Poverty Incidence amongst Individuals by Location ................................................ 20 Table 6: Composition of the Poorest and Richest Quintiles by Location ............................... 21 Table 7: Poverty Incidence amongst Individuals by Demographic Group ............................. 23 Table 8: Poverty Incidence by Education and Labor Market Status . ..................................... 28 Table 9: Structure of Total Household Income by Quintile Group . ...................................... 28 Table 10: Correlates of Household Poverty ................................................................ 30 Table 11: Proportion of Households Reporting Engaging in Selected Coping Strategies ...... 32 in the Last 6 Months ................................................................. 32 Table 12: Household Food Security and Coping Strategies by Quintile ................................ 33 Table 13: The Utilization of Labor Resources ................................................................ 36 Table 14: Labor Force Participation Rates by Age Group and Gender .................................. 37 Table 15: Labor Force Participation Rates by Settlement and Age Group ............................. 37 Table 16: Reasons for Non-Participation by Age Group and Settlement ............................... 38 Table 17: Labor Force Status by Educational Level ................................. .............................. 38 Table 18: Employment by Sector ................................................................ 39 Table 19: Employment by Occupation ................................................................ 39 - ii - Table 20: Private Undertakings and Informal Trading by Economic Sector .......................... 40 Table 21: Labor Force Participation Ratios 1991-1998 (Registered) by Gender ................... 42 Table 22: Registered Unemployment 1992-1998 ............................................................ 43 Table 23: Reasons for Changing Jobs and/or Labor Force Status 1997-1999 ........................ 43 Table 24: Average Monthly Wage per Household by Settlement .......................................... 44 Table 25: Share of the Employed Affected by Wage Arrears by Quintile ............................. 45 Table 26: Ownership of Benefits in Kind ............................................................. 46 Table 27: Average Nominal Wage, Minimum Wage and Subsistence Minimum ................. 46 in Tajik Rubles and US dollars, 1992-1998 ............................................................ 46 Table 28: Trends in Crop Area, Yield and Production since 1991 ......................................... 53 Table 29: Agricultural Sector Reforms under the ESAF, 1999-2000 ..................................... 54 Table 30: Rural-Urban Distribution of Households by Regions ............................................ 54 Table 31: Distribution of Individuals in Households by Rural and Urban ............................. 55 Table 32: The Age Composition of Population by Rural and Urban ..................................... 55 Table 33: Educational Status by Rural and Urban ............................................................ 55 Table 34: Participation Rates and Land Available for Rural Households .............................. 56 Table 35: Contribution of Own Farn ing to Household Income ............................................. 56 Table 36: Ownership of Livestock and Other Assets ............................................................ 57 Table 37: Land Availability per Capita by Quintile for Rural Households ............................ 57 Table 38: Numbers of Farms and Estimated Arable Areas (June 1999) ............................... 62 Table 39: Educational Facilities and Students, 1998 ............................................................ 73 Table 40: Actual Education Spending by the State Budget, 1992-1999 ................................. 75 Table 41: Education Budget Plan and Execution, 1998-2000 ................................................ 75 Table 42: Actual Education Expenditure by Item (% of total), 1997 ..................................... 76 Table 43: Actual Education Expenditures (as % of Plan), 1994-1998 ................................... 76 Table 44: Planned Education Expenditures, 1997-1998 ......................................................... 77 Table 45: Local Education Expenditure Per Capita, 1994-1999 ............................................ 77 Table 46: Net Enrollment Rates in Basic Education by Quintile ........................................... 80 Table 47: Net Enrollment in Basic Education by Region and Quintile .................................. 81 Table 48: Net Enrollment in Basic Education by Settlement and Region .............................. 83 Table 49: Reasons for Students' Absence in Basic Education by Region .............................. 84 Table 50: Reasons for Students' Absence in Basic Education by Quintile ............................ 84 Table 51: Attendance Rates for Post-Basic Education by Level of Education ....................... 85 for Heads of Household ............................................................ 85 Table 52: Net Enrollment in Basic Education by Education of Household Head ................... 87 Table 53: Official Health Indicators ............................................................ 91 Table 54: Infectious Disease Incidence per 100,000 ............................................................ 92 Table 55: Tuberculosis Cases ............................................................ 92 Table 56: Condition for which Medical Assistance was Sought by Age and Gender ............ 93 Table 57: The Prevalence of Breast Feeding ............................................................ 93 Table 58: The Prevalence of Acute and Chronic Malnutrition ............................................... 94 In Children under 5 Years ............................................................ 94 Table 59: Primary Health Care Facilities ............................................................ 95 Table 60: Health Personnel per 10,000 Population ............................................................ 95 Table 61: Health Personnel Salaries ............................................................ 96 Table 62: Actual Health Spending by the State Budget, 1992-1999 . . 96 Table 63: Health Budget Plan and Execution, 1998-2000 ...................................................... 97 Table 64: Planned Health Expenditure by Item (% of total) .................................................. 97 - ijI - Table 65: Planned Local Health Expenditure by Institution (% of total), 1994-1998 ....... .... 97 Table 66: Local Health Expenditure Per Capita, 1994-1999 .................................................. 98 Table 67: Self Reported Morbidity by Quintile of Per Capita Househol,d Expenditure ......... 99 Table 68: Reasons Why People Did Not Seek Medical Assistance by Qluintile .................... 99 Table 69: Health Care Access ................................................................100 Table 70: Distribution of Health Care Utilization and Access by Quintile .......................... 101 Table 71: Degree of Difficulty in Paying for Health Care .................................................... 101 Table 72: Place of Birth of Last Child ............................................................... 102 Table 73: Average Length of Stay in Hospital ............................................................... 103 Table 74: Estimated Eligible and Actual Beneficiaries of the CCP (April 1998) ................ 109 Table 75: Pensioners in 1993-1998 (thousands) ............................................................... 110 Table 76: Households Eligible for, and in Receipt of, Social Protection Benefits ............... 111 Table 77: Pension Eligibility and Payment by Region ......................................................... 112 Table 78: Reasons for Not Receiving a Benefit given by Eligible Households ................... 113 Table 79: Average Value of Pensions and Salaries .............................................................. 114 Table 80: Summary of Targeting Options ............................................................... 118 Table 81: Budgetary Funds Available to Support the Poor .................................................. 119 Table 82: Existing NGOs Working in the Social Sectors in Tajikistan ................................ 131 Table 83: Govermnent Spending on Social Sectors 1999-2000 ........................................... 136 Table 84: The Effect of Economic Growth on Poverty Reduction by Year 2005 ................ 137 Table 85: Pssible Poverty Reduction Indicators and the Situation in 199,9/2000 ................. 137 Table A2.1 The Risk of Household Poverty by Household Size . ............................ 148 Table A2.2 Composition of Poor Households by Household Size . .......................... 148 Table A2.3 The Risk of Individuals Living in Poor Households by Age . ..................... 149 Table A2.4 Composition of Individuals Living in Poor Households by Age . .................. 149 Table A2.5 The Risk of Household Poverty by Type of Settlement . ......................... 150 Table A2.6 The Risk of Household Poverty by Region .................................... 150 Figure 1: Cumulative Distribution of Per Capita Household Income and Expenditures ........ 19 Figure 2: Proportion of Individuals who are Poor by Region ................................................. 22 Figure 3: Proportion of Individuals who are Poor by Age Group .......................................... 25 Figure 4. Proportion of Individuals who are Poor by Household Composition ..................... 26 Figure 5: Food Security ............................................................... 32 Figure 6: Median Household per Capita Net Education Expenditures ................................... 78 (as percent of Total Household per Capita Expenditures) ....................................................... 78 Figure 7: Gender Gap in Net Enrollments: Urban Areas ........................................................ 82 Figure 8: Gender Gap in Net Enrollments: Rural Areas ......................................................... 83 Figure 9: Educational Attainment by Poverty Quintile .......................................................... 86 Figure 10: Median months of arrears by quintile ............................................................... 112 - iv - PREFACE This poverty assessment was carried out in November 1999. Much of the analysis is based on the results of the Tajikistan Living Standards Survey which was undertaken earlier in the year. The work was also based on a considerable amount of complementary data collected through other surveys, social assessments and statistics from a wide variety of sources. The World Bank team was led by Michael Mills (Principal Economist), and included Chris De Neubourg (labor market), Cevdet Denizer (economist), Annelies Drost (education), Jane Falkingham (poverty profile and social protection), Isabel Hemming (health), Taies Nezam (social development), Ray Purcell (agriculture) and Michael Thurman (governance). The initial planning work for the task was undertaken by Geremia Palomba (Economist). Team assistance was provided by Carmen Laurente. The poverty assessment was carried out under the overall guidance of Maureen Lewis (Sector Leader for Economics in the Human Development Department of the Europe and Central Asia Region) and Annette Dixon (Director, Human Development Department of the Europe and Central Asia Region). The mission worked with three particular teams of national experts. The first team was formally established as a Working Group on Poverty and Social Protection by the Prime Minister, and was headed by Mr. B. T. Turaev (Deputy Chairman, State Statistical Agency). Its members included: Messrs/Mesdames Valiev D. A.; Alifbekov Z. A.; Mirzoeva A. D.; Alimov A. L.; Djononov S.; Sharipova E. M.; Nosirov E. M.; Shohmakhmadova S.; and Sharipov Z. The second team, with which the mission worked, was the SPRITE ("Social Protection in Transition Economies") team headed by Professor Komilov. The SPRITE team organized two seminars on poverty for a large number of officials and national experts. The first seminar was held in November, 1999, when the poverty assessment mission was undertaken; and the second seminar took place in May 2000, after the draft report had been prepared. At these seminars, there were major presentations on poverty issues by selected national experts and also by World Bank team members, as well as substantial small-group discussions. Following the first seminar, the Prime Ministerial Working Group on Poverty and Social Protection compiled a first draft of a National Program to Alleviate Poverty and Develop Social Protection Mechanisms in Tajikistan. In March 2000, the President established another team, headed by Mr. G. Babayev (State Economic Advisor) to produce an Interim Poverty Reduction Strategy Paper. This team also participated prominently in the May 200 seminar. In February 2000, the World Bank also organized a country economic mission. The purpose of that work was to focus on the macroeconomic framework, an analysis of growth possibilities, the factors hindering a supply response from the reform measures already put in place, and the restructuring needs of the agricultural and industrial sectors in particular. The poverty assessment report and the country economic report complement each other closely. - 1 - SUMMARY AND CONCLUSION 1. Poverty is the central development issue in Tajikistan. The country is now one of the poorest in the world, with an estimated per capita GNP of no more than $330 (1997). Poverty is not new in Tajikistan, as the country came to independence with widespread material poverty. That was made worse by the civil conflict in the early 1990s. However, with the cessation of violence in the country, with economic growth having resumed since 1997, and with reforms now being made to restructure the economy, it should be possible to make significant progress in poverty reduction during the coming years. 2. This report suggest that there should be four key pillars to a. poverty reduction program strategy: (i) the stimulation of high labor productivity and economic growth; (ii) the provision of basic social services; (iii) the targeting of assistance to the poorest; and (iv) the development of a well-governed and safe environment. Ultimately, economic growth is the main mechanism to ensure that future generations in Tajikistan will be able to find gainful and highly productive employment and to bring prosperity to all; and so action on the labor market and on agricultural and rural development policies form the first of the proposed four key pillars for a poverty reduction strategy. However, poor people also need to have sufficient human capital so that they can take advantage of existing and new opportunities. Access to good quality basic schooling and health services is not only a human right, it is also essential for poverty alleviation. The third key pillar of the proposed poverty alleviation strategy concerns the targeting of assistance to the poorest, that is, those people who have special difficulties or needs and for whom targeted assistance is required. An efficient and effective social safety net is also required to facilitate the transition process and to obtain widespread public support for some of the difficult reform decisions (such as public sector retrenchment) which need to be taken. The fourth pillar of the proposed poverty alleviation strategy is the development of a well-governed and safe enviromnent in which people can live and look after themselves. A profile of poverty. 3. To develop an appropriate strategy for poverty reduction, a clear understanding of the nature of poverty in the country is essential. Within Tajikistan itself, poverty is viewed as a multi-dimensional phenomenon. When residents are asked to define it, they answer lack of clothing, food shortages, inadequate salaries, low pensions, lack of land or livestock and a range of other indicators. In order to obtain a firm quantitative foundation for assessing the dimensions of poverty, the Tajikistan Living Standards Survey (TLSS) collected data on demographics, housing, expenditures, consumption and other areas. Clearly, the extent of poverty depends on the choice of poverty line. However, whatever measure is chosen, it is clear that the level of material poverty in Tajikistan is high, with four out of five 'poor'; a third 'very poor'; and nearly 20 percent 'extremely poor' (below $1.075 PPP). Over half of all households had not eaten any meat in the 7 days prior to the survey, 61 percent had not eaten eggs and 85 percent had not eaten confectionery. Three-quarters of the population are very concerned about how they will provide for basic necessities in the next 12 months. -2- Summary Table: Poverty Measures in Tajikistan (a) Proportion of population below minimum consumption line 96% (b) Proportion of population poor (below State Statistical Agency poverty line of 83% TR20,000) (c) Proportion of population very poor (below 50% of State Statistical Agency 33% poverty line) (d) Proportion of population extremely poor (below $1 .075 PPP a day) 17% Source: TLSS 4. The data from the TLSS may be compared with data from other sources. The Tajik Social Investment Fund (TASIF), Government agencies, donor agencies (UNDP and the Asian Development Bank) and international NGOs have all completed various social assessments of Tajikistan. According to the TASIF, some 60 percent of the population consider themselves poor. While the food security situation is acute, there is no evidence of widespread severe malnutrition. Nevertheless, a series of anthropometric and food security surveys over the last six years clearly indicate that there is a significant amount of undernutrition and that there was a steady decline in the nutritional status of children through the 1990's. The situation was especially serious in some regions in the period 1995-1997. With the upturn in economic growth since 1997, it is possible (but not yet certain) that the situation may have improved somewhat since then. Together the data from all of these sources provide a picture of the extent of poverty, the geographical and demographic patterns of poverty, poverty and the labor market, social services and expenditures, and infrastructure and assets. 5. The distributions of both incomes and expenditures within Tajikistan are very heavily skewed, with households highly concentrated in a narrow range of values in the bottom four- fifths, with the result that headcount measures of poverty are very sensitive to even small changes in the value of an absolute poverty line. A strategy for reducing poverty should focus on those families and individuals who are at the bottom of the distribution, and also on the most vulnerable segments of society. Households with many children experience a higher risk of living in poverty than adults do, and the situation is undoubtedly very serious for street children and institutionalized children. Poverty rates among the very old (age 75 or older) are also high and are highest among single and disabled pensioners. As expected, data show that the risk of being poor is inversely related to education. Persons with no education were more than two and a half times as likely to be living in a household in the bottom quintile as individuals with higher education. According to the TLSS, less than half of all households have access to piped water; nearly a quarter are reliant on water from rivers or lakes or ponds, 75 percent of households have no source of hot water, 85 percent rely on an outside latrine, two thirds of households heat their home for no more than four months, and only 14 percent of households have a phone inside the home. 6. The findings of the TLSS further demonstrate that substantial proportions of households are food insecure. Food was the principal household expenditure during the month preceding the interview, and one third of households had to incur debt in order to buy - 3 - food, according to the TLSS. In terms of coping strategies, 72 percent of households reported consuming food grown by the household. But financial hardship is not limited to the poorest group. Even among well-off households, nearly 30 percent reported eating smaller portions and reducing the number of meals a day. About 25 percent thought they would have to sell household assets, and more than 25 percent would have to borrow to make ends meet. Of all households eating less than two meals a day, more than 50 percent reported that they would have to reduce the number of meals even more. Employment and poverty. 7. According to the TLSS, 56 percent of the population participate in the labor force, a relatively low number compared with other republics of the FSU. Some 90 percent of men aged 30 to 54 belong to the labor force. Older residents, many of whom work in agriculture, are continuing to work; 30 percent of men and 15 percent of women are active after age 60. A large number of people are not in the labor force for traditional reasons such as school enrollment, pensions, disability, and home care. Yet a disturbingly large- number of young people are unemployed and are no longer seeking gainful employment. Gender differences in school attendance above the age of 16 are also large, especially for individuals between 20 and 24, with women far less enrolled in schools. Women also stay out of the labor force, even at young ages, for reasons related to family responsibilities. 8. The formal part of the economy still remains state-dominated and many people work in the public sector, especially in state companies and state farms. However, the state sector is shrinking due to lack of resources and it cannot employ more people. The old labor absorption mechanisms of the Soviet system no longer work and new mar]ket mechanisms are too weak to absorb the excess supply. Service jobs make up the second most important sector. But the private sector is still in its infancy. The majority of private activities are concentrated in informal family businesses and street trade. Officially, the unemployment rate is 3 percent, but a large number of people are among the "hidden" unemployed, meaning that they are formally employed but are not working because they are either on forced holidays or are working part-time involuntarily. When these hidden emaployed people are included in the unemployment rate, the level may be as high as 33 perc ent. The average wage is no more than one third of the subsistence minimum. 9. To alleviate poverty, the functioning of the labor market needs to be improved. Since labor is the only source of wealth that is accessible to (nearly) all, poverty cannot be effectively alleviated without providing people with jobs. A strategy aimed at enhancing economic growth is therefore extremely important. The Government needs to improve conditions for private business initiatives, and it is already looking carefully at ways in which the environment for private sector development could be improved. The regulations and laws on bankruptcy, taxes and registration are not known widely nor implemented well, and this discourages private business initiatives enormously. The Institute of Labor has started the process of formulating a new law on employment and new regulations regarding minimum wages, but clear regulations are needed. Further banking reform is also highly needed, given that access to a capital market and effective protection in the banking and insurance businesses are preconditions for the development of a capitalist economy. The Government also needs to help to improve entrepreneurial skills and support new private business initiatives. Incubation centers for private businesses could be established to provide initial -4 - training, counseling, monitoring and supervision of new businesses. Many programs that would normally be conducted under the control of local public authorities could probably be used to enhance private sector businesses. 10. It is necessary to identify the constraints to the development of the private sector, and the Government is already planning to conduct a comprehensive survey of private enterprises in the three major municipalities of the country. It is also forming private sector advisory groups, comprising municipal officials and representatives from the private sector. The country also needs to develop the institutional capacity needed to analyze labor market developments and to monitor labor market policy. Now that national survey data are available through the TLSS, a major effort needs to be made to build up the national capacity in social policy monitoring and evaluation. In the short term, this means: (i) setting up a policy monitoring and evaluation unit within the Ministry of Labor and the Institute of Labor; (ii) seeking technical assistance to support the activity; and (iii) considering the publication of an annual labor market report. To encourage the development of a functioning labor market, it is also essential that the Government should also pay its own wage arrears and respect its own employment contracts. Agriculture and poverty. 11. Agriculture plays a vital socioeconomic role in the economy and in the lives of a majority of the population. As other sectors have collapsed, the relative role of agriculture has increased. Agriculture now makes up 30 to 40 percent of GDP, more than 30 percent of export revenues, and 30 to 40 percent of Government tax revenues mainly through the taxation of cotton. It also provides formal occupations for more than 50 percent of the labor force. The workers on state farms have tended in the recent past to receive payments in-kind. Informal agricultural employment and self production on small plots now contribute to the basic livelihood of most of the rural population. Great changes in cropping patterns have taken place since 1991. The figures suggest that 200,000 to 300,000 arable hectares, or 20 to 30 percent of the total production area, may now be out of use. The main reason is the breakdown of the irrigation and drainage systems. In termns of production, there has been a decline in the land productivity of cotton since 1991 and an increase in total grain area and production. The tendency to increase the area of wheat, which is the main staple, has been constrained by rust disease. As noted earlier, the TLSS results show that rural areas are relatively worse off than urban areas, even when household economies of scale are taken into account. When asked about survival strategies, participants rank their household plot and livestock at the top of their lists, ahead of migration, trade, humanitarian assistance, wages and pensions (in descending order of importance). Given that the first pillar of the strategy to alleviate poverty is to stimulate labor-intensive economic growth, the agricultural sector should be a focus of attention. 12. Improved access to land needs to be considered a priority in a poverty reduction strategy. The country's area available for crop production is limited as mountains cover much of Tajikistan's surface. Arable land is some 960,000 hectares, or 7 percent of the total area of 143,000 square kilometers. Much of the arable land is subject to arid or semi-arid conditions, and irrigation systems exist on some 720,000 hectares. Pastures are a significant resource, with 3.6 million hectares under permanent pasture. Between 1995 and mid-1999, some 120 state farms were privatized and most of them were converted "dekhan" (private) - 5 - farms. Since June 1999, the program of land reform has accelerated, and targets have been set for the conversion of a further 160 kolkozes into private dekhan farmrs by March 2000 through issuing land share certificates to employees. The World Bank is assisting the land privatization process through its Farm Privatization Support Project (FPSP). Some small lots of land have also been distributed to households to supplement existing plots. These additional lots have allowed a greater level of food self-sufficiency and opportunity for sales. 13. For many of the most vulnerable, improved and extended land access could make a major difference. It is therefore recommended that: (i) an equitable and inclusive distribution of access to land should be ensured for rural households; (ii) there should be transparent legal protection of land access and use rights for individual dekhan shareholders; (iii) comprehensive information should be provided to shareholders on individual land rights and on choices of farm management systems; (iv) participatory consultation approaches should be undertaken for land stakeholder decision making; (v) there should be maximum flexibility of farm management arrangements to enable both individual and collective and mixed systems on former kolkozes; (vi) applied research, advisory and other support services (including farmer to farmer approaches) should be introduced to promote intensive, small scale technologies and farm management; and (vii) the status of some 200 farms designated for retention under the state should be reviewed. 14. Cotton production is extremely important. It dominates the agricultural economy, is the country's single most important cash crop, and covers more than 5;0 percent of the national irrigated area. The production and delivery of cotton is given absolute priority at all levels of the Governnent. In order to achieve improvements in resource use in the agricultural sector, part of the privatization program has already focused on the state cotton ginneries and the state cotton-marketing organization. Though rescheduled for completion by September 1999, the process is still ongoing. State orders for cotton were phased out in 1996, but some rayon Governments continue to play a key role in administering the current system. This system of producing and delivering cotton has a negative effect on poverty. First, the interest of the centralized system to control the number of producers and contracts has slowed down the land reform program, and has delayed realizing its potential to reduce poverty. Second, despite the fact that the financial returns on cotton are potentially good, the actual returns paid to producers are extremely low. The wages of much of the rural population working on kolkozes are also extremely low. As can also be seen from the TLSS data, the non-payment of wages and wage arrears at the local level has a major impact on poverty. For the future, therefore, it is recommended that: (i) the privatization program for the state ginneries and the central cotton marketing organization should be sustained; (ii) there should be freedom of choice for the cropping decisions of new private farms; and (iii) new institutional arrangements need to be developed to pernit the practice of small scale cotton production. 15. The irrigation and drainage systems have seriously deteriorated since the breakup of the Soviet Union. As a result, irrigation volumes have been reduced by as much as 50 percent. A good deal of irrigated land has been lost to production due to irrigation or drainage failure. For the future, it is recommended that: (i) irrigation rehabilitation should be pursued on the basis of economic viability and where there are potentially good economic retums on investments; (ii) rainfed, livestock and non-agricultural technologies should be investigated and introduced for areas where irrigation is not economically viable; (iii) user - 6 - charges for irrigation water should be applied and extended; and (iv) institutional reform of irrigation management systems should be carried out at the community, district and regional levels. Recently a credit for a Rural Infrastructure Rehabilitation Project has been approved by IDA, and this project should contribute to developments in this area. 16. Through carefully targeting interventions, micro-finance has an important role to play in assisting the most vulnerable groups. Some international donors and NGOs have already extensively used micro-credits for rural communities. Sustainability is always an issue with micro-credit because such schemes are typically donor driven and subsidized. Because loans are small, it is not uncommon for administrative costs to be higher than the value of the loan. To ensure sustainability, very high rates of repayment and low administrative costs are needed. As a result, it is recommended that: (i) financial sector restructuring should be sustained; (ii) credit should only be extended to private entities; (iii) micro-grants should be targeted toward the most vulnerable groups; and (iv) micro-credit should be targeted toward vulnerable groups and areas, and community savings and credit schemes should be introduced and extended. Education and poverty. 17. Access to good quality basic schooling and health services is not only a human right, but is also essential for poverty alleviation. Tajikistan's historically impressive educational statistics--especially for women, girls and ethnic minorities--place it in the ranks of high income countries, being in this respect well ahead of most other FSU countries. However, this situation is rapidly changing with the decay and destruction of school buildings, the flight of teachers both away from the country and also away from the profession, lack of basic textbooks and teaching materials, and absence of many children especially in war- affected areas, who are unable to attend school for much of the year. 18. The collapse of budgetary revenues over the 1992-1997 period forced severe reductions in social expenditures in all categories. Reported expenditures on education fell to the lowest levels in Central Asia. Education had been free and compulsory for all citizens at all levels until the mid-1990s. However, since then, numerous attempts have already been made to diversify sources of revenue to generate additional income in the education sector. Both formal and informal fees have also been introduced in basic education in the form of parental contributions to schools or payments for specific extra-courses. Unfortunately, there is insufficient knowledge about the extent and use of formal and infornal fees and about their equity effects in the education sector. Although some supplementary private payments for basic education may be necessary in the circumstances of Tajikistan, and may have actually prevented a collapse of the schooling system, there needs to be extreme caution over the potential equity and other effects of this policy on children, particularly those coming from poor families. Specifically, fees need to be transparent, subsidies consistent and well targeted, and the revenue from fees should be devolved rather than centralized. It is believed that is already strong evidence that the lack of textbooks has already adversely affected the poor. While the Government is currently addressing the need to distribute textbooks, it is important that it ensure that the poor have equitable access to the required textbooks as part of the reform measures. -7 - 19. In the present circumstances of Tajikistan, there needs to be a keen focus on the issue of access to education especially at the basic level, particularly by the poor. Access to education comprises barriers and constraints on both the supply side and also the demand side. If children are not attending school, it may be due to problems with the quality and activities of the schools, or it may be due to factors at home or more generally relating to poverty. For the nation as a whole, about 97 percent of children in the relevant age group are reported to be enrolled at primary school, and about 85 percent are enrolled at lower secondary school. Although these rates are still very high on an international comparative scale, considering the country's per capita income level, they show a marked decline compared to the country's historically strong and almost universal access to basic education. In particular, enrollment rates in grades 6-11 have dramatically decreased. The variations in the enrollment rates by income groups, especially at the lower secondary level, also show that the very poor are less likely to enroll their children than non-poor households. Further analysis should be undertaken into the barriers and constraints facing the poor in accessing education, and into the factors causing high absence rates. There should azlso be an analysis of regional, geographical and gender disparities within the education sector, and the links with poverty status. 20. Training should also be provided in establishing a monitoring and evaluation system for schools and in education policy development. There should also be a dialogue between the Ministries of Finance and Education in order to ensure that the education sector receives sufficient budget, and that the distribution of the education budget explicitly takes into account the needs of the poorest groups. The budgeting system needs to be reformed to change the normative approach, introduce incentives and ensure that efficiency savings (for example, from changes in student/teacher ratios) remain within the sector. Objectives and standards should be established to secure equal access to high quality education for all citizens of the country. The rehabilitation and restructuring of schools should be continued and accelerated; and new policies to raise extra funding for schools (such as through the rental of textbooks and the creation of parent-teacher associations) should be developed with particular attention to the needs of the poor. Of particular importance is the need to ensure that all teachers' salaries are paid regularly on time by the local Governments. 21. Increasingly questions are being asked about the validity of the current school curricula in terms of catering to a more modern labor market, rather than preparing students for a state controlled working environment. There is need for revisions to be made to the curriculum (and consequent retraining of teachers) to ensure that there is a strong emphasis on the application of knowledge and new skills. While the skills provided by basic education are important for all people to have, it is also important that students can continue with post- basic education which is crucial for the personal and social development of children and also for the economic and social development of the country. Therefore, in addition to giving the highest priority to basic education, it is also important that the Government focuses on the needs of the vocational and technical training system. A large share of t]he students at the vocational and technical schools are likely to come from poorer families, and these schools are in urgent need of upgrading. Most of them are presently outdated, inefficient and inflexible, and need to be restructured, reoriented and in some cases downsized in order to provide more short-term and flexible training courses. One important way for people in the rural areas to get themselves out of poverty is through improved practices in farming, farm business or small businesses generally; and the vocational and technical educational system - 8- could play a major role in this. So there is need for a review of the vocational and technical training system in order to enable the training schools to provide more short-term and flexible training courses and to contribute to improved practices in farming, farm business and small businesses generally. Health and poverty. 22. The health system is in a very serious situation, with the poor in particular suffering from lack of access to decent quality services. Tajikistan used to rank with middle upper- income countries in terms of the major health indicators. There are major constraints on the accuracy of the data, but it is believed that life expectancy has declined since independence although it may have improved somewhat since its low point in 1994. The official Government statistics suggest that in recent years there has been an improvement in infant and maternal mortality rates, and a decrease in birth and death rates, but the health status of the people has probably deteriorated in some important ways in recent years. In particular, the rates of respiratory diseases, intestinal infections, malaria and typhoid all rose in the period 1995-1998. The findings of the TLSS further demonstrate that substantial proportions of households are food insecure. Surprisingly, according to the TLSS results, the majority of people reported that their health status over the last year had been good or very good. Even so, health status in Tajikistan remains precarious, as the country has to cope with a double burden of disease, inflicted by both chronic non-communicable diseases and infectious diseases. 23. At independence, Tajikistan inherited the standard Soviet health system, which was generally comprehensive but highly centralized and inefficient. There were a large number of hospitals, hospital beds, doctors, and nurses and medical technicians. The exigencies of transition and civil war, with the associated sharp economic decline and resource constraints, have created a widening gap between the health care budget and the actual costs of care. Availability of care does not constitute accessibility of care; and in Tajikistan, although there is still an oversupply of facilities, beds and personnel, there is also a striking absence of even rudimentary medical equipment, supplies and pharmaceuticals. Another particular problem concerns the salaries of health workers which are less than half of the average of all workers' salaries. 24. Access to health care for the poor is now a particular problem, especially as cost recovery schemes and private sector expansion are introduced. As social inequalities in Tajikistan have increased, health inequalities have developed accordingly. Of great concern, the poorest 20 percent of the population are reporting illness and seeking care at lower rates, as they face financial barriers to care more frequently. The poorest are able to devote a smaller proportion of household expenditure to health care, and they are more likely to be seen by a feldscher than are the non-poor. They (and their children) face a higher risk of water-borne disease than the richest quintile, having less access to centralized or piped water. Some women of reproductive age are particularly vulnerable to the effects of poverty on health. 25. There has been significant discussion and planning in the Ministry of Health in recent years about the reform of the health care system and especially about concentrating more resources at the primary level. The World Health Organization has been assisting with this. -9- Despite the plans made, implementation has been constrained and only modest actual resource reallocation has yet occurred. The ongoing IDA-funded project will help the Ministry of Health with the further implementation of the reforms, but it will only focus initially on two rayons. There will soon be a parallel social sector rehabilitation project funded by the Asian Development Bank, and that will extend the reformns further. But much remains to be done. The incentive system will also need to be changed. Health care financing is still input-based, and this rewards excess capacity and unnecessary hospital admissions of extended duration. In general, the health delivery system still emphasizes curative care, rather than disease prevention and health promotion. 26. If a greater proportion of the available resources were to be allocated to primary health care, there would be an improvement in equity as well as efficiency. The poorest quintile of the population access primary care more frequently than tertiary care; and so redirecting resources to primary care rather than to hospital based services (more frequently used by the non-poor) would target the poor in Tajikistan. Young children and women of reproductive age, who are the most frequent consumers of health care, and among the most vulnerable, also access most services at the primary level, and therefore improvements in primary and maternal health services would additionally benefit this group. But the quality of care should also be improved in the peripheral facilities, and health worker capacity enhanced, in order to increase utilization rates generally of primary care. Strengthening primary care, and primary care providers (feldschers and midwives), is therefore the key strategy for restructuring the health delivery system to the benefit of the poor. Social protection and poverty. 27. The third key pillar of the proposed poverty alleviation strategy concerns the targeting of assistance to the poorest, that is, those people who have special difficulties or needs and for whom targeted assistance is required. During the last five years, civil conflict and the economic and social dislocation associated with transition have exacerbated the disadvantage of some of the "traditionally poor"-- pensioners, families with large numbers of children and single parent families-- and given rise to a new group, the "working poor". And while the number of vulnerable people has increased, the ability of the Govermnent to fund social protection has decreased dramatically. The value of benefits has fallen in real terms and the number of actual recipients has been reduced considerably below even the Government's target. 28. In view of the large number of extremely poor people in Tajikistan, it is important to examine the organization and the financing of the existing safety net programs and to evaluate whether they are appropriate and effective in reaching the poor. Vulnerable groups are currently supported by a complex mix of both formal and informal safety nets, with benefits in cash and in-kind from Government, NGOs, and the extended family and wider community. Much of the support to the poor is provided by informal safety nets--by community groups, NGOs and the family. 29. There are two main types of official social protection benefits available in Tajikistan: the cash compensation program offering social assistance benefits to the poor; and pensions providing social insurance benefits to the elderly, disabled and bereaved. The annual Government outlay for the cash compensation program should be about TR6 billion, - 10- assuming that the eligible population is about one million people (just less than 20 percent of the total population). In practice, however, the cash compensation payments since 1997 have been considerably less. In order to improve payment rates, two reforms were recently made to the program: the paperwork required for claiming the benefit was simplified, and the duration for the period of the claim was increased to three months to reduce transaction costs. However, these reforms have had little impact. The other social benefit available is the pension program, which also needs to be reformed. The value of pensions relative to salaries has declined from 50 percent in 1995 to 35 percent in 1998, and pensions have become almost worthless. 30. The cash compensation program could be redesigned in several ways. One alternative is consumption-based proxy-means testing. Proxy-means tests use a set of characteristics that prove to be good predictors of poverty in order to establish eligibility for a social program. A second and simpler way is to target transfers according to some fixed observable characteristic of the recipient household or individual, such as the number of dependent children or single parent status. The poverty profile demonstrated that the number of children in the household is one of the strongest correlates of poverty in Tajikistan. Therefore, any change that affects children will also affect the vast majority of households. One option would be to introduce a child benefit that is linked to school attendance, which might improve attendance rates and help alleviate poverty as well. Another approach would be to allow exemptions from fees for textbooks and other materials for the poorest children. Another alternative could be an in-kind benefit such as free school lunches, which would be targeted at the poorest children and might address some of the problems of malnutrition and school attendance. 31. Yet another alternative is for individuals to select themselves for assistance. Self- targeting programs such as public works (self targeting based on below market wages) have been found to be effective mechanisms of providing income and/or food security in other parts of the world. Typically such programs are hybrid in nature, combining poverty alleviation (i.e. cash transfers) with a work requirement to ensure self-targeting and with infrastructural development that promotes economic growth. Such a program could build on the experiences of the Tajikistan Social Investment Fund. Given the pressing need in Tajikistan for both of these elements, the development and expansion of public employment programs deserves attention. 32. In summary, it is clear that the existing system of social protection should certainly be reformed, given that it is ineffective and inefficient. Having an efficient and well targeted safety net will also be important as the Government moves ahead with civil service reform and especially with staff retrenchments. The Prime Ministerial Working Group on Poverty has already evaluated the alternative mechanisms for targeting the most vulnerable, and the Government intends to pilot a reform of the cash compensation program based on school enrollment or attendance, with local community involvement. The reformed system of cash compensation should be monitored closely and adjustments then made on the basis of experience gained. In addition to reform of the cash compensation system, there is need for reform of the pension system and an expanded program of public works. - 11 - Governance and poverty. 33. Poor governance contributes significantly to the problem of poverty in Tajikistan. The main historic factors in the political life of Tajikistan are the legacy of 70 years of Soviet rule, the civil war and unrest that raged throughout the republic from 1992 until the peace accords of 1997, and provincial alignments of power. However, the recent political developments now bode well for an accelerated program of poverty reduction. Since the signing of the peace agreement between the Government and the United Tajik Opposition in June 1997, there has been an important foundation for peace and securit,y. The Government is also embarking on a major program of public sector reform. 34. The fundamental aspects of governance in Tajikistan are inherited directly from the Bolshevik politics. Tajikistan's public administration reflects a structure similar to other FSU republics, with a republican Government overseeing a three-tier local Government structure at provincial, district and village levels. The absence of an overall body of regulations covering civil service management, together with inadequate pay and incentives, have resulted in low morale and motivation among Government employees, difficulties in retaining competent staff, and distortions of reward systems and administrative structures as ministries resort to offering a range of non-monetary benefits in order to retain and motivate key staff. Responsibility is also fragmented. The structure of decision-making is still highly centralized, and there is also an absence of sufficient checks and balances in the Government -- the judiciary is extremely weak compared to the legislative, and it is especially weak compared to the executive branches of the Government. The civil war has resulted created an atmosphere of lawlessness in which armed extortion is common. The lack of a strong judiciary and police force make it hard for impoverished persons to seek redress under the law. There are also many problems with the existing legal system. For example, the legislation on the rights of citizens lacks sufficient transparency. The Parliament, possessing insufficient expertise in international legal standards, still tends to formulate laws in the Soviet manner. 35. The constraints on the rule of law encourages abuse of power and position in the local economy of Tajikistan, and this negatively affects the livelihood of the population, especially in rural areas where over 70 percent of the population lives. Most of the rural population is engaged in agriculture -- even doctors and teachers, who are often not paid their meager salaries, receive their primary income from subsistence agriculture. Agriculture is managed in such a way that farmers have few rights and must circumvent the law as a coping mechanism. Private business and trade create an opportunity for small retailers to ply their wares and for farmers to sell produce from their kitchen gardens. However, a "cut" is taken from most trade in bazaars by inspectors, policemen, and criminal groups. Due to the collapse in the economy, state budgets no longer suffice to support public services; but it is the poor who are least able to cope with these problems, since they possess neither the money to bribe nor connections to wield influence. 36. It is highly encouraging that the Government has already expressed its strong political will to implement key reforms in public administration, and it is now actively preparing a program of governance reforms. Implementation has already started withl, for example, the establishment of a new state procurement system. The Government also iformulated an anti- - 12- corruption program, which includes removing price distortions in the economy, reviewing laws and decrees, reforming agriculture and land rights, training law enforcement officers and the judiciary, and combating drug trafficking. Some important progress has been made in some of these areas. For example, there have been attempts by the Constitutional Court to improve the standards of legislation. There has also been some improvements in the work of the police in the anti-corruption efforts. A major program to eliminate drug trafficking has also begun. In addition, the number of military checkpoints has declined dramatically, following the peace accord of 1997 and a subsequent decree of the President. 37. There are several further measures that could be taken to improve the lot of the poor in Tajikistan through better governance. These include "top-down" measures, such as strengthening the judiciary and police force, as well as "bottom-up" measures such as encouraging citizens to take joint action through non-Governmental organizations. An obvious priority in reducing lawlessness is to reduce the number of weapons in the country. One way to do this might be means of a weapons buy-back program. Another priority in reforming governance is to improve the capacity of the judiciary to increase the checks and balances in Government and to help strengthen the rule of law. This could be done by increasing their salaries and providing their offices with basic equipment. The training of judges could also be improved through seminars and technical assistance. Governance could be further improved by rendering the law more transparent. Campaigns to educate the citizenry concerning their rights and duties under the law would also improve transparency in governance. 38. In addition to the measures designed to improve governance at upper levels of the state, it is necessary also to provide impetus to improving governance from the "bottom-up". Creating more Non-Governmental Organizations (NGOs) in rural areas could help to organize the population to fight any abuse of positions corruption and to provide them with legal aid. NGOs could also improve governance in education, health care, and environmental management. Community institutions such as the mahallah and jamoat might also encourage more popular participation in governance. The further liberalization of marketization in agriculture and the total elimination of implicit as well as explicit state orders would reduce the opportunities for local officials to abuse their powers in determining the terms of contracts. - 13 - INTRODUCTION 1. Poverty is the central development issue in Tajikistan. The country is now one of the poorest in the world, with an estimated per capita GNP of only $330 (1997). Poverty is not new in Tajikistan, as it came to independence with widespread material poverty. As shown in Table 1. in 1989 the country had the lowest mean per capita income amongst the Soviet republics, well below 50 percent of the mean income in Russia, and the highest proportion of the population in poverty. At independence, the human development index (HDI) in Tajikistan was also lower than in Russia and other parts of Central Asia. Despite the material poverty, however, literacy rates in Tajikistan at independence were at the level of highly industrialized countries; infant mortality rates were low; and access to education and health services was free and almost universal. These standards were, however, largely the result of substantial budgetary transfers from Moscow. Table 1: Poverty and Human Development in USSR (1989) Relative per Prop. of UNDP Human Life Adult capita mean Population with Development Expectancy Literacy (%) income* Per Capita Index (1991) (years) (1991) (1991) Monthly Income below RR75 Russia 100 5 0.858 Kazakhstan 80 15 0.774 69 97.5 Kyrgyz Republic 58 33 0.685 68 97.0 Turkmenistan 57 35 0.697 66 97.7 Uzbekistan 51 44 0.664 69 97.2 Tajikistan 46 51 0.629 70 96.7 All USSR 89 11 Source: Atkinson and Micklewright. 1992; United Nations Development Program (UNDP) Human Development Report, 1994. Note: * Per capita gross household income for workers and collective farm workers in 1989. 2. Soon after independence, Tajikistan entered a period of civil conflict and massive economic decline which lasted until 1997. The country lost its traditional mrtarkets for imports and exports, and also suffered from the end of budgetary transfers from Moscow. These developments led Tajikistan to severe economic decline. As shown in Table 2, gross domestic product (GDP) declined precipitously and inflation soared after 1991. At the same time, the Government budget shrank drastically; and between 1992 and 1997, the Government's revenue as a percentage of the GDP was reduced by more than half. These events had an adverse effect on the material welfare of people, and also oln their capacity to live a healthy and secure life. The conflict disrupted people's lives and dlestroyed the civil society; the economic decline worsened the already high levels of poverty, and created pockets of deep distress and social exclusion; and the sharp decline in public spending reduced the Government's ability to assist needy people and to provide basic services such as education and health to all. - 14 - Table 2: Summary Macro-Economic Indicators, 1991-1999 Year Percentage Change Consumer Price Governtment in GCP Index Revenue (% GDP) 1991 -7.1 1992 -29.0 26.6 1993 -11.0 27.1 1994 -18.9 350.4 44.5 1995 -12.5 609.3 15.2 1996 -4.4 418.3 12.1 1997 1.7 88 13.7 1998 5.3 43.2 12 1999 5.5 (est.) 20.3* (est.) 12.8 (est.) Source: Govemment and IMF; World Bank for CPI. Note: * Estimated on monthly consumer price index from January through August. 3. There have been two major developments in the last three years. First, the scale of the civil conflict was reduced considerably after the signing of the peace agreement in June 1997, and since then the peace accord has largely been implemented. Second, various structural economic reforms were introduced; at the same time growth resumed, and inflation was brought under control. The economy grew slightly in 1997, accelerated in 1998 to 5 percent, and is projected by the International Monetary Fund (IMF) to be 5.5 percent in 1999 and to increase to 6-7 percent annually over the next few years. Inflation has also been relatively low since 1997. Tight fiscal policies have narrowed the budget deficit to 3.8 percent of GDP, largely through expenditure adjustments. Government revenues are, however, projected to be no higher than 14 percent GDP in the near future due to the extremely weak tax collection capacity; therefore the Government budget will continue to be highly constrained. Moreover, the economic structure of the country depends significantly on the export of only a few commodities, particularly cotton and gold, and is thus exposed to the effects of external price shocks. At the same time, unfortunately, the human development indicators have been worsening. In 1997, the UNDP ranked Tajikistan as 115 out of 174 countries using the HDI. -15- CHAPTER I: A PROFILE OF POVERTY A. Measuring Poverty in Tajikistan 1.1 Events through the 1990's exacerbated the inherited problem of material poverty in Tajikistan, and also gave rise to new dimensions of the phenomenon. To develop an appropnate strategy for the alleviation of poverty, it is critical to have a firm understanding of the nature of poverty in the country. This requires a wide range of information, both qualitative and quantitative. Fortunately, there is a considerable amount of qualitative data from work completed in last few years. This includes various social assessments which have been undertaken; a large volume of data collected by the Tajikistan Social Investment Fund (TASIF); many specific reports produced by Government agencies such as State Statistical Agency (SSA); and also other reports by donor agencies (such as UNDP and the Asian Development Bank) and international non-govermnent organizations (NGOs).l 1.2 These sources of information show that in Tajikistan, poverty is viewed by most people as a multi-dimensional phenomenon.2 When asked to define it, respondents typically emphasize (i) lack of clothing, (ii) food shortages, (iii) inadequate salaries, and (iv) low pensions. Some respondents also include lack of land or livestock in the definition. In the war-affected areas, such as in Khatlon and the Karategin Valley, local communities also include shortages of construction material and insufficient shelter. In contrast, communities residing in valleys (in Leninabad, Khatlon and the Regions of Republican Subordination (RRS) also include inadequate drinking and irrigation water, and low crop yields. Mountainous/remote communities in areas such as Gomo-Badakhshan (GBAO) also include (i) lack of physical infrastructure (transportation, communication, electricity), (ii) lack of social infrastructure (health, hygiene, education), (iii) unemployment, and (iv) and underdeveloped production and distribution channels. 1.3 In addition, to obtain new and comprehensive quantitative data, the Tajikistan Living Standards Survey (TLSS) was carried out in Spring 19993 . This survey collected data on demographics and housing, expenditures and consumption, food security and agriculture, education and health, employment and migration etc. While the data are believed to be reliable, as with any survey and analysis of poverty, there are many approaches to its measurement. These include issues such as whether to use an absolute or relative approach, the choice of the poverty line, the selection of the welfare indicator, the valuation of in-kind benefits, and the valuation of home production. The survey results are also limited as they only represent one point in time, and exclude the homeless. The main approach used here, ' These include a Food Security Survey carried out by ECHO; a Household and Bazaars Suarvey carried out by Save the Children US; and recent nutritional surveys by German-Agro Action (in the Regions of Republican Subordination and Leninabad) and by Action Against Hunger (in Khatlon and Khulyab). 2 Between November 1997 and October 1999, the TASIF conducted seminars and surveys in 11 rayons (in all major regions of the country, including urban and rural, mountainous and valley, and war-affected communities) to inform the targeting of its micro-projects. In each rayon, participants (local government officials, mahalla representatives, farm managers and laborcrs, women and pensioners, etc.) identified community-specific poverty criteria; and then, based on these criteria, selected the poorest jamoats and kishlaks. 3The survey methodology is discussed in more detail in Appendix 1. The issue of equivalence scales is discussed in more detail in Appendix 2 - 16- however, is first to focus on five possible absolute poverty lines (based on expenditures and using an intertnediate assumption for economies of scale); then to examine the issue of the distribution of expenditures; and finally to focus on the poorest 20 percent of households. The data from the TLSS are compared with the qualitative information obtained from a wide variety of other sources. In this way, a picture is built up of the extent of poverty (using different assumptions); the geographical and demographic patterns of poverty; and the correlates of poverty with the labor market, the use of social services, the structures of incomes and expenditures, and infrastructure and assets. B. The Level and Depth of Poverty 1.4 The picture of poverty obtained from the various social assessments and seminars conducted by the TASIF suggest that about 60 percent of the population consider themselves as poor. This is confirmed by data from the TLSS in which just over six out of ten households ranked themselves as being on the bottom three rungs of a nine-step ladder. It is widely felt that the most vulnerable groups are pensioners, the disabled, families who have lost their breadwinner, large families (with five or more children), teachers and doctors, and the unemployed. In cotton-growing areas, collective farm workers are considered vulnerable; in war-affected areas, families without shelter and/or livestock are considered vulnerable; and in mountainous areas, families without land are considered vulnerable. Although subjective, this is an important perspective on poverty in Tajikistan. 1.5 The people's own perception of poverty can be compared with the quantitative results from the TLSS. Using the poverty lines suggested by the SSA, taking a rate of TR20,000 for the "poor" and TRlO,000 for the "very poor", and assuming that there are some moderate economies of scale enjoyed by households containing more than one person (assuming a theta of 0.75), about 83 percent of the population may be considered to be very poor and 33 percent may be considered to be very poor (using expenditures rather than incomes). This is a head-count of poverty. In addition, however, it is also useful to consider the poverty-gap index (P1), which gives an indication of the depth of poverty and the cost of eliminating it by showing how much would be necessary to bring every household up to the minimum subsistence level. Table 3 shows that the gap is very considerable (at 35.8) if the Government's poverty line of TR 20,000 is used. Finally, the Foster-Greer-Thorbecke (P2) measure shows the severity of poverty, by weighting the poverty gaps of the poor by the size of those gaps, thus giving a greater weight to those who are most poor. This measure can thus be thought of as the sum of two components: an amount due to the poverty gap, and an amount due to inequality amongst the poor. In the case of the TLSS data, Table 3 indicates that the index is 19.0 using the Government's poverty line of TR 20,000. 1.6 If on the other hand, the poverty line is based on a purchasing power parity (PPP) poverty line of $1.075 a day4, the poverty rate would then only be about 16.3 percent (with 4In May 1999, $1 could be exchanged for TR1,200 on the foreign currency exchange market. However, this exchange rate does not take into account differences in the cost of living between the USA and Tajikistan. There is currently no official consensus on the appropriate exchange rate to use to calculate PPP for Tajikistan. The World Bank has, however, recently produced an estimate for 1999 of $1=TR234. Thus an "extreme" poverty line of $1.075 PPP a day is equivalent to TR7,557 a month, whilst a more moderate poverty line of $2.15 PPP a day is equivalent to TR15,1 11 a month. - 17 - P1 at 4.4 and P2 at 1.8). Using an unofficial minimum consumption basket of TR32,083 per person per month5 would imply that about 96 percent of the population may be considered to be poor. Table 3: Poverty Measures in Tajikistan (a) Minimum consumption basket 1998 (TR32, 083) % poor 95.7% P1 56.4 P2 36.8 (b) $2.15 PPP a day (poverty line = TR15,111) % poor 65.4% P1 22.9 P2 10.9 (c) $1.0 75 PPP a day (poverty line = TR 7,55 7) % poor 16.9% P1 4.4 P2 1.7 (d) State Statistical Agency (poverty line = TR20, 000) % poor 82.6% P1 35.8 P2 19.0 (e) State Statistical Agency (poverty line = TR] 0,000) % poor 32.8% P1 9.2 P2 3.9 Source: TLSS Notes: (i) Sample size 14,142 individuals. (ii) Based on expenditures rather than incomes. (iii) Assumption that theta (for economies of scale) equals 0.75. 1.7 There are several major conclusions from the TLSS. First, the level of poverty varies enormously depending on the poverty line used. From Table 3, it may be seen that the estimate of poverty ranges from 17 percent (using the $1.075 a day poverty line) to 96 percent (using the minimum consumption basket). Using the criteria developed by the Government's SSA, the poverty rate is slightly over 80 percent. Second, as shown in Appendix 2, reported household income is lower than expenditure, except at the very top of the distribution beyond the 98th percentile, and therefore poverty defined according to income will always be higher than when defined according to expenditure. Income data is usually subject to under-reporting; and in the case of Tajikistan, it is exacerbated by the growth of the informal sector, arrears in payments in the formal sector and the de-monetization of the economy. Figure 1, which shows the cumulative distribution of both per capita household income and expenditures, helps to explain the sensitivity of the results to the various poverty lines. The distributions of both incomes and expenditures within Tajikistan are very heavily skewed, with most households highly concentrated in a narrow range of values in the bottom 5 The Governrent has defined a provisional minimum consumption basket and its value in 1999 was estimated to be TR32,083 per person per month. This figure is made up of an allowance of TR23,934 for food-stuffs (based on a basket of goods that provide a rninimum calorific value), TR2,947 for non-fbod items and TR2,695 for services. However the minimum consumption basket has yet to be formally adopted by parliament, and in any case is not used to define the level of social transfers. -18 - four-fifths, with the result that headcount measures of poverty are very sensitive to even small changes in the value of an absolute poverty line. Figure 1: Cumulative Distribution of Household per capita Income and Expenditures, 1999 Cumulative Distribution of Hlousehold per capita Incone & Expenditure, 1999 100 90 - _ 80 U 70 _ _ _ _ _I ° 60 0) 40 0 30 - 20 - 10 0 0 10000 20000 30000 40000 50000 60000 Tajik MbWes Source: TLSS 1.8 Compared to other Central Asian republics for which data is available, the TLSS results are consistent with the widespread view in the country that the vast majority of people are poor. Inequality also seems to be lower in Tajikistan than elsewhere in the region (see Table 4). While the TLSS data show that individuals living in households in the top 10 percent of the expenditure distribution spent four times as much as those living in the bottom 10 percent, the gini coefficients for the distribution of both income and expenditure are less that in the neighboring countries. However, inequality has increased over time. Atkinson and Micklewright (1992), using the 1989 Family Budget Survey data estimated that the gini coefficient for the distribution of per capita income was 0.308 in Tajikistan. Ten years later, according to the TLSS, it had risen to 0.47. Table 4: Distribution of Income and Expenditure in Selected Countries Gini coefficient from LSMS Individuals in Poverty (%) data: selected years (per (assuming theta equals 1) capita expenditure) Azerbaijan(1995) 0.35 68.1 Kazakhstan(1996) 0.35 34.6 Kyrgyzstan (1993) 0.54 (1996) 0.46 68.7 Tajikistan (1999) 0.47 96.0 - 19- C. Identifying the Poorest 1.9 The vast majority of the population of Tajikistan are poor, at least using the minimum consumption basket as the poverty line. A strategy for the reduction of poverty in the country needs to take this into account. But the choice of poverty line is essentially a political one, and it is also necessary to focus on those at the bottom of the distribution in order effectively to target policy interventions on the most vulnerable in society. The main analysis of the profile of poverty now focuses on individuals who live in households in the bottom 20 percent of the distribution of per capita expenditure. This approach is simple and transparent, and has the advantage of avoiding a degree of complexity that may threaten a broad endorsement of the poverty line. 1.10 Poverty can be looked at in two ways: firstly the risk of being poor faced by individuals with different characteristics, and secondly the composition by characteristics of those that are poor. For each of the following parameters, the first table presents the percentage within each quintile of the distribution of per capita expenditure (with the incidence of poverty being defined as the percent of any particular ,group in the bottom quintile); and the second table presents the composition of both the richest and poorest quintile, as well as that of the population as a whole. 1. 11 Poverty by Location. The TLSS data can be used to draw sonie conclusions about the geographic profile of poverty in the country. Table 5, for example, shows that only 18.6 percent of individuals living in urban areas can be said to be poor, compared with 23.4 percent of those living in rural areas. Thus the relative risk of poverty for those in urban areas is 0.83, compared to a relative risk of poverty for those in rural areas of 1.04. Put another way, urban dweller are 20% less likely to be poor than on average, whilst rural dweller are 4% more likely to be poor. This is partly due to the fact that rural households tend to be larger than urban ones, with a mean household size of 7.6 people in rural areas compared with 5.7 in urban areas. However a slight differential between those in urban and rural areas remains even when household economies of scale are taken into account. Table 5: Poverty Incidence amongst Individuals by Location Poorest 2 3 4 Richest 20 20 percent percent All individuals 22.4% 21.7% 20.5% 19.1%Mo 16.3% Urban 18.9% 19.9% 19.4% 19.8%Yo 22.0% Rural 23.4% 22.2% 20.9% 18.9% 14.6% Dushanbe 7.1% 17.5% 12.9% 25.91% 36.6% GBAO 39.1% 26.9% 18.5% 10.2% 5.4% RRS 17.0% 16.0% 20.7% 22.9% 23.4% Leninabad 22.4% 23.7% 21.7% 18.1%fo 14.2% Khatlon 26.8% 24.2% 21.2% 17.1% 10.7% Source: TLSS - 20 - 1.12 Poverty is much lower in Dushanbe than elsewhere in Tajikistan, whilst GBAO is the worst off region, followed by Khatlon, Leninabad and RRS. Individuals living in GBAO are three-quarters more likely to be poor than on average, with a relative poverty rate of 1.75. They are also particularly exposed to some potential natural disasters, especially as it is in that region that Lake Sarez is situated. The extent of poverty in that region also contrasts sharply with Dushanbe, which has a relative risk of poverty of 0.32. However, as Table 6 shows, although poverty in GBAO is relatively high, that oblast only accounts for 6.9 percent of all poor people. Table 6: Composition of the Poorest and Richest Quintiles by Location Poorest 20 percent Richest 20 percent All Tajikistan All individuals 100.0% 100.0% 100.0% Urban 18.5% 29.7% 21.9% Rural 81.5% 70.3% 78.1% Dushanbe 2.1% 14.5% 6.4% GBAO 6.9% 1.3% 3.9% RRS 19.2% 36.4% 25.3% Leninabad 26.1% 22.8% 26.1% Khatlon 45.7% 25.0% 38.1% Source: TLSS 1.13 The data from the TLSS may be compared with information obtained from other, more qualitative sources. Starting with GBAO, the poorest region according to the TLSS data, there is evidence not only of the extent of poverty there, but also of some deterioration over time6. This is seen from the results of three nutritional surveys carried out in 1994, 1996 and 1998. The latest survey found that, although the rate of deterioration in the health and nutritional situation in the oblast was slower during the 1996-98 period than during 1994-96, it is still getting worse and is now acute. The key findings of the 1998 survey include: (i) 6.1 percent of children between 6 and 59 months of age are wasted (a weight for height measure), reflecting an acute or recent nutritional deficit; (ii) 53 percent of children between 6 and 59 months of age are stunted (a height for age measure), reflecting chronic malnutrition; (iii) 26.8 percent of children between 6 and 59 months of age are under-weight (a weight for age measure); and (iv) there is a high rate of iron and vitamin deficiencies among women of reproductive age (15-59) and children under five, reflecting that the food provided by the humanitarian agencies or grown locally does not ensure good nutritional health. Although there are some important similarities in the health and nutrition situation in GBAO compared to other parts of the country, the extent of stunting and also iodine deficiency in children 6 to 12 years of age, and the prevalence of anemia in women, and are 6 As discussed later in this report, this agriculturally disadvantaged region has moved from being 15 percent self sufficient in wheat in 1995 to a projected 80 percent self-sufficiency in 1999. The impact of land redistribution there to individual households on their livelihoods has undoubtedly been dramatic and positive. It is therefore surprising that the nutritional status of the children has apparently deteriorated during this period. This phenomenon requires some further analysis. - 21 - more severe in GBAO. This confir-ms that the poverty situation is probably most severe in that part of Taj'ikistan. Figure 2: Proportion of Individuals who are Poor by Region 4 u0 ne BO RS einbd Khto 1.4 Ante pr f h cuty hr tepvet staio3emst b eatvl bdi Khto. cin gistHne cnute ure n 98t3ass0tentrto neso th oulto i htobatbyfcsigo hide bten25n 9 ots.Ti gopi chilren n Kugan-usabe and A7. peren So th cideninad Khlyabtreafetenwt sthpounatinirhonic mhatolnutritifon;using) aout6neret childrenbewn haad beemnts vacinae frorpi meuaslesl(v thermsvlerarlesins ofirston anboie dfefciencby, anod manyeccommunicable keidinses areon the incrveyase: due aoto ignorance,nd poor qualdrnity watrgadsaniTueation.4 Inerestingly the sretembleethtchildren in urbana areasfce it atn are morte mal-noutrished.i aot87eceto 1.15re Histouricall,Tuwas alway thoughcet Lfteniaa chblasen was thelyaparto fftecutry with sthenleas povrty Ihrni 199,l erantgrnAtion;bu (GAA)eno conlduced had sureyn toacciaess the nurition, theinraltheoigoane and fodscrtpiutonr qandlivin cotrnditionstaion.Lentrsinabdglasthan also in the RSS, regions where the organization is providing food, non-food and rehabilitation assistance. The organization had conducted a similar survey in 1996, which serves as a baseline. The 1998 survey found the food security situation of the population to be "critical". Of the households surveyed, 44 percent in RRS and 11I percent in Leninabad had experienced a day without food during the week preceding the survey (even though the survey took place in September, the most food-secure month). Most hiouseholds complained about regular food shoirtages; and in the RRS, the food situation had deteriorated since the war. The nutritional situation of the households was found to be "rnarginally inadequate" - 22 - due to low protein consumption. This had not changed significantly in the last two years. Most children are breastfed until the age of 18 months, but babies between 6-12 months are not adequately fed because they do not receive supplementary food. The nutritional status of children under five years of age showed that: (i) 46 percent suffered from chronic malnutrition; (ii) 20 percent are affected by severe growth retardation (stunting); and (iii) 11 percent suffer from acute malnutrition (wasting). Acute malnutrition was more prevalent in the RRS than in Leninabad. Children under 18 months of age were most affected by wasting due to the high prevalence of disease at this age. Children over two years of age were more affected by stunting. 1.16 In summary, then, the poverty picture obtained through the TLSS is generally borne out by other sources of data. Taking the prevalence of stunting of children, the situation is particularly bad in GBAO (53 percent); it is very serious in most of Khatlon (39 percent); and it is still a major problem even in Leninabad (20 percent). 1.17 Poverty by Demographic Group. According to the results of the TLSS, children experience a higher risk of living in poverty compared with adults, with a relative risk of between 1.03 and 1.11. The risk of poverty increases sharply according to the number of children under 15 living in the household. Only 7 percent of individuals living in households with no children are poor, compared with 31 percent of those living in households with 5 or more children. People living in households with children comprise the vast majority of the poor: over four-fifths live in households with at least 3 children and nearly a half live in households with at least 5 children. Therefore targeting large households with children may represent an important option for reaching the bulk of the poor. Table 7: Poverty Incidence amongst Individuals by Demographic Group Poorest 2 3 4 Richest 20 20 percent percent All individuals 22.4% 21.7% 20.5% 19.1% 16.3% 100% Age in Years 0-5 23.1% 23.1% 20.8% 18.6% 14.4% 100% 6-15 24.8% 21.6% 20.0% 19.1% 14.5% 100% 16-64 20.8% 21.5% 20.9% 19.3% 17.6% 100% 65 + 22.7 18.9% 20.3% 19.4% 18.7% 100% No. of Children under 15 Zero 6.8% 15.1% 16.2% 15.7% 46.2% 100% 1-2 16.5% 18.6% 20.5% 23.3% 21.1% 100% 3-4 20.2% 22.2% 21.6% 20.4% 15.6% 100% 5 or more 30.7% 23.9% 19.9% 15.1% 10.5% 100% No. of Elderly Zero 21.2% 21.1% 20.2% 19.4% 18.1% 100% 1 25.5% 23.6% 20.8% 17.8% 12.3% 100% 2 21.9% 21.4% 21.8% 20.1% 14.9% 100% 3 ormore 78.3% 21.7% 100% Gender of Household Head Male 21.1% 21.9% 20.8% 19.6% 16.5% 100% Female 28.6% 20.3% 19.5% 16.4% 15.0% 100% Source: TLSS - 23 - 1.18 Besides poverty among children in general, there are some particular groups of children in distress. For example, there has been a growing problem of street children who started to appear in the early 1990's after the deterioration of the economic and social environment. The problem of street children in Dushanbe was assessed by UNICEF in 1995 and Medicins San Frontieres (Holland) in 1997. Both studies relied on qualitative methods of interviews and observations. The first study interviewed 184 children in nine different sites around the city and some "experts"; the second interviewed 203 children in various sites in each of the city's four districts, and concerned government and international community officials. Estimates of the total number of street children in Dushanbe range between 1,000 and 2,000. The majority of these children work or beg to buy food arid/or supplement their family income. However, it does seem that the majority of the street children do have homes in which to stay at night; and so the situation, while serious and deteriorating, is not totally dire. 1.19 Another group of concern is institutionalized children. In 1998, Orphans and Refugees Aid, an international NGO, prepared a report on the care of children in internats (state institutions). Children enter intemats for a variety of reasons: (lue to the inability of parents to feed or clothe them (60-70 percent of them), physical or mental disability, delinquency, the death of both parents, their abandonment by one or both parents, or other reasons such as their parents being disabled, alcoholics or in prison. F'rior to the collapse of the Soviet Union, internats were well funded and provided comprehensive care. Now, due to a variety of factors-lack of funds for staff salaries, for building maintenance, and for food and clothing-the physical, intellectual and emotional growth of children is jeopardized. During their stay in the internat, the children are socially and emotionally excluded from the larger society; and they also become dependent on the institution. So, when they leave the internat as a young adult, most are unable to cope with the outside world. There are few opportunities for further education or meaningful employment. It is not surprising that many become involved in criminal activity. 1.20 Older people are also more likely to be poor than other adults, with 22.7 percent of people aged over 65 living in the bottom quintile of per capita household expenditure, compared with 20.8 percent of persons aged 16-64. However, older people make up only a small proportion of the total poor (4.1 percent). The risk of poverty is greater for people living in a household containing one older person as compared with a household containing none or two. Most older people live with younger household members, and single elderly person households are very rare, comprising less than 1 percent of all households. Households with three or more people aged 65 and over are also very unusual in Tajikistan (constituting only 2 percent of all households). However, people living in such households are significantly more likely to be poor than other types of household with a relative risk of poverty of 3.5. Interesting, if these households were not poor, then they are located at the other end of the spectrum i.e. in the top quintile. This may reflect that fact that the receipt of three pensions, when paid, is sufficient to push a household a long way up a relatively flat distribution. In Russia, poverty was found to be inversely related to age. However in Tajikistan the relationship between age and poverty is far from straightforward. Figure 3 shows that poverty rates amongst the very old (aged 75+) are higher than even those for children. Therefore, elderly people living alone, although numerically very small, may be amongst the poorest of the poor. - 24 - Figure 3: Proportion of Individuals who are Poor by Age Group 28 26__ 24__ . 22 20 1 8 12 0-9 10-19 20-34 35-54 55-74 75+ Source: TLSS 1.21 Again it is useful to compare the TLSS results with information obtained from other, qualitative sources. Many people in Tajikistan do believe that pensioners and the elderly are a particularly vulnerable group, but they also realize that they are not a homogeneous group. In total, about 548,000 people (9.2 percent of the entire population) receive pensions, and of them about 65 percent (354,700) receive a pension of less than TR 2,000 per month. About 10 percent of pensioners (56,000) receive a social pension (allowance) because they did not work and qualify for a full pension. On the other hand, some people are still able to retire early because they provided a special service to the state, for example, through working in a "strenuous" industry. Overall, the Ministry of Social Protection (MSP) considers about 9,000 single and disabled pensioners to be especially vulnerable. The MSP provides assistance (food, clothing and coal) to about 6,000 members of this group in their homes or residential centers. Some of the single pensioners are Russians whose families have emigrated to other parts of the FSU. 1.22 In general terms the TLSS results do not seem to show any significant difference in the poverty rates for men and women in Tajikistan. However, poverty is here defined by the expenditure of the household and assumes that all household resources are shared equally among their members. This may not be a realistic assumption in Tajikistan, and there is some evidence that the circumstances of transition may have tended to increase gender-based disparities within the household rather than reduce them. In 1996, UJNHCR and Relief International conducted a survey to assess the situation of women in the country. That survey concluded, for example, that the collapse of the social allowances and services has impacted women's lives more significantly than those of men, since the burden of caring for children and the elderly fall disproportionately on women. There is also some evidence that - 25 - women's wages have fallen more than men's. In the Soviet period a high proportion of public sector workers were women (especially in education and health). These are the sectors now where wages have often not been paid and where real pay rates have suffered the greatest fall in value. The greater decline in the relative value of women's wages may, therefore, mean that the proportion of household resources 'enjoyed' by women and children is declining. Figure 4. Proportion of Individuals who are Poor by Household C omposition 20- , 1 ad + kids 2 ads 2 ads, 1-2 2 ads, 3+ 3+ ads 3+ ads, 1-2 3+ ads, 3 kids kids kids Idds Source: TLSS 1.23 There is also a particular problem of female-headed households. The civil war created approximately 25,000 female-headed households, predominately in Khatlon and Gharm. Some women lost their husbands during the war. Others lost their husbands to emigration. A few of the latter group abandoned their families and created new families. In all of these cases, the woman became the head of the household. According to the Save the Children 1998 Socio-Economic Survey of Households, Farms and Bazaars, female-headed households have less access to land, irrigation and livestock. They are also less food secure, but receive more humanitarian assistance. Even with this assistance, their monthly income is less than male-headed households. Table 7 confirms from the TLSS dlata that individuals living in female-headed households experience a greater risk of poverty (28.6 percent) than those in male headed households (21.2 percent). However, such households account for only a fifth of all poor people. - 26 - 1.24 The civil war displaced over 10 percent of the country's population, the war-related displacement was relatively temporary, with most people (95 percent of the internally displaced people and 65 percent of the refugees) returning by the end of 1995. The process was completed after the signing of the peace agreements in July 1997. However, those who had migrated for economic reasons did not return as the situation in the country did not improve significantly. A few also did not return for political reasons. A 1998 UNHCR survey in Khatlon, where most of the displacement occurred, compared earlier returnees (arriving before July 1997), recent returnees (arriving after July 1997) and non-returnees. The key findings from this survey are that: (i) returnees have larger homes, but fewer inhabitable rooms (due to the destruction of some 35,000 homes during the war); (ii) returnees have access to less agricultural land, sow less wheat and produce smaller yields; (iii) returnees have less livestock now, but had more livestock before the war; and (iv) the majority of the recent returnees, most of the earlier returnees and about half of the non- returnees receive humanitarian assistance. In general, there is considerable agreement among the international community that the repatriation of refugees and internally displaced people in Tajikistan has been a success because the Government and Tajik people wanted the situation stabilized, and humanitarian assistance was linked to rehabilitation and development. 1.25 Poverty by Education and Labor Market Status. As is the case in most regions of the world, the risk of being poor in Tajikistan appears to be inversely related to education. According to the TLSS, individuals aged 16 and over with no schooling experience a relative risk of poverty of 1.23 (Table 8). Individuals with primary education or general secondary education are no more likely to be poor than on average. However, individuals with vocational or specialized secondary education are less likely to be poor than on average, with a relative risk of 0.76; and those with some higher education were much less likely to be poor, with a relative risk of 0.46. Thus persons with no education are over two and a half times as likely to be living in a household with a per capita expenditure in the bottom quintile as those with higher education. Overall, however, individuals with no education make up a very small proportion of the poor - a reflection of the fact that less than 2 per cent of the population have no formal schooling. The vast majority of the adult population has at least general secondary education; and so do the majority of the adult poor. 1.26 In contrast to education, there is no clear relationship between labor market status and poverty in Tajikistan. Table 8 shows that a lower proportion of those in the bottom quintile live in a household where the head is employed (61.1 percent) than on average (63.4 percent); but also a lower proportion live in a household where the head is unemployed (5.3 percent) than on average (6.3 percent). The group with the highest risk of poverty are those where the head is out of the labor market, but not retired. Individuals in these households experience a relative risk of poverty of 1.49, and constitute 11 percent of the poor. It may therefore be that labor market status is no longer a good indicator of access to resources. It is not whether a person has a job that is important, but whether a person has a job and receives a salary or wage that is the critical factor. It is also important to note that income from the labor market accounts for only a third of the total income of households in the bottom quintile of consumption, and only just under a half for those in the top quintile. - 27 - Table 8: Poverty Incidence by Education and Labor Market Status. Poorest 20 2 3 4 Richest 20 percent percent All individuals 22.4% 21.7% 20.5% 19.1%go 16.3% Education Age less 15 24.2% 22.2% 20.3% 18.9%/o 14.5% Primary 22.3% 21.8% 21.4% 18.5% 16.0% Secondary 22.6% 21.5% 20.4% 19.1%/0 16.3% Vocational 17.1% 21.1% 22.3% 20.3%g4 19.2% Higher Household Head Labor Market Status Employed 21.6% 21.7% 19.8% 19.1 %0 17.8% Unemployed 18.6% 23.6% 27.4% 19.6% 10.8% Not in work force 33.4% 19.8% 15.1% 16.1%to 15.6% Retired 22.2% 21.7% 23.5% 19.1 % 13.5% No. of Adults Unemployed Zero 21.4% 21.6% 20.4% 19.5%Xo 17.1% 1-2 27.5% 19.9% 23.4% 15.8% 13.3% 3 or more 22.5% 27.5% 16.1% 21.7% 12.2% Source: TLSS 1.27 Poverty by Income and Expenditure Patterns. Labor incomne remains the most important source of income for all households. However, the combined income from the sale of foodstuffs and households assets, on average, accounts for a similar share. Remittances are also very important, whilst the role of the state in providing a safely net is very small. Even amongst those households in the bottom fifth of the distribution of expenditure, social transfers only account for eight percent of income. Table 9: Structure of Total Household Income by Quintile Group. oPorest 2 3 4 Richest 20 20 percent percent Labor income 32 34 34 38 42 Saleoffoodandcrops 16 18 17 14 10 Sale of private belongings 14 13 18 16 15 Sale of commercial goods 1 3 2 2 3 Rent 1 - - I I Withdrawal of savings 2 2 2 3 3 Remittances 14 12 9 10 12 Aid 8 7 7 15 4 Social transfers 6 6 5 5 5 Other 7 5 5 6 6 Total 100% 100% 100% 100% 100% Source: TLSS Note: the table excludes the imputed value of home production. - 28 - 1.28 As might be expected, the share of total household expenditure on food is higher for those households at the bottom of the distribution than at the top, with the poorest household spending 79 percent of all expenditure on food. The imputed value of home production contributes a significantly greater proportion of total food expenditures for the poorest groups, as does the value of food gifts including humanitarian aid, whilst the share of expenditure on the cash purchase of food varies little across the distribution. Expenditures on 'other' items, which includes clothing, books, durables, holidays and weddings, also varies sharply between the rich and poorest households. 1.29 Ownership of Assets and Infrastructure. In Tajikistan, as in other countries of the FSU, in the past there was little or no relationship between a household's ownership of consumer goods and its level of income. This is because under the Soviet Regime, consumer durables were allocated by the command economy rather than by the market economy. However, since independence the sale of household assets has emerged as a key household coping strategies. The TLSS results do indeed show that there is a significant relationship between household per capita expenditure and a range of durables. The strongest correlation is between ownership of a color television, followed by stereo and then video recorder. While this situation is partly due to the differential acquisition of consumer durables by the better-off in the recent past, the TLSS data show that overall there has been a relatively low level of consumer spending throughout the Republic. Under five percent of households in Tajikistan have purchased a major household appliance (refrigerator, vacuum cleaner, washing machine) in the last 7 years, indicating the widespread nature of financial hardship. The most popular item purchased was a stereo/cassette player. 1.30 The majority of households plot (79 percent) in Tajikistan have access to an individual garden, even though many of them are very small; and, in fact, access to a plot is inversely related to poverty, with a higher proportion amongst the bottom quintile having a garden plot than amongst the top quintile. For the less poor groups, having a garden plot is less essential, whereas for the poor it is a critical part of their coping strategy. However, each individual in the poorest quintile has an average of only 5.5 sotkas per head, compared with 11.5 sotkas in the top quintile. Similarly ownership of at least one cow is greater amongst the rich than the poor. 1.31 Other interesting findings from the TLSS include the following: (i) 7 percent of households report that their home was damaged during the war, of which a quarter experienced significant damage and a third almost completely destroyed; (ii) less than half of all households have access to piped water, nearly a quarter are reliant on water from river/lake /ponds, and a further eighth on spring water; (iii) of those who have piped water, a quarter reported that water was only available for five hours a day or less; and only 36 percent reported 24 hour availability; (iv) only a half of households reported that their water quality was good/excellent and a half reported fair/poor; (v) 75 percent of households reported no source of hot water; (vi) only 14 percent of households have a flush toilet and 85 percent rely on an outside latrine; (vii) the. most common source of fuel used by households for cooking is wood (43 percent), followed by manure (17.5 percent) and cotton stem (12.3 percent); (viii) a third of households had only heated their home for 3 months or less in the last year; two-thirds heated it for 4 months or less; and (ix) 14 percent of households have a phone inside the dwelling, 17percent rely on neighbors, and 54 percent have no access to a phone. - 29 - Table 10: Correlates of Household Poverty Demographic variables Number of children in the household 1.32 ** Number of adults in the household 1.17 ** Region ** Dushanbe 1.00 GBAO 7.00 ** RRS 1.87 Leninabad 3.38 ** Khatlon 4.79 ** Socio-economic variables Number of adults employed in the household 0.85 * Head of household educational group * None 1.00 Primary (le 7 year) 0.91 General Secondary (8-12 years) 1.00 Vocational/Specialized 0.68 Some Higher 0.48 # Other 1.19 Ownership of consumer durables Stereo/tape player Yes 0.39 Color TV Yes 0.37 Radio Yes 0.52 ** Sewing machine Yes 0.60 ** Car Yes 0.42 * Bicycle Yes 0.45 ** New Refrigerator (since 1992) Yes 0.23 # Household assets Sotkas of land per capita 0.98 * Number of cows Zero 1.00 One 0.88 Two 0.48 ** Three or more 0.35 ** Housing Gas supply Yes 0.46 * Coal/peat/wood stove Yes 0.64 # Note: ** significant at (p<0.001), * significant at (p<0.01), # significant at (p<0.05). 1.32 Multivariate Analysis of the Correlates of Household Poverty. Many of the correlates of poverty discussed above are related to each other. In order to establish the strength of their relationship after controlling for other factors, a logistic regression was carried out on the TLSS data, with the independent variable beingr whether or not the household is poor as defined by being in the bottom quintile of per capita household expenditure (Table 10). This model does not allow conclusions about causation: for example, it may be that ownership of a durable or assets, such as a car or cattle, facilitates a - 30 - livelihood strategy that protects the household against being poor; or it may be that ownership of a durable (or lack thereof) reflects the purchasing power of the household. Nevertheless, the correlates of poverty (both the strength of the relationship and also the statistical significance) are still interesting. D. Coping Strategies 1.33 Households with limited resources are increasingly employing a range of different strategies to survive on limited resources. The most basic necessity within any household is food, and by far the most important coping strategy with regard to ensuring its supply is its self-production. 84 percent of all households reported having access to an individual garden plot and 72 percent of households reported consuming food grown by the household in the last 7 days. Access to land is therefore a critical factor in many households' survival. Humanitarian assistance is also important, as 23 percent of all households reported the receipt of food gifts in the last seven days and their imputed value accounted for a tenth of total expenditure for the poorest households. However, what is most striking is the widespread nature of behavior change within Tajikistan. Even amongst the most well-off households, nearly 30 percent reported having reduced the number of meals a day, and a similar proportion reported eating smaller portions. This rose to over 60 percent amongst the poorest households. An indication of the pervasive nature of financial insecurity for households across Tajikistan is the fact that over a quarter of all households had sold assets in the last month, and a third had had to borrow from relatives, friends, and neighbors. These proportions were roughly constant for households across the entire distribution. In addition to the coping strategies already employed by households, respondents claimed that they would use a variety of coping strategies over the next six months. A higher proportion of households thought that they would have to modify their diet still further and/or find other work A quarter thought that they would have to sell household assets and over a quarter would have to borrow to make ends meet. Two percent thought that they would have to resort to begging. E. Food Security 1.34 The information presented on household coping strategies suggests that financial hardship is far from being limited to the poorest group. This conclusion is supported by information on food security and insecurity. In particular, it is possible to compare the TLSS data with the results of a nation-wide survey conducted by ECHO in June 1997 to assess the profile and location of food insecure households. According to the TLSS, on average households in Tajikistan in 1999 ate 2.5 meals a day. This is only a slight reduction from the 2.6 reported by the ECHO Food Security Survey in 1997. The average number of meals a day varied from 2.2 for the poorest households to 2.8 for the richest. However, most worrying from a nutritional point of view is the rise in the proportion of households claiming to eat just one meal a day from 10 percent in 1997 to 13 percent in 1999. Over a quarter of the poorest households ate an average of one meal or less a day, and over half ate two or less. Alarmingly of all households eating less than two days a day, over 50 percent stated that they thought in the next 6 months that they would have to reduce the number of meals eaten still further. - 31 - Figure 5: Food Security (kg of Flour in Stock per Household by Household grouip) 20 14 zz/ kg 10 zE 8-/ Poorest 2 3 4 Richiest 20% 20% Source: TLSS 1.35 There has also been a significant change in the diet of the average Tajik since independence, with a shift towards cheaper foodstuffs. Traditionally the mainstay of the Tajik diet was soup - made with a variety of ingredients including vegetables and meat as well as noodles. In addition vegetables stuffed with rice and meat, plov (a rice and meat dish) or meat and potatoes were consumed several times a week, particularly in works canteens. Today, the TLSS results suggest that meat, eggs and confectionery have become luxury items that even the richest only eat a few times a week and that the poorest rarely have access to. Over half of all households had not eaten any meat in the 7 days prior to the survey, 61 percent had not eaten eggs and 85 percent had not eaten confectionery. Table 11: Proportion of Households Reporting Engaging in Selected Coping Strategies in the Last 6 Months. Food Security % of households shifting to cheaper foods 85 % households reducing number of meals per day 44 % households selling assets to buy food 28 % households borrowing money to buy food 34 % households no meat consumption preceding week 53 1.36 Food security is difficult to measure precisely, especially in a one-off survey, but some general conclusions may be drawn about food insecurity at a population level in Tajikistan. Food was the principal reported household expenditure over the month preceding the interview. Bread is the foundation of the diet, and reported protein consumption (meat, dairy, pulses) is low. Meat, a culturally significant part of the Tajik diet, is now rarely eaten; slightly over half of respondent households had eaten no meat in the seven days preceding the survey. Other than reducing the number of meals eaten daily, and reducing portion size (responses which are difficult to objectively evaluate), coping strategies showed little - 32 - variation between quintiles, with about one third of all households having to incur debt in order to buy food, for example. Table 12: Household Food Security and Coping Strategies by Quintile Bottom Top Tajikistan Shifted to cheaper foods 84% 84% 85% Reduced # of meals per day 61% 29% 44% Ate smaller food portions 63% 29% 47% Sold household assets 29% 27% 28% Borrowed 37% 30% 34% Begged 7% <1% 2% 1.37 The TLSS respondents were asked a range of questions concerning their current stock of various foodstuffs, and their perception concerning the household's position with regard to food over the next 6 months and basic necessities in the next 12 months. The store cupboard in most Tajik households appears to have been almost bare in May 1999, with very low per capita stocks of dried fruits and beans and no stocks of preserved fruits. This is perhaps not surprising as the survey was conducted after winter and spring, and before the main harvest period. There is a very strong correlation between food stocks and poverty, highlighting that the poorest households have little cushion by way of stored food. Although the survey was conducted before the harvest, households were not uniformly optimistic that their situation with regard to food would improve in the coming months. Just under a half of all households thought it would definitely get better, but this fell to only 30 percent for amongst the poorest households. This may reflect that even if food availability increases with the harvest, many households felt uncertain that their access to food would improve. In fact, just under a fifth of all households thought that their food situation would get worse, and this rose to a quarter amongst the poorest. F. The Development of a Pro-Poor Country Development Strategy 1.38 In order to develop a poverty alleviation program for Tajikistan, the Government has established a Working Group headed by the Deputy Chairman of the SSA. Other members include the Deputy Chairman of the Social Protection Fund and senior officials in the President's Office and various ministries. The Working Group has already started the process of analyzing data on poverty in the country. The main recommendations which have so far come out of the deliberations of the Working Group and which are endorsed by the analysis here are the following: (i) there is clearly a massive problem of poverty, although the extent of it depends crucially on the poverty line used; (ii) more work needs to be done to develop better indicators of poverty and there is also an urgent need to identify the poorest groups; (iii) it will be critical to monitor trends in living standards and poverty over time, including through the carrying out of repeat surveys using the TLSS methodology; - 33 - (iv) national capacity needs to be developed further, both in terms of data collection and analytical skills; and (v) the Working Group on poverty identification and monitoring should be supported and assisted in its activities. - 34 - CHAPTER II. EMPLOYMENT AND POVERTY 2.1 The previous chapter has shown that poverty in Tajikistan is a widespread and complex phenomenon. However, particularly with the cessation of conflict and violence in the country, the resumption of economic growth in the last couple of years, and the progress being made in the restructuring of the economy, there are some grounds for optimism that poverty can be significantly reduced over the coming years. It is suggested in this report that there are four key pillars for a poverty alleviation strategy. These relate to: (i) the stimulation of high labor productivity and economic growth; (ii) the provision of basic social services;