Report No. 26213-AL Albania Poverty Assessment November 5, 2003 Human Development Sector Unit Europe and Central Asia Region Document of the World Bank CURRENCYEQUIVALENT (Exchange Rate Effective April 30,2002) Currency Unit = Lek Lek 1=US$0.007 US$1 =Lek 145 GOVERNMENT'S FISCAL YEAR January 1 - December 31 ABBREVIATIONSAND ACRONYMS CAS Country Assistance Strategy CBN Cost of Basic Needs ECA Europe and Central Asia ESW Economic and Sector Work EU European Union GDP Gross Domestic Product GPRS Growth and Poverty Reduction Strategy (Government PRSP designation) IBRD International Bank for Reconstructionand Development IDA International Development Association ILO International Labor Organization IMF International Monetary Fund INSTAT Institute of Statistics LCS Living Conditions Survey LFP Labor Force Participation LSMS Living Standards Measurement Survey MOA Ministry of Agriculture and Food MOES Ministry of Education and Science MOF Ministry of Finance MOH Ministry of Health MOLG Ministry of Local Government MOLSA Ministry of Labor and Social Affairs MTEF Medium Term Expenditure Framework NES National Employment Service NFRE Non-Farm Rural Enterprise NSSED National Strategy for Socio-EconomicDevelopment (akaGPRS) OECD Organization for Economic Cooperationand Development PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper UBN Unmet Basic Needs USAID United States Agency for InternationalDevelopment Country Director: Orsalio Kalantzopoulos Sector Director: Annette Dixon, Armin Fidler Sector Manager: Mamta Murthi, Arup Banerji .. 11 Table of Contents ACKNOWLEDGEMENTS ............................................................................................... XI EXECUTIVE SUMMARY .............................................................................................. xi1 CHAPTER I: INTRODUCTION ......................................................................................... 1 1 BACKGROUND: WHY A POVERTYASSESSMENT FORALBANIA? 2 THE ECONOMIC CONTEXT .................................................................................... 3 3... ..................1 MEASURING POVERTY INALBANIA .................................................................... 6 4. STRUCTURE OF THE REPORT ............................................................................... 9 CHAPTER 11:POVERTY INALBANIA .......................................................................... 10 1. INTRODUCTION ..................................................................................................... 10 2. INCOMEPOVERTY ................................................................................................ 10 3. NON-INCOME DIMENSIONS OF POVERTY ........................................................ 14 4 22 5.. SUBJECTIVE POVERTY ......................................................................................... A PROFILE OFPOVERTY ...................................................................................... 24 6. DECOMPOSINGINCOMEINEQUALITY ............................................................. 33 7 35 8 36 9... CONSUMPTIONPATTERNS .................................................................................. DETERMINANTS OF POVERTY ............................................................................ CONCLUDING REMARKSAND POLICY IMPLICATIONS ................................. 37 CHAPTER 111: SPATIAL DIMENSIONS OF POVERTY ........................................ 40 1 OVERVIEW . .............................................................................................................. 40 2. POVERTY INTHE REGIONS ................................................................................. 41 3 RURALPOVERTY ................................................................................................... 42 4 RURALAND MOUNTAINAREAS -POCKETSOF SEVEREPOVERTY ............44 5... POVERTY INPERI-URBAN AREAS ...................................................................... 47 6. POVERTY AND INEQUALITYMAPS .................................................................... 49 CHAPTER IV: LABOR MARKET DEVELOPMENTS AND POVERTY ........................ 55 1 OVERVIEW . .............................................................................................................. 55 2. EMPLOYMENT ....................................................................................................... 57 3 UNEMPLOYMENT .................................................................................................. 62 4 WAGE DEVELOPMENTS ....................................................................................... 65 5... LABOR MARKET AND POVERTY ........................................................................ 69 6. CONCLUDING REMARKS AND POLICYIMPLICATIONS ................................. 71 111 ... CHAPTERV: HEALTH ................................................................................................... 73 1 OVERVIEW .............................................................................................................. 73 2 STATUSOFPOPULATIONHEALTH .................................................................... 74 3 HEALTHINSURANCE ............................................................................................ 76 4 HEALTHCAREUTILIZATION .............................................................................. 77 5 EXPENDITURESONHEALTHCARE ................................................................... 81 6 CATASTROPHICHEALTHEXPENDITUREAND POVERTY ............................. 85 7 DETERMINANTSOFHEALTHCAREDEMAND ................................................. 87 8........CONCLUDINGREMARKSAND POLICYIMPLICATIONS ................................. 87 CHAPTERVI: EDUCATION ........................................................................................... 91 1 OVERVIEW 91 2 EDUCATIONALATTAINMENT ............................................................................. 91 3... .............................................................................................................. REASONSFORLOW ENROLMENT ...................................................................... 97 4. EXPENDITUREONEDUCATION .......................................................................... 98 5 EDUCATIONAND LABORMARKETS ................................................................ 100 6.. CONCLUDINGREMARKSAND POLICY IMPLICATIONS ............................... 103 CHAPTERVII: COPINGWITHPOVERTY-THE ROLE OF SOCIALPROTECTION PROGRAMS ................................................................................................................... 105 1 INTRODUCTION . ................................................................................................... 105 2. IMPACT OF ALBANIA'S CASH SOCIAL PROTECTION PROGRAMS 108 .. 3. NDIHME EKONOMIKE: KEY ISSUES ................................................................. 117 CHAPTERVIII: COPINGWITHPOVERTY-ROLE OFINTERNAL AND EXTERNAL MIGRATION .................................................................................................................. 121 1 OVERVIEW 121 2 BACKGROUND ...................................................................................................... 121 3 TYPESOF MIGRATION ....................................................................................... 124 4.... ............................................................................................................ REMITTANCES ..................................................................................................... 131 5. CONCLUDINGREMARKSAND POLICY IMPLICATIONS ............................... 133 CHAPTERIX: LOOKINGAHEAD ............................................................................... 135 1 REDUCINGPOVERTY . .......................................................................................... 135 2. AREAS FORFUTURERESEARCH ...................................................................... 140 iv List of Tables 3 Table 1. 2: GDP and Sectoral Growth................................................................................................ Table 1. 1: Macroeconomic Indicators. 1998-2002 ............................................................................. 4 4 Table 1.4: Employment and Unemployment (`000 unless otherwise indicated) .................................... Table 1. 3: Exports and Imports (percent o f total) ............................................................................... 5 Table 2. 1: Poverty and Inequality inAlbania................................................................................... 11 Table 2. 2: Comparing poverty lines................................................................................................ 12 Table 2 3: Sensitivity of headcounts to changes inpoverty line........................................................ Table 2. 4: Poverty rates in selected countries inthe ECA region....................................................... . 13 14 Table 2. 5: Self-assessmentof water quality ..................................................................................... 15 Table 2. 6: Electricity ..................................................................................................................... 16 17 Table 2. 8: Literacy by age group and location ................................................................................. Table 2. 7: Unmet BasicNeeds ....................................................................................................... 17 Table 2.9: Educational attainment................................................................................................... 18 Table 2. 10:Enrolment rates by location.......................................................................................... Table 2. 11: Moderate and severe malnutrition inchildren under five years of age.............................. 18 19 Table 2. 12:Antenatal care ............................................................................................................. Table 2. 13: Labor Force Participation o f Working Age Populationby Gender (Age > 14).................21 20 Table 2. 14: Labor Force Participation: Gender Differences By Region (Age > 14) ............................ Table 2. 15: Ownershipand size of agricultural land......................................................................... 21 22 22 Table 2. 17:Value of durable goods by location............................................................................... Table 2. 16: Ownership of livestock among households engagedin agriculture .................................. 22 Table 2. 18: Value of durable goods by economic status................................................................... 22 Table 2. 19: Satisfaction with financial situation and mean consumptionper capita............................ 23 Table 2. 20: Comparing objective and subjective poverty.................................................................. 24 Table 2. 21: Maincharacteristics of the poor.................................................................................... 25 Table 2. 23: Poverty, livingconditions and access to services ........................................................... Table 2. 22: Poverty Measures and Shares by Age Group and Location............................................. 26 26 Table 2. 24: Net enrolment by economic status ................................................................................ 28 Table 2.25: Poverty Measures and Shares by Educational attainment and Location............................ 29 Table 2.27: Labor Force Participationfor Poor andNon-poor (Age > 14) ........................................ Table 2.26: Poverty and child malnutrition...................................................................................... 30 31 Table 2.28: Poverty measuresby employment status o f household head............................................ 31 32 Table 2. 30: Income shares by poverty status (percent) ..................................................................... Table 2. 29: Poverty Ratesby Number o f Employed People and Location ......................................... 32 Table 2. 31:Incidence of poverty by main source of income (at least 50%) ........................................ 33 Table 2. 33: Gini decomposition by income source (non-agricultural income assumedto be zero) ....... 34 Table 2. 32: Gini decomposition by income source........................................................................... 34 35 Table 2. 35: Sources of consumption by location.............................................................................. Table 2. 34: Main budget shares by location .................................................................................... Table 2.36: Mainbudget shares by economic status......................................................................... 35 35 Table 2. 37: Food expenditure patterns by location........................................................................... 36 V Table 3. 1:Poverty rates by Region................................................................................................. 41 Table 3.2: Sources of Income by Urbanmural and RuralQuintiles (percent) ..................................... ................................... 42 43 Table 3. 4: Ownership of Agricultural Land (percent)....................................................................... Table 3.3:Average Value of HouseholdIncomeperMonthby Source (Leks) 43 Table 3. 5: Distribution of Land. by Land Size and Region. Quintile and Poor/Non-Poor (percent)......44 Table 3. 6: Poverty and inequality indices (%)................................................................................. 50 Table 3.7: Poverty and inequality indices by PREFECTURE(%) ..................................................... 51 Table 3. 8: Poverty and inequality indices by DISTRICT (%)............................................................ 51 Table 3. 9: Poverty and inequality indices by Mini-Municipality of Tirana City (%) ........................... 53 Table 4. 1:Labor force participation by gender and education(Working age population. age 14+) ...... 56 Table 4. 2: Labor force participation by region(working age population. age 14+) ............................. Table 4. 3: Employment by sectors (end of period) ........................................................................... 56 58 Table 4. 4: Employment by gender andeducation(percent of working age population. age 15-64) ...... 58 Table 4. 5: Employment by region (percent of working age population. age 14+) ............................... 59 Table 4. 6: Duration of wage employment........................................................................................ 59 Table 4. 7: Type of employment by employer .................................................................................. 60 Table 4. 8: Albania employment data. 2002 ..................................................................................... 61 Table 4.9: Non-farm householdenterprises by industry.................................................................... Table 4. 10: Labor force and unemployment. 1991-2001(thousands) ................................................ 62 62 Table 4. 11:Unemployment by gender andeducation(percent)......................................................... 63 Table 4. 12:Labor force and unemploymentby gender andeducation(percent of working age Table 4. 13: Labor force and unemploymentby region (percent of working age population. age 14+) ..64 population. age 14+) ................................................................................................................ 63 Table 4. 14: Average monthly wage by industry. public and private sector......................................... Table 4. 15: Average monthly wage by employer by region (inLeks) ........................................ 66 66 66 Table 4. 17: Reportedaverage monthly wages by industry. 2002....................................................... Table 4. 16:Average monthly wage by sector and education(Leks) .................................................. 67 Table 4. 18: Private/public wage differential by educationalattainment ............................................. Table 4. 19: Poverty rate by the number of earners ina household..................................................... 68 69 Table 4.20: Labor market characteristics (percent ofworking age population. age 14+) ..................... 70 Table 4.21:Average monthly wage by poverty status ...................................................................... 70 Table 5. 1:Health Care Services...................................................................................................... 73 Table 5.2: Life expectancy at birth.................................................................................................. 74 Table 5. 3: Causesof Mortality by groups of diseases (per 100.000 population) ................................. 74 Table 5.4: Self-reportedmorbidity across gender. regions. and poverty levels ................................... 75 Table 5. 5: Outpatientcare utilization (percent ofthose ill)............................................................... 78 Table 5. 6: Health care personnelper 1000habitantsacross districts.................................................. 79 Table 5. 7: Outpatient utilization rate amongthose reporting an illness across poverty status and age...79 Table 5. 8: Antenatal care by poverty status and regions ................................................................... 80 Table 5.9: Assistanceat birth by type of provider. across regionsand poverty status (proportions)......80 Table 5. 10:Public Expenditure on Health....................................................................................... 81 Table 5. 11:Outpatient care expenditure by regions ......................................................................... 82 vi Table 5 12: Yearly total hospital expenditure .................................................................................. . Table 5 13: Average informal payments inout-patient care among those makinginformal payments 83 Table 5 14: Per capita monthly health expenditure and share of health expenditure on total household .. ..83 per capita expenditure across consumption quintiles .................................................................. 85 Table 5 15: Proportion o f people with catastrophic health expenditure Table 5 16: Catastrophic headcount and catastrophic excess measures for healthexpenditure 86 Table 5 17: Poverty measuresbefore and after healthcare payments ... .............................................. .............85 ................................................. 86 Table 5.A. 1: Chronic diseases reported across poverty status ........................................................... 89 Table 5.A. 2: Reported illness inthe past four weeks across poverty status, percentages ..................... 89 Table 5.A. 3: Health care utilization across gender. region of residence. poverty status. and whether or not the person has a health license. percentages ......................................................................... 90 Table 5.A. 4: Percentage of people seeking care across regions and age groups .................................. 90 Table 5.A. 5: Pre-natal consultationby type o f provider across regions and poverty status (proportions) ............................................................................................................................................... 90 Table 6. 1: Highest diploma obtained across regions ......................................................................... 92 Table 6 2: Net enrolment rates across gender. regions. and income levels .......................................... 93 Table 6 3: Gross enrolment rates in secondary school .. ...................................................................... 95 Table 6.4: Proportion o f people that finished 8 years o f basic education and continue to secondary school ..................................................................................................................................... 96 Table 6. 5: Reasonsfor not attending pre-school across regions. proportion o f 3-5 years old ............... 97 Table 6 6: Reasons for non-enrollment among 14-17 years old (proportions) . ............................ Table 6. 7: Public expenditure on education as percentage o f GDP and total public spending ..............98 Table 6 8: Total household expenditure ineducation across levels and consumption quintiles 99 Table 6 9: Individualtotal expenditure ineducation by school level and income 99 Table 6 10: Labor force status of Albanian men across education levels. percentages....................... Table 6 11: Women's labor force participation across education levels and age Table 6 12: Mean monthly earnings ofpaid workers across education levels. gender. and age ..... ............98 ................................ 101 ............................... ..........101 102 Table 7. 1: Sources of Income (Mean Income; Leks per Month) ...................................................... 109 Table 7 2: Structure of Albanian HouseholdIncome. by Source. Location of Residence. and Income . Quintile for Location (Percentage of total income for household category) ................................ 110 Table 7 3: Distribution o ftransfer income across households receiving transfer income from that source . (percentage of households receivingthat income source) .......................................................... 111 Table 7 4: Incidence Among Households of Receipt of Transfer Income. by Quintile. Poverty Status. . Area of Residence and Source of Transfer Income (percentage o f households within category) .. 112 Table 7 5: Estimatedimpact of income transfers on post-transfer rates o f poverty. ifspecified sources o f . transfer income had not existed. by source o f transfer and assumed share of transfers usedto increase consumption. for all households in 2002..................................................................... 116 Table 7 6: Estimatedimpact of income transfers on post-transfer rates o f extreme poverty. if specified . sources of transfer income had not existed. by source o ftransfer and assumed share o ftransfers used to increase consumption. for all households in 2002 ................................................................. 117 Table 8. 1: Internal mobility: household members born in current municipality ............................... 124 vii Table 8.2: Movement ofhouseholdheads inthe 1990s................................................................... 125 Table 8. 3:Householdcharacteristics ofrecent temporary migrants................................................. 128 Table 8. 5: Location of Remittersand Recipients............................................................................ Table 8.4: Individual andHousehold characteristics of children who left home............................... 130 132 Table 9. 1: Share ofIncome by Income Sources (percent. rural populationonly) .............................. 138 ... v111 Listof Figures Figure2.1: Cumulative Density Functionsby location....................................................... 12 Figure2.2: Access to water....................................................................................... Figure2.3: Sanitation.............................................................................................. 15 15 Figure2.4: Enrolment by gender................................................................................ 20 23 Figure2.6: Household basic needs requirement............................................................... Figure2.5: Satisfaction with current financial situation...................................................... 23 Figure 2.7: Poverty rates by household size and location ................................................... 25 Figure2.8: Enrolment andPoverty.............................................................................. 28 Figure2.9: Educational attainment by economic status and location....................................... 29 Figure2.10: Educational attainment ofthe poor by gender.................................................. 30 Figure3.1: Income cumulative density functions by region................................................. 41 Figure 3.2: Poverty Headcount by Prefecture by District.................................................... 54 Figure 3.3: Per Capita Consumption............................................................................ 54 Figure3.4: Poverty Headcount by Communes andMunicipality........................................... 54 Figure4.1: Albania labor force participation rates, 1993-2001.............................................. 55 Figure4.2: Albania employment rates, 1991-2001............................................................ 57 Figure4.4: Composition ofthe unemployed................................................................... Figure 4.3: Duration of employment inthe informal sector (percent o ftotal employment) ..............61 64 Figure4.5: Realwage development, 1993-2001............................................................... 65 Figure4.6: Average monthly wage by sector and gender (in Leks)......................................... 68 Figure 5.1: Population o fthe people with health license..................................................... 77 Figure6.1: EducationalAttainment (cohort 16-35)........................................................... 92 Figure 6.2: Percentageo f students receiving tutoring ...................................................... 100 Figure7.1: Distribution o fHousehold's Transfer Income (from all sources, by source)...............113 Figure 7.2: Distribution ofNon-Poor Household's Transfer Income (from all sources, by source of income) ................................................................................................... 113 Figure 7.3: Distribution o f Poor Household's Transfer Income (from all sources, by source of income). .................................................................................................. 114 Figure 7.4: Distribution of ExtremelyPoor Household's Transfer Income (from all sources, by 114 Figure8.1: Internalmobility ofhousehold heads............................................................. source of income) ...................................................................................... 125 Figure 8.2: Out-migrationby district........................................................................... 126 Figure8.3: Immigrationby district., .......................................................................... 126 Figure 8.4: Migration net inflow by district.................................................................. 127 129 Figure8.6: Remittances and GDP.,............................................................................ Figure8.5: Temporary external migration.................................................................... 131 ix List of Boxes Box 2.1: MillenniumDevelopment Goals....................................................................... 16 Box 2.2: No payphone? There is Vodafone!................................................................... 27 Box 4.1: Labor Market Definitions .............................................................................. 57 Box 4.2: Why is the unemployment rate so highinTirana?................................................. 65 Box 5.1: Informal payment for health care, peri-urban areas of Tirana and Durres...................... 84 Box 6.1:Access to school in peri-urban areas of Tirana and Durres.,..................................... 94 Box 6.3: Informal payments in education...................................................................... 100 Box 6.2: Determinants of secondary school enrolment ....................................................... 95 Box 7.1: Social Protection spending, 1998.................................................................... 107 Box 7.2: Cash transfer programs................................................................................ 118 Box 9.1: Investment climate inAlbania., ...................................................................... 136 Box 7.3: Target Efficiency....................................................................................... 118 Box 9.2: Administrative corruption inAlbania................................................................ 137 X ACKNOWLEDGEMENTS The Albania Poverty Assessment is a collective effort o f a team led by Mukesh Chawla (ECSHD). The multi-sectoral team comprised: Gero Carletto (poverty profile and migration), Maria Eugenia Bonilla Chacin (education and health), William M. Tracy (qualitative and quantitative analysis), Moukim Temourov (labor market), Verdon Staines (social protection), Mansour Farsad(macroeconomic analysis), Frauke Jungbluth and Julia Bucknall (rural development), Lorena Kostallari, Meritxell Martinez, and Nicole La Borde (program assistant) all from the World Bank, the Institute of Statistics o f Albania, INSTAT (poverty profile), the Ministry of Labor and Social Affairs of Albania, MOLSA (poverty mapping), and consultants, the Human Development Promotion Center and Andrew Mast (qualitative analysis). Maureen Lewis and Jean-Jacques Dethier were the peer reviewers. Orsalio Kalantzopoulos, Country Director, Albania, Annette Dixon, Director ECSHD (at the time of writing), and Mansoora Rashid and Mamta Murthi, Acting Sector Managers, HDE, ECSHD (at the time o f writing) provided valuable guidance and support. A number o f people contributed by commenting on various drafts and generously providing information: Kathryn Funk, Asad Alam, Ruslan Yemtsov, Edmundo Murrugarra, Cem Mete, Radwan Ali Shaban, Akiko Maeda, Nora Dudwick, Jan Rutkowski, Philip Goldman, David Hotchkiss, LindaPicchino, Hermine de Soto, Mansour Farsad, PierellaPaci and Alexandre Marc. The preparation of the poverty assessment would not have been possible without intensive collaboration with the Instituteo f Statistics ofAlbania (INSTAT). The Institute's staff not only provided the data onthe first round of the Albania Living Standard Measurement Survey 2002, they also worked together with the Poverty Assessment team in generating of the poverty profile. The team owes a special gratitude to the Institute of Statistics ofAlbania for cooperation and generous sharing of information. This report has benefited from the collaboration of other government agencies, and the donor community present in the country. Poverty mapping, used as a backgroundpaper for this assessment, was financed by the Albanian Ministry of Labor and Social Affairs, which also provided assistance for the LSMS program. The team i s grateful to the representatives o f the donor community who advised during the conceptualization stage o f the poverty assessment.Their views on the critical issues related to poverty in Albania helpedto identify and focus on the key aspects o f poverty that have been analyzed inthis report. xi EXECUTIVESUMMARY Albania has enjoyed strong economic performance inthe last five years, growing at an average o f above 7 percent annually for most of the period. Inflation remained at levels of 4 percent or less, exports and imports grew steadily, and both the current account and the domestically financed fiscal deficit improved significantly. Several structural reforms have also been carried out in the last five years, and have involved banking, land reforms and privatization. Almost all small and medium enterprises have been privatized and significant progress has been made in the privatization o f strategic sectors, like telecommunications. Regulation in the financial sector as well as core public sector functions has been strengthened, and the Government has adopted a new bankruptcy law and a deposit insurance scheme. Significant steps have also been taken in the area o f public administration, and a comprehensive civil service law has been adopted. Legal and judicial capacities have been strengthened, accountability rules have beentightened, and an anti-corruption action plan is under implementation. Despite the impressive performance o f the economy, however, poverty in Albania has remained high and per capita income, at around US$1,230 in 2002, has remained one of the lowest among transition economies. Inan effort to adopt policies and measures to share widely the benefits of growth and reduce poverty, the Government of Albania outlined a poverty alleviation strategy in the Interim Poverty Reduction Strategy Paper (2000) and developed it in detail two years later in the Growth and Poverty Reduction Strategy (GPRS), the precursor to the current National Strategy for Socio-Economic Development (NSSED). Inthe early stages of development, however, the NSSEDwas constrained by the availability of data on poverty, and while it did make good use of the existingquantitative and qualitative information, an accurate and complete account o f poverty was not available. In recognition of the data limitations, the first round of a five-year Living Standards Measurement Survey (LSMS) program was launched in 2002. This survey provides valuable information on a variety of issues related to living conditions of the people o f Albania, including details on income and non-income dimensions o f poverty inthe country, and forms the basis ofthe present assessmentofpoverty inAlbania. The Poverty Line The poverty line used for this assessment is the first absolute poverty line based on nationally representative household level expenditure data ever estimated for Albania, and is based on an actual consumption bundle o f a sample reference population drawn from the survey, which was then converted to a monetary measure using the cost-of-basic-needs methodology. This resulted in the choice of an absolute poverty line of 4,891 Leks per capita per month. Based on this measure, one-quarter of the Albanian population, or close to 780,000 individuals, fall below the poverty line. Extreme poverty amounted to about 150,000 persons, or less than 5 percent of the population, who were unable to meet even the basic food requirements. A large number o f individuals are clustered around the poverty lines, however. Increasingthe poverty line by 10 percent increases the percentage of poor individuals by 25 to 50 percent, depending on the poverty line used. This i s true for negative shifts inthe poverty line as well. Poverty in Albania has marked spatial and regional dimensions, with rural areas and the Mountain region being consistently poorer than rest o f the country, according to all definitions o f poverty. Poverty headcount in rural areas is 66 percent higher than in Tirana, and 50 percent higher than in other urban areas. Per capita consumption inrural areas, at 7,224 Leks, i s about four-fifths of the consumption levels inurban areas. Households inthe most remote districts inthe Mountain region inthe north and northeast o f the country do not fare well in terms of poverty, and almost half o f residents of this area are poor, and more than a fifth live in extreme poverty. Also, the depth o f poverty inthis area i s muchmore pronounced than in any of the other regions, with a poverty gap index of over 11 percent. Average consumption in the mountainous regions i s two-thirds of consumption levels in Tirana, and about 20-30 percent lower than xii the rest of the country. Poverty rates and depth in all other regions are around or below the national averages. Tirana Other Urban Rural Total Extreme Extreme Extreme Extreme Poor poor Poor poor Poor poor Poor poor Headcount (percent) 17.8 2.3 20.1 4.8 29.6 5.2 25.4 4.7 Meanper capita consumption (Leks) 9,043 8,468 7,212 7,801 Rural Poverty Sixty-six percent o f the poorest quintile and 61 percent of the second quintile are rural households. Inall, 34.8 percent of the rural population i s poor (poor 29.6 percent and extreme poor 5.2 percent). A significant amount of consumption (more than 30 percent) in rural areas derives from own-production, which indicates the importance o f small-scale agricultural production for food security and maintenance of minimum consumption levels. Rural areas depend largely on agriculture as their main source of income (44 percent), followed by employment (19 percent), public transfers (17.8 percent), private transfers (12.6 percent) and non-agricultural production (5.9 percent). Compared to urban areas, the share of all non-farm income sources is lower in rural areas. Urban households depend on public and private transfers for 41.5 percent of their income, while inruralareas only 30.4 percent o f income comes from these sources. Almost 90 percent o f the rural population own and use agricultural land. Rent o f agricultural land is undertaken rarely. The average land size owned by a household is around 0.72 hectares, indicating that there i s very little room for enlarging agricultural operations. From an international perspective, the data presentedconfirm the picture of Albania being one of the poorest countries in Europe. The PPP poverty rates presented(2 and 4 dollars per day) are higher than those of neighboring countries, and put Albania more or less in line with Caucasus countries like Georgia, and slightly ahead o f Azerbaijan. Despite growing concerns about widespread inequality in transition economies, inequality i s found to be moderate, with a consumption-based Gini index of 0.28. The 90/10 ratio, another measure of inequality, is moderate at 3.6. The main source of income for poor households is agricultural business (37 percent), followed by wage employment (27 percent). For non-poor households dependent work i s the single most important source (35 percent), followed by agricultural (25 percent) and transfers (14 percent). Non-agricultural business provides only 6 percent of poor families' income, slightly less than transfers (8 percent) and Ndihme Ekonomike (7 percent). Compared with poor households, non-poor households derive a significantly higher share of their income from transfers. For both groups, pensions provide about 14 percent of the total. Likewise, no difference can be noted in the share of income provided by unemployment benefits (less than 1 percent), maternity, social care and other minor public transfers and other sources (both less than 1percent). Albania's rural areas, particularly in the Mountain region in the north, are amongst the poorest. Compounding the problems for those Albanians who live in these areas is the harsh climate and rugged, mountainous geography. The majority o f residents make their living through subsistence farming, which provides a significant source of the rural families' daily food intake. However, most do not have enough land to provide adequate sustenance or variety, and most rely on a small set o f staple crops and products for their meals. ... Xlll Determinants of Poverty While household size is a strong indicator of poverty, low levels of educational attainment and unemployment are perhapsthe maindemographic factors drivingpoverty.Primaryschool enrolment rates are lower amongthe poor, and lowest amongthe extreme poor.The pattern is muchmore pronounced for enrolment rates in secondary school. The incidence of poverty among adults holding a primary school diplomaor less is double the povertyrate among individualswith a vocational diploma or higher. More than 80 percent ofthe poor havenot completed secondary school or holda vocationalschool diploma. Only 33 percent ofthe individuals15 years of age andolder holda full timejob, while another 21percent work on a part-timebasis (3 1percent in rural areas), bringingthe total ofworking peopleto 53.9 percent. About one half of rural workers only work part-time,and a large portion of them report working only a few hours a week. Irrespective of how employment is defined, the poor exhibit substantially higher unemployment rates when comparedto the non-poor.The unemployment rate among the poor based on the standard ILO definition (at 14.2 percent) is almost double the rate among the non-poor.Among the extreme poorthe unemployment rate is virtuallytriple.Usingthe relaxeddefinitions, unemploymentrates amongthe extreme poor is above 30 percent.A significantlyhigherproportionofmarginalworkers(those classifiedas discouraged, seasonaland laidoff workers) are found amongthe poor. Results of multivariate analysis confirm the negative and strong correlation between education and poverty.Both in rural and urban areas, familieswith higher shares ofmemberswith secondary and higher education are less likely to be poor. As expected, the effect of higher education is larger than that of secondary education. Poverty is more likely to be found in large households: the probability of a household being poor increases with household size. Families with a higher share of children are also more likely to be poor. Employmentstatus also matters, and when a higher share of family members is unemployedthe householdis significantlymore likely to be poor. Poor Non-poor Total Meanhouseholdsize 5.7 4 4.3 Labor market characteristics of the poor Albania enjoys a high labor force participationrate (66 percent), but the levels of skills are low, and almost two-thirds ofthe labor force havingjust basic or less than basic education. Almost two-thirdof the working population is either self-employedor work as unpaidworkers in a householdfarm or non-farm enterprise.Most self-employmentis in agriculture, and only one-thirdof the employedearn wages. About 10 percent of the employed in Albania work inthe informalsector, and this share beingabout 24 percent in urban areas. There are more men than women in the informal sector. There is evidence of long-term employment in the informal sector, with almost 75 percent of individuals being engaged in informal xiv activities for more than three years, and about 10 percent working inthe informal sector for more than 10 years. Unemployment remains widespread, and the unemployment rate i s especially high in urban areas. Long- term unemployment reigns - it takes, on average, more than one year to find ajob. Tirana has the lowest labor force participation and the highest unemployment rate compared to other geographical areas in the country. After sharp declines of the early and late-nineties, real wages in Albania are growing again. Over the 1999-2001 period, real wage increased, on average, by 10 percent annually. Still, wages in Albania are very low: in the year 2002, an average monthly salary was 18,680 Leks (about $US 133), and almost 50 percent of the wage employed had earnings of less than $US 100 per month. L o w pay is mainly associated with low educational attainment, and also with working in a low paying industries with insecurejobs, such as services, agriculture, and the social sector. Women earn significantly less than men with equal qualifications. Private sector earnings are higher that earnings in the public sector, and the wage differential increases with educational attainment. Non-Income Dimensions of Poverty Non-income dimensions o f deprivation compound income poverty, and living conditions disparities are more pronounced than are readily apparent by looking at the income dimension alone. Modernization o f the country has benefited existing urban areas, particularly Tirana, at a much faster pace than the rural areas. The coverage o f basic infrastructure services i s nearly universal in existing urban areas, but much less so in rural and recent peri-urban areas, though existing urban areas face serious quality and reliability issues. Health There are large income and regional inequalities in health status and in access to health care services. There are large differences between rural and urban areas as regards access to water and sanitation, and while all urban dwellers have access to a sanitation facility inside their house, only half o f households living in rural areas have access to sanitation facilities. There are also large income differences in the percentage o f people visiting an outpatient care provider, and while the poor and people living outside the Tirana region are more likely to report an illness than the non-poor, they are less likely to seek care when ill.Ontheotherhand,residentsoftheTiranaregionarelesslikelytoreportanillness,butmorelikelyto seek care when ill.Overall, only 58 percent o f people who report an illness or injury seek outpatient care. There are large regional differences in inpatient care utilization as well. Almost all women receive pre-natal care, and trained personnel attend almost all mothers at birth. However, there are large income and regional differences in the number o f pre-natal consultations and in the type o f provider attending the women at birth. Comparatively speaking, however, women in the Mountain areas are the least likely to receive pre-natal care, and receive the least number o f clinical consultations. In contrast, women in Tirana receive two times as many pre-natal consultations than women in other parts o f the country. There are large out-of-pocket expenditures in health care. Expenditure in medicines represents the largest share o f these expenditures. Informal payments are widespread, especially in hospital care where 60 percent o f people reported making under-the-table payments. These payments also represent an important share o ftotal expenditure, accounting for about a quarter o f total expenditure on hospitalization. Health care expenditures represent a significant share o f total household monthly expenditure, and vary with the income level o f the household. Though the poor pay less in absolute terms, they spend a larger share o f their income on health care than the rich. Poor households are not only more likely to experience xv catastrophic health expenditure, they are also more likely to have larger excesses inhealth payments over the catastrophic threshold. Education There are large income and regional inequalities in access to education facilities and education outcomes. While Albanians 21 years and older have 8.5 years of schooling on average, there are large variations across income and place of residence, and an adult living in a poor household, on average, has two years less of schooling compared to one living in a better off household. Similarly, adults living in Tirana have about 3.5 years more of formal education than those living in rural areas. There are large variations as well inthe total years o f completed schooling across income and place of residence, and even though the percentage o f those who finish basic education is high, there is a sharp decrease inthe number o f students who continue schooling beyond basic education. There are also large differences between rural and urban areas, and while these differences are small in the basic years of schooling, they increase at the upper secondary school level. There are large variations in enrolment rates at all levels of education across regions and income levels. Enrolment rates for the poor, and especially the extremely poor, are much lower than those of the non- poor; similarly, enrolment rates in rural areas are much lower than urban areas. Enrolments in pre-school have greatly declined during the transition period, from more than one in two children between 3 and 5 years o f age in 1990 to less than one inthree in 2002. The main reason for the low enrolment at the pre- school level appearsto be the non-availability o f pre-schools, the number o f which has fallen by one-third inthe last decade. Enrolment rates inupper secondary schools are low, and have dropped duringthe transition years to about half of the 1990 levels. Enrolment rates in secondary school also vary largely across income and region of residence, and the net enrolment in secondary education of poor children i s less than half that o f the non- poor. Across regions, while almost 70 percent of the children of the relevant age group are enrolled in Tirana, and close to 60 percent inother urban areas, only a quarter of the children insecondary school age in the rural areas are enrolled. Reasons for not enrolling in secondary school tend to be economic, and many students leave to work after completing basic education. Public spending on education has fallen to levels almost half of what it was in 1990. The reduction in public spending on education has been balanced by increase in private expenditures, which on average represents about 1.4 percent of the total household monthly per capita expenditure (almost 2.5 percent for households with students). Households in the upper quintiles of the income distribution pay on average about 2.5 times more than those in the poorest quintile. However, as a percentage of total monthly expenditure, poorer households pay a similar or even a higher percentage (the difference, however, is not statistically significant) o f their total household expenditure. Informal payments are widespread, and in both basic and secondary schools, more than 40 percent of the students report making such payments to schools and teachers. Coping with Poverty-The Role of Social transfers Social protection in Albania today consists of three major groups of public programs, plus private transfers between and within families. The public programs consist of a comprehensive social insurance system, social safety net programs, and labor market policies. These are discussed in detail below. The Social Insurance System provides for comprehensive protection against income loss due to old age, disability, loss of a primary earner, death, unemployment, general sickness, occupational injury and work related illnesses, childbirth and maternity. The Social Safety Net consists of three cash social assistance programs: a targeted poverty benefit in cash (called Ndihme Ekonomike or economic aid, which provides a means-tested cash benefit for eligible families with little or no earned income); a regular monthly xvi allowance to those disabled since childhood; and price compensation paid to pensioners and their families, and a program o f social care services for orphaned, disabled, and elderly people. In a broad sense, the current safety net has replaced an inherited social protection strategy in which people were assured o f guaranteed employment at low wages on collective farms and in govemment enterprises, while a system o f administered pricing within a closed economy kept prices for consumer staples correspondingly low. In 2002, averaged across all households, public transfers represented 21.3 percent o f household income. Private transfers represented a further 14.4 percent. Together, therefore, these reported transfers provided more that one Lek out o f every three. That made them a more important source o f household income, on average, than either earnings from employment (3 1.5 percent) or agriculture (25.0 percent). For the population as a whole, transfer income sources that were quantitatively important as a share o f household income @e., those contributing more than 2 percent) consisted o f all categories o f pensions [urban old- age pensions (10.5 percent), rural old-age pensions (3.8 percent), and other pensions (2.4 percent)], Ndihme Ekonomike (3.1 percent), and remittances from relatives abroad (12.8 percent). In short, public pensions and private transfers from Albanian relatives abroad dominated, with Ndihme Ekonomike contributing substantially less. Among all Albanian households, the overall poverty rate o f 25.4 percent in 2002 would have been 11.8 percentage points higher - that is, 37.2 percent - if there had been no public transfer programs. Relative to this higher, implied level o f poverty before public transfers, the public safety net reduced the number o f poor households by almost one-third. The major contributions came from urban, rural, and other pensions and Ndihme Ekonomike (in that order). The other public transfers programs made modest additional contributions. Similarly, the extreme poverty rate would have been 11.2 percent, 6.46 percentage points higher than its actual rate o f 4.74 percent higher. Thus, relative to this higher, implied level o f extreme poverty before public transfers, the public safety net reduced the number o f extremely poor households by 58 percent. The individual programs had the same relative importance in bringingabout this reduction as they did for the overall poverty rate. These estimates assume that all transfers contributed 100 percent to additional consumption. Remittances from abroad constitute a distinctive feature o f transfer patterns in Albania. On average, one Lek o f household income out o f every eight came from this source. However, among the 22 percent of households receiving such transfers, their average size was 13,600 Leks per month. That is equivalent to 47 percent o f average monthly household income for the country. Moreover, private transfers from abroad were much more likely to be received by non-poor households (23.5 percent) than by poor households (16.8 percent) or extremely poor households (10.1 percent). Similarly, when these transfers were received, they were substantially larger on average for non-poor households (14,400 Leks per month) than for poor households (8,900 Leks per month) or extremely poor households (6,100 Leks per month). Exploration i s warranted o f the extent to which receipt o f private transfers from abroad was itself sufficient to move families out o f poverty and/or into the higher income quintiles. Of households receiving Ndihme Ekonomike, 76 percent were poor and only 24 percent were non-poor. Moreover, 43 percent o f all households receiving Ndihme Ekonomike payments were in the lowest quintile, which contains four out o f every five poor households. Another 24 percent were in the second quintile, which contains the remaining poor households. Only 12 percent o f Ndihme Ekonomike recipients were in the two top quintiles. However, poverty did not assure receipt o f Ndihme Ekonomike. Only one-quarter o f all poor families receivedNdihme Ekonomike; and those families representedthree- quarters o f all families receiving the benefit. Fully 69 percent o f households in severe poverty and 75 percent o f all poor households did not receive Ndihme Ekonomike. (Some o f these, o f course, might not have met all the eligibility conditions, like ownership o f assets and family composition). xvii Pensions and remittances from Albanian relatives are the largest sources of transfer income. Urban old- age pensions are disproportionately received by the two upper quintiles; but the 25 percent of households that are poor represent 37 percent of the recipients of urban pensions. Rural old-age pensions are disproportionately received by the two lower quintiles and the poor. It i s notable that poor households receive a more than proportional share o f each differentform of transfer income examined. Copingwith Poverty -Internal and External Migration Migration i s perhaps the single most important political, social, and economic phenomena in post- communist Albania, and has been a dominating fact of everyday life in the last decade. Since 1990, approximately one-fifth of the total population o f the country has left and is living abroad, and there have been large-scale movements of population from rural to urban areas. Migration, whether rural to urban or international to Italy or Greece, is probably the most important livelihood coping strategy employed by rural households, and serves as an important escape valve for unemployment and other economic difficulties. Official estimates make remittances the largest source of foreign exchange, greater than the combined value of exports and foreign direct investment, currently constituting 14percent o f GDP. Inthe past decade Albania has experienced a demographic transformationfrom rural to urban, and from the North Eastern Mountains to the districts of the Coast and Tirana. Tirana is by far the principal destination o f internal migrants, a process that appears to have accelerated and become increasingly concentrated in the second half of the 1990s. Migrationto Tirana is strongest towards poorest peri-urban areas, re-enforcing a vicious cycle of poverty and adding strain to already overstretched municipal services. Further, there is evidence of a local rural-urban migration within the North Eastern region. Whether these urban centers in the north are only interludes to more distant and lucrative types of migration for these households remains unclear, although the large numbers of migrants from these provincial urban centers in the North to Tirana seem to suggest that many, indeed, eventually move on. Given the internal migration to the urban areas o f the Coast, as well as local rural-urban migration, creating economic opportunities in urban areas beyond Tirana could help in re-establishing a more sustainable rural-urban flow, as Tirana alone i s struggling to accommodate and service the increasing number ofmigrants flocking to its poorest neighbourhoods. The vast majority o f temporary migrants, mainly from rural areas in the Center and the North-East of the country, travel to Greece to seek short-term employment opportunities to complement the meager earnings from agricultural activities. Despite the higher transaction costs, in recent years the flow of temporary migrants to more distant destinations such as Italy and Germany has increased substantially. Evidence indicates that among the non-migrants, a large share has considered the option o f migrating, and many have tried and failed. About 458,000 Albanians currently live abroad. Information from Greece and Italy, the two principal receiving countries, puts the number of legal residents in these countries as approximately 600,000 in 2000-2001, or one-sixth of the Albanian population. An astonishing 35 percent of children of household members currently live abroad. For children who left home in the 1990s, the share reaches exodus proportions, with one child in two currently living abroad. The poverty headcount for households with children abroad is significantly lower (12- 18 percent, depending on the destination), as compared to households with children in Albania (3 1 percent). The amount of remittances from children living abroad i s estimated to be over US$200 million per year. Almost one-half of Albanian households have access to migration networks, either through direct migration o f a household member or through their children living abroad. This is most likely an underestimate, and it is comparable to a country like Mexico, which has over 100years of migrationtradition and experience, compared withjust over a decade for Albania. Migration is a double-edged sword, however, and while in the past migration out o f rural areas was perceived as a potential solution to the problem of rural poverty in Albania, it is also increasingly seen as xviii contributing to social dislocation, agricultural labor shortages and rapid deterioration in the provision o f social services in urban areas. Furthermore, the large-scale legal and illegal migration o f Albanians to other parts o f Europe constitutes an important political and social challenge to the governments and citizens o f the European Union. Stricter migration laws, increasing costs and associated risks also make migration less o f a viable solution, particularly for the less-endowed households. And finally, although migration and the resulting remittances are likely to continue to be an indispensable ingredient in the Albanian recipe for rural development, the need and necessity to devise more appropriate, sustainable strategies to lift rural households out o f poverty and promote the country's growth cannot be over- emphasized. Peri-urban areas The peri-urban areas lie on the outskirts o f urban centers, and are often situated on or near sources o f fresh water (rivers and streams, wells). They are usually composed largely o f informal settlements and structures, and often lack connections to basic utilities (water, electricity, waste disposal). Most residents o f the peri-urban areas are recent arrivals from other areas in Albania who have migrated to the urban centers with the hope o f employment and a better life for themselves and their families. As such, the population o f the peri-urban areas i s both fluid and difficult to gauge. Often times these residents migrated from other areas o f Albania en masse and ended up in communities that mirror the communities intheir previouslocation, and are comprised primarily o f extended families. A huge problem in peri-urban areas is the lack o f basic services. Chief among these is the government's inability to provided clean runningwater to many households. Many families have to get their water from a nearby well, or from a source o f running water not located near their residence. Because o f a corresponding lack o f waste removal systems in these peri-urban areas, many families obtain their water from sources near where they dispose o f their waste (both human waste and trash). Many o f the same problems hold for the delivery o f basic electrical services as well. Many residences are connected to electric wires, but most indicate that service is intermittent (at best), and that the power provided is insufficient to runbasic household appliances. Families moving to urban areas have only served to overcrowd the already insufficient school systems. Class size has increased dramatically in the peri-urban areas, and many schools have started runningon multiple rotations (one set o f classes half o f the day, a completely different set the other half) in an effort to relieve stress on teachers and classrooms. This has resulted inmany students coming to school for only a few hours each day. Access to health services in peri-urban areas is very limited, and in the peri-urban areas o f Tirana there is still no network o f health centers that would be in line with the number o f inhabitants. Transport and Infrastructure The need for improving transport and road infrastructure cannot be overemphasized. As noted earlier, there are significant differences in poverty levels across regions and across rural-urban dimensions, with rural and the Mountain regions being consistently poorer than the rest o f the country according to all definitions o f poverty. In this situation, improving access to farm inputs, processing o f farm outputs and reduced isolation across many dimensions would necessarily involve significant transport improvements. Further, considering that the growth o f the manufacturing sector i s essential for overall economic development, and investments in all candidate manufacturing sub-sectors i s predicated on reliable transport links for flows o f raw materials and finished goods, across the borders and within the country, it i s imperative that transport and road infrastructure be strengthened. Likewise, the benefit o f proximity to existing higher-order health and education facilities within present urban areas can be extended into peri- urban areas if passenger transport services can operate into the peri-urban by means o f basic, all-weather, road infrastructure. In light o f the large volume o f migration from rural areas, improving roads could potentially be preferable to constructing new public service facilities in peri-urban and rural areas. xix Poverty-Reducing Strategies An areathat warrants priority is employment generation, sincethere exists a strong linkagebetweenbeing unemployedand beingpoor. The predominant labor market and poverty issue in Albania is how to deal with the big excess supply of labor, giventhe limitedmarket mechanismsavailable. The government is no longer able to keep labor demand artificially inflated, and the newly created market mechanisms are too weak to offset employment losses. In addition, the wages are low and there is not much room for price- adjustments. The excess labor supply has brought high rates of unemployment, hidden unemployment, and a largely subsistence-based informal sector. As a result, large segments of the labor force, such as younger people, new graduates, women, and people of pre-pension age do not participate in the labor market. The recently adopted NSSED sees the labor market development and the promotion of employment as one of the main components of the poverty reduction policies. The document foresees preparation of relevant labor market legislature to insure provision of the professionaltraining and the development ofthe workforce,the formalizationofthe labormarket, andthe promotionofemployment. A prerequisite for the success of any labor market program is a favorable business environment and reasonably flexible labor market. Development of the private sector is critical for sustaining economic growth in Albania and providingemployment.Creatinga business environment favorable to the growth of the private sector should therefore be a priority for the Government, which needs to take measures to transforminformalbusinessesto formalbusinesses, increasethe levelof competitivenessofthe economy, strengthenthe legaland institutionalframework, andencouragethe inflow ofcapitalto the economy. A recently concluded study of the investment climate in Albania highlights bureaucratic bottlenecks, weak law enforcement and strong corruption as major impedimentsto the growth of private investment, and emphasizesthe needfor removalof administrativebarriersas the key to encouragingnew investment. The study finds that investment procedures in Albania are seriously flawed and characterized by over- complexity, lack of clarity and transparency, and investor unfriendliness. Private business faces bureaucratic delays, discretionaryand arbitrary treatment, and high costs, includingfrequent "unofficial payments", all ofwhich increasethe costs of investment. Inpracticalterms, the Governmentneedsto take measuresto removeexistingconstraintsto firm creation and growth.Most common obstaclesto privatebusiness development inAlbania are corruption, poor law enforcement, insecure property rights, bureaucratic harassment of small firms (frequent inspections, audits, etc.) andother lengthyadministrativeandregulatoryprocedures. Like other transition economies, Albania inherited labor market institutions and regulations developed under central planning,which have provento be ill-suitedto the needs of a market economy. This legacy of rigid regulationshas posed a considerable hindrance to labor market performance in the country. The labor market reforms in Albania have to (a) improveemployment flexibility through lessening dismissal costs (e.g., lowering severance pay, introducingtemporary employment contracts); (b) enhance working time flexibility (e.g. through providing for reschedulingof working hours); (c) increase wage flexibility (e.g.through eliminatingrigidwage grids); and (d) reduce some of the non-cashbenefits.Simultaneously with the deregulation of labor relations the reforms have to support their devolution via increased participationoftrade unions and employers' organizations in collectivebargaining. The results of the analysis point to three priority areas for policy makers interested in improvingaccess and effectiveness of health care. First, coverage under health insurance needs to be increased, both in terms of number of persons covered as well as number of services providedunder the coverage. Health insurance is not only important as a source of fundingfor the sector; it also providesfinancialprotection for the insured against catastrophic health expenditure. Inparticular, the poor and vulnerable sections of the population,who are at risk of falling into povertyas a resultof a single incidence of illness,needto be covered by health insurance. Indeed, even though the poor and vulnerable are the ones most likely to be xx covered by health insurance at present, there remain large groups o f the poor and vulnerable that are actually not covered. Further, health insurance only covers the services o f the GP or the family doctor, and provides no protection against catastrophic health expenditure. InTirana, health insurance covers all outpatient care and not surprisingly, people in this region spend the least on outpatient care, on average, compared to people in the rest o f the country. There is clearly a strong case for increase inthe number o f services offered. Another area that needs immediate attention is informal payments, the eradication o f which will have positive effects on access to health services. Informal payments in outpatient care represent about 11 percent o f the total expenditure, in contrast to hospital care, where they represent close to 25 percent o f the total expenditure. On average, Albanians spend almost 700 Leks per capita per month on health care, which is about 18 percent o f the average per capita consumption o f a poor family. However, there are large differences in the amount households spend across consumption quintiles.Households belonging to the poorest 20 percent o f the consumption distribution on average spend about half what households in the richest quintile spend on health. As a percentage o f total consumption, households in the poorest quintile o f the income distribution spend almost 10 percent o f their per capita monthly consumption on health. Besides increasing the burden o f out-of-pocket expenditure on health, particularly for the poor, informalpayments also undermine all intent to improve quality and efficiency in the sector. Anecdotal evidence also shows that quality o f care has deteriorated, particularly in outpatient care and particularly outside Tirana, where a large number o f people choose to bypass these services and go directly to hospitals. Improvingthe quality o f the outpatient care services i s necessary for enhancing the scope and utilization o f health services, so that they can be used for more than to just obtain prescription for drugs. As far as education is concerned, secondary school enrollments need to be increased, and quality o f education at this level improved. Secondary schools should impart market-relevant skills and training, emphasizing flexibility, adaptability and analytical skills over rigidly drawn curriculum. New vocational programs, in line with potential demand for skilled labor, needto be introduced to prepare students for the job market. New secondary school programs are especially needed in rural areas, where the lack o f these programs i s one o f the main causes o f low enrollment. Another area that needs immediate attention i s improvement in the provision o f basic services and infrastructure, including access to quality health and education services and to basic sanitation services. It i s likely that for many households social exclusion, the lack o f access to basic services and infrastructure, and to secondary and higher education, are factors hindering their ability to lift themselves above the poverty line. The Government has already pledged renewed effort towards making marked improvements to its public service infrastructure, particularly in relation to delivery to the poorest, and it i s imperative that these focus on rural and remote regions where the dismal condition o f basic services i s most perceptible. Since a large number of the poor live in rural areas, it i s imperative that the Government develops a rural development strategy focused on poverty reduction in the short run. A large share o f the population is engaged in agriculture, but the farm size is, in many cases, too small to be a viable commercial option or simply one that allows a family to escape poverty. Since the agricultural sector provides the most substantial share o f income for the rural poor, addressing the factors hindering the development of this sector has to be an essential component o f any policy mix aimed at reducing rural poverty. Agriculture i s the key income generating activity in rural areas, and even richer households rely heavily on agriculture as their main source o f income. Sales o f livestock are more important for poorer households and households that live in mountain areas. Measures that provide access to better small scale xxi processing could potentially benefit the poor and the more remote farmers and increase their income generation from livestock. Flexible and locally adapted rural extension service can do a lot to improve current production patterns. Poorer farmers have higher production costs and less income. They are not so diversified and tend to be less engaged in processing. These are all issues that can be addressed by extension services that take the local conditions into account and advise farmers based on market possibilities. Increased quality and quantities will potentially be used by small to medium enterprises that start processing activities in rural areas and can thus be a potential outlet for poor farmers. In terms o f future growth, attention must necessarily shift to the manufacturing sector, which currently contributes only 11.6 percent o f the country's GDP, and employs only 4 percent o f the workforce in the country, making it the fourth-largest employer, after agriculture, wholesale retail and repairs, and construction. The agriculture sector employs more than half o f the labor force, but most are unskilled and low-skilled, and over 15 percent are part-time. The extent o f disguised employment in agriculture is not well understood, but there i s always the distinct possibility that land consolidation and technology could well push a large number o f the relatively unskilled out o f the agricultural sector. As it is, migration is effectively sending a large number o f persons from rural areas to urban and peri-urban centers o f the country. Employment in construction is already quite high. At the same time, there is a limit to the extent that construction activities can be expected to grow, and it does not seem likely that the construction sector will be a catalyst for the creation of new revenue generating activities. Stimulating the manufacturing sector could potentially create several employment opportunities. Within the manufacturing sector, there exist several industries that could potentially offer different employment opportunities, and it could be possible to further target the industries with the highest growth potential. Consider, for instance, textile manufacturing, which i s the largest employer within the manufacturing sector, which also has the lowest wages. Another sector i s leather goods which, given Albania's proximity to Italy and wide familiarity with the language, combined with recent trends in dispersed manufacturing, could well afford Albania the opportunity to ease into production. This opens up intriguingoptions, and i s an area that needs to be explored indepth. Another area that deserves greater attention is tourism. Albania is a country o f great natural beauty, enhanced by the mountains in the north and the sea in west. At the same time, Albania is close to Greece and Italy, countries that attract lots o f tourists from within Europe and outside. Development of tourism within Albania presents itself as an attractive option, one that needs to be explored and exploited more aggressively. There is little doubt that the many government programs and services identified above could potentially lead to increased benefits for the people, the realities o f resource-constraints have to be recognized. However, given the overall scarcity o f funding even for existing programs, it is not obviously clear how new and more programs could be financed. This makes the task o f poverty reduction even more challenging and underscores the need for better understanding the trade-offs across different items o f government expenditure. xxii CHAPTER I:INTRODUCTION 1. Background:Why aPovertyAssessmentfor Albania? Albania has undergone a process of major economic and social changes in the last decade following the collapse of the communist regime. On the economic front, following a sharp contraction in the early transition years (1991-92), GDP grew by about 9 percent annually in real terms between 1993 and 1996, during which inflation was brought down substantially, and both the current account balance and the fiscal deficit improved sign,ificantly. With the collapse of the large pyramid schemes in 1997, however, economic growth also fell into the negative territory, registeringa negative growth o f 7 percent in real terms in 1997. A recovery program initiated by the government inthe aftermatho f the crisis yielded quick dividends, and the economy bounced back to register an average growth rate o f 7 percent over the next three years. In 1999, the influx of Kosovo refugees put a heavy strain on Albania's economic and administrative capacity. However, the economy was able to weather the storm with strong support from the donor community. By the end of the 1 9 9 0 ~Albania had roughly regained its 1990 GDP level, ~ inflation had been brought under control and the current account and the domestically financed fiscal deficit had improved significantly.' The strong macroeconomic performance continued in 2000 and GDP registered a 7.3 percent growth in that year. GDP growth declined somewhat during the second half o f 2001, due mainly to power outages and the slowdown inthe world economy after September 11. The slowdown continued in2002, and GDP growth dropped to 4.7 percent due to reduction in public investment, continuation of the energy crisis, political uncertainty and a slow down in reforms, particularly the privatization o f enterprisese2Cost-push factors put upward pressure on prices since mid 2001, and average inflation, which was 3.1 percent in 2001, rose to 5.4 percent in 2002 due mainly to higherfuel and some food prices. Inflationdeclined to 3.0 percent in 2003, and is expected to remain at that level for the year as a whole. Growth is also expected to rebound to around 6 percent in2003. Several structural reforms have also been carried out since the late 199Os, and have involved banking, land reforms and privatization. Almost all small and medium enterprises have been privatized and significant progress has been made in the privatization of strategic sectors, like telecommunications. Regulation in the financial sector as well as core public sector functions has been strengthened, and the Government has adopted a new bankruptcy law and a deposit insurance scheme. The registry on movable property has been established and progress has been recorded in the development o f a land market. Significant steps have also been taken in the area of public administration: a comprehensive civil service law has been adopted; a wage structure has been developed; a medium term expenditure framework has been put in place; legal and judicial capacities have been strengthened; accountability rules have been tightened; and the anti-corruption action plan has been revisedand is under implementation. Despite the impressive performance of the economy, however, poverty in Albania has remained high and per capita income, at around US$1,230 in 2002, has remained one o f the lowest among transition economies. Partly a reflection o f the difficult conditions in Albania at the outset o f the transition and 1 For a detailed discussion of the Albanian economy during the early years of transition see "Albania Beyond the Crisis," Country Economic Memorandum (CEM), December 1998. The "Public Expenditure and Institutional Review", April 2001, elaboratesonthe more recent fiscal developments inthe economy. Albania's GDP figures should be treated with caution; they are not based on any national accounts, and there are no reliable private sector activity and investment data. The Government has recently changed its methodology for calculating GDP, and while agriculture production is directly estimated, the growth in other sectors is measured on the basis of a new sales index, supplementedby several sector specific indicators(e.g., for electricity). 1 partly o f the early economic recession of 1991 and the civil unrest o f 1997, per capita income did not register a significant rise even though the economy recorded high growth rates of 9 percent annually during 1993-96 and 7 percent during 1998-2001. Inan effort to adopt policies and measuresto share widely the benefits of growth and reduce poverty, the Government of Albania outlined a poverty alleviation strategy inthe InterimPoverty Reduction Strategy Paper (2000) and developed it in detail two years later in the Growth and Poverty Reduction Strategy (GPRS), the precursor to the current National Strategy for Socio-Economic Development (NSSED). The NSSED is a comprehensive program that aims to address poverty through the systemic reforms that involve creation o f a growth-oriented business environment that attracts investment and creates employment opportunities for the poor, improvement of human capital through improved quality and access to social services such as health and education, and empowerment of the poor through stronger public accountability and increased participation in decision making. The strategy highlights the importance o f attaining macroeconomic stability, and obtaining an adequate social protection network as pre-conditions to growth and poverty reduction. Besides giving education, health, social support, governance, and infrastructure priority, the specific objectives set by the NSSED are to: (i)to increase real GDP by 25 percent; (ii)to reduce the number of people living in poverty; (iii)to improve water, sanitation and energy infrastructure; (iv) to decrease maternal and infant mortality rates and disease incidence; and (v) to increase primary and secondary enrollment rates as well as average years of schooling. Finally, and given the large scope of the government's reform agenda, the NSSED also highlights the importance of monitoring and evaluating its implementation. Inthe early stages of development, however, NSSED-or GPRS, as it was knownthen-was constrained by the availability of data on poverty, and while it did make good use of the existing quantitative and qualitative information, an accurate and complete account of poverty was not available. The problems were further confoundedby the roughly 10 percent decrease in population (since last census) that Albania has experienced over the last decade, accompanied by significant shifts of the population from rural to peri-urban areas, all of which led to huge economic, demographic, and social changes. Quantification and analysis of poverty in Albania were recognized as urgent and necessary in order to both document comprehensively the poverty situation and to construct a meaningful responseto the problem. In recognizing the data limitations, the GPRS helped the Government to, inter alia, complete the first round of a five-year Living Standards Measurement Survey (LSMS) program in 2002 and implement a census in 2001, which provides age-specific and regional population breakdowns. The survey provides valuable information on a variety o f issues related to living conditions o f the people o f Albania, including details on income and non-income dimensions of poverty in the country, and forms the core basis of the present assessment o f poverty in Albania. As a first priority, detailed poverty numbers were computed and a poverty profile constructed, which have since been used extensively to provide accurate and comprehensive informationfor the further development ofthe NSSED. The present assessment o f poverty takes the analysis a step further, and gets into more details on a number of key issues. The IDA has been assisting the implementation of the NSSED through programmatic operations. The Poverty Reduction Support Credit-I (PRSC-I) was the first o f these operations aimed at supporting the first phase of the strategy. This was followed by Poverty Reduction Support Credit41 (PRSC-11), the second programmatic operation, which continues to support the Albanian Government in its reform agenda. A third programmatic operation, PRSC-111, i s currently underpreparation. The NSSED and the PRSC-I have recorded impressive success on a number o f fronts. First, Albania has maintained a satisfactory macro framework; since 1997 the economy has been growing and inflation has beenkept under control. Second, a pension reform program has been put in place; this reform is projected to reduce the pension-related deficit to 1.2 percent of GDP annually. Third, the strategy has strengthened 2 the government's monitoring and evaluation capacity; a population census and the first round o f the Living Standard Measurement Survey have been completed and are being analyzed. And fourth, accountability and transparency o f public sector operations is being strengthened. The second stage in the Albanian Poverty Reduction Strategy hasjust started. The NSSED will continue its strong emphasis on social service delivery and policy monitoring and evaluation. This new stage will also deepen its focus on rural and private sector development. Accordingly, the PRSC-I1 is structured around the following four components: (i)promoting sustainable growth and supporting private sector development; (ii) strengthening the government's capacity to monitor and evaluate the policy agenda; (iii) improving service delivery and social safety net effectiveness; and (iv) improving core public sector functions and institutional arrangements. The rest o f this chapter is organized as follows. Section 2 presents the macroeconomic framework and gives a general idea o f economic performance and macroeconomic forecast. Section 3 discusses the measurement o f poverty in Albania, and gives a brief account o f previous efforts. Section 4 provides a road map for the rest o f the chapters o fthis poverty assessment. 2. The Economic Context Economic performance has been strong in the last five years, particularly following the 1997 crisis, and growth has averaged about 7 percent annually. As Table 1.1 shows, inflation has been contained at low levels. Both exports and imports have grown steadily, resulting in an increasing current account balance. Foreign exchange reserves have been steady, and external debt has remained at around US$O.S billion through the period. Current-accountbalance(US$ m) -65.1 -155.4 -156.3 -218.5 -370 Foreign-exchangereserves (excl gold; US$ m) 348.5 369.1 352.2 362.5 Private Transfers (Remittances; US$m) 440 327 439 543 Total external debt (US$ billion) 0.6 0.7 0.8 0.8 Debt-service ratio, paid (percent) 2.4 2.4 2 2.4 IExchange rate (Lkto US$; average) I 150.63 I 137.69 I 143.71 I 143.49 I140.74I Source: Bank of Albania, Monthly Statistical Report; IMF, International Financial Statistics. Private remittances are and will continue to play an important role in supporting growth. Inthe year 2000 remittances represented close to a fifth o f the total GDP, more than twice the revenues coming from 3 export^.^ This large flow o f remittances has artly offset the deficit in the current account, which in the year 2002 equaled about 8.7 percent of GDP. L r Agriculture accounts for the largest proportion of the GDP, and as Table 1.2 shows, its salience has declined only marginally from 35.9 percent of the GDP in 2000 to 34.2 percent in 2001. Industry accounts for about 13 percent of the GDP, followed by construction and transport, both around 10 percent o f the GDP. These two sectors, however, have recorded high growth rates of over 14 percent in recent years, more than double o f the 6.5 percent growth rate in industry and ten times more than the very slow movements inagriculture. RealDercent change. Year onvear Percent oftotal GDP I I 2000 I 2001 I 2000 I 2001 I Industry 5 6.5 13.2 13.2 Agriculture 4 1.4 35.9 34.2 Construction 17 14 9.7 10.3 Transport 14 13 9.5 10.1 Other services 9 8 31.8 32.2 GDP 7.8 6.5 100 100 As Table 1.3 shows, textiles are Albania's principal exports and account for over 65 percent of all exports. Food and beverages and building materials (both around 9 percent) comprise the next largest export category. Machinery accounts for almost a quarter of all imports, followed closely by food and beverages (19.6 percent), textiles (14 percent), minerals and fuels (13.7 percent) and building materials (12.4 percent). Italy absorbs over 70 percent o f all exports, with Greece (12.1 percent of all exports) and Germany (6.8 percent) following as the other destinations. Italy also accounts for most o f the imports (36.2 percent), followed by Greece (27.6 percent). Principal exports 2001 Percent Principalimports2001 Percent Textiles 65.6 Machinery 23.6 Food, beverages& tobacco 9.0 Food, beverages& tobacco 19.6 Buildingmaterials 8.6 Textiles 14.0 Leather goods 3.5 Minerals, fuels & electricity 13.7 Wood 3.4 Building materials 12.4 Main destinations of exports 2001 Percent Mainorigins o f imports 2000 Percent Italy 71.2 Italy 36.2 Greece 12.1 Greece 27.6 Germany 6.8 Germany 5.5 Yugoslavia 2.8 Turkey 5.4 Denmark 1.1 Bulgaria 2.4 UNDP (2000). The World Bank (2002b). 4 The private rural sector is the largest employer, accounting for over 760,000 people, or over 60 percent of the labor force. The state sector is the second largest employer, accounting for around 190,000 people, or about 15 percent o f the labor force. The number of people unemployed has fallen in the last year, down from 192,000 in the second quarter of 2001 to 176,000 in the second quarter of 2002, for an unemployment rate of 14.2 percent in 2002 (Table 1.4). Fiscal deficit and money supply have remained well within the monetary and budgetary targets agreed with the IMF in the recent years. Despite slippages in revenue collection and shortfalls in privatization receipts, overall fiscal deficit has been on a declining trend from 9 percent of GDP in 2000 to 6.6 percent in2002. As a result, domestic borrowing has been contained to remain at around 3 percent of GDP. The overall fiscal deficit i s expected to further drop to 6.2 percent o f GDP in 2003. As regards the external sector, stronger than expected imports (of fuel and electricity generators) led to a worsening of the trade balance during the last two years. However, the deterioration in the current account balance was partly cushioned by exceptionally high private remittances stimulated by the inflow of euro area cash held by Albanians working abroad. With remittances projected at a normal level, the current account deficit is estimated around 8 percent in2003. One of the key aspects o f economic policy in Albania has been the strict maintenance of fiscal discipline since the second half o f the 1 9 9 0 ~As. ~ a result, the fiscal deficit has been reduced and there has been a significant decline in the level of domestic financing o f the budget deficit, resulting from a moderate increase in revenues and strict limits on expenditures. Despite these improvements, a number of fiscal challenges remain. Direct taxes constitute a small share of total taxes, and the collection of VAT has fallen short of its potential. Total revenues as a share o f GDP remain significantly lower in Albania than in most other transition economies, despite comparable tax rates. This is a reflection of poor tax administration and significant tax evasion. Total expenditures have been on the rise, but the quality of expenditures, particularly those in support of poverty reduction objectives, needs significant improvement. Interest payments on domestic debt, although declining steadily, still account for a large share (14.5 percent) of current expenditures.6 Albania needs to maintain fiscal sustainability to ensure continuous growth and steady decline in poverty. Continued adherence to fiscal discipline to further reduce the budget deficit is necessary in the coming For detailed analysis of fiscal developments during the 1990s see "Albania-Public Expenditure and Institutional Review" World Bank, April 2001. In 2002 the share of interest expenditures in current expenditures dropped to 15.3 percent, after averaging more than 25 percent during 1997-2000. This reduction was largely due to significantly lower interest rates on treasury bills. 5 years. The challenge is to sustain a fiscal deficit that i s compatible with low levels o f inflation and public debt while allowing the necessary financing for key investment expenditures, social services, and targeted support for the poor and the vulnerable. To ensure fiscal and debt sustainability, there i s a need to further reduce the budget reliance on domestic borrowing in order to lower the high domestic debt burden. Reduced interest expenditures will provide additional resources for priority areas without increasing the overall expenditures. This will require strong revenue efforts as well as inflows o f foreign assistance to support Albania's public investment needs. Fiscal sustainability also requires development and implementation o f a comprehensive pension reform program so as to reduce its reliance on budgetary transfers. A second-round o f pension reform started in2002 with the objective to address the overall long- term problems o f the coverage, affordability, equity, and adequacy. These would require appropriate measures to improve the efficiency o f operations o f the institutions in charge o f the public pensions, introduce changes in the structure o f rural pensions, and create incentives for private organizations that offer voluntary and supplementary pensions. Strict fiscal discipline needs to be combined with better allocation and more efficiency o f public expenditures. Within the overall resource envelope, the allocation o f expenditures needs to be improved in line with strategic objectives identified in the NSSED. The Medium-Term Expenditure Framework (MTEF), adopted in 2000, which aims to improve the strategic resource allocation and the efficiency in the use o f public expenditures, needs to be fully integrated with the NSSED process and other initiatives, such as the Stabilization and Association process and fiscal decentralization. Efforts should be directed at improving service delivery and at identifying and orienting public spending toward NSSED priorities. While education and health care expenditures need to be given priority, social protection transfers need to be rationalized to ensure the delivery o f effective social assistance through cash and community-based service programs. Concurrent with these measures, capacities for policy development and monitoring need to be strengthened to improve policy evaluation, while closer linkages between policy objectives and expenditures need to be developed and strengthened. Accountability mechanisms need to be further strengthened to allow Albania to fight corruption, encourage high levels o f investment, particularly foreign direct investment, maintain fiscal discipline, and ensure that resources are used efficiently and transparently. Attainment o f these objectives would be crucial not only to sustain growth and reduce poverty but also to facilitate integration into the European economy. Albania's prospects for growth, development and poverty reduction are good, especially in view o f a more favorable and stable political climate in Southeastern Europe and enhanced prospects for a closer association with the European Union. In the medium term, economic growth i s projected to remain strong, with GDP growing by about 5 percent a year. Increases in exports and private sector activities are critical to achieving this growth. Prospects for growth and reducing poverty also critically hinge on addressing the energy sector problems, implementing structural reforms, particularly those directed at improving competitiveness environment, reducing obstacles for businesses, strengthening the financial sector, and fostering private sector activities. 3. Measuringpoverty in Albania Previous data gathering for the measurement o f poverty and inequality in post-communist Albania has been sporadic, inconsistent and hardly comparable over time. The paucity o f relevant statistical data at the household level was a major constrainingfactor for the Government during preparation o f the NSSED. 6 One of the early attempts to measure poverty in Albania was a study done by the World Bank in 1996 using the data of a Household Budget Survey carried out in Tirana in 1993 and 1994 with financial assistance from UNDP and technical assistance from the National Institute of Statistics and Economic Studies of France (INSEE). The Hl3S introduced a new methodology to Albania, as the survey was a combination of household budget and multi-topic surveys. Information from a total of 3,179 households were collectedusing direct interviews as well as a 14-dayrecordkeepingof all frequent expenditures.It is important to emphasize that the survey only covered the urban area of Tirana, and it was not intendedto berepresentativeofthe country. The 1993-94 Tirana HBS was followed by the 1996 LSMS, conducted in the whole country excluding Tirana, on a sample of about 1,500 households.Resultsfrom the survey suggest that poverty in 1996was a ruralphenomenonwith almost 90 percent of the poor residing in ruralareas and 60 percent of the heads of poor households being self-employed in agriculture. The incidence of rural poverty was found to be five times higher than urban poverty and was highest in the North of Albania, where a large portion of families owned on average less than 0.5 hectare of land. Many households in these regions require subsidized wheat and cash transfers to survive the winter. Limited access to public services, often aggravatedby poor rural infrastructure,especially roads, was a contributing factor to ruralpoverty. Prior to 2002, the most recent available information on poverty and living standards comes from the Living Conditions Survey (LCS) conducted by INSTAT in 1998 as part of the PHARE program. The LCS was the country's first attempt to measure poverty based on a nationally representative sample of households from a sampling frame constructed by INSTAT. The LCS survey was designed to be multi- purpose: its objectives were to provide a global picture of the structure of living conditions inthe country and a dataset for multi-purpose analysis inthe area of social development, national accounts, poverty and prices. The survey was designedadoptingthe methodologyof the EuropeanCommunity HouseholdPanel (ECHP). It was conducted in the fall of 1998 on a representative sample of 11,523 households both in urbanand rural areas, selected from the INSTAT sampling frame. The LCS provides information on the level and sources of household income, including income from work both in the public and private sector, self-employment, agriculture sector, social assistance payments, as well as income from property and remittances. According to estimates based on the LCS, income poverty in Albania was very high, both in level and depth, with one out of every three Albanians falling under the (relative) poverty line, and over one half of them living in extreme poverty. Income poverty was found to be predominantly a rural phenomenon, with 80 percent of the poor living in a rural setting. Poverty is defined as having an income below 60 percent of median income, whereas extreme poverty is defined as having an income below 40 percent of median income. Similar poverty rates are obtained when measures of absolute poverty (e.g., $2/day and $l/day) are used. Rural poverty rates are significantly higher than urbanpoverty rates. The 1998 survey results show that households most likely to be poor are those that are large, where the householdhead is either unemployedor is self-employed in private agriculture, and where the household head has low levels of ed~cation.~Households with three or more children are significantly more likely to be poor, and a large number of children live in poverty. The survey findings also show that almost two- fifths of households whose head has only primary education are poor. Among the households with an employed head, poverty is disproportionately placed among agricultural workers, with 2 out of every 5 poor working in agriculture. Another finding of the 1998 survey is that poverty levels are high among households living on social assistance, and one out of every five poor is a pensioner. 'Alsosee Rashidet al. (2001) and World Bank (1997). 7 The numbers reported by the LCS, however, may be misleading, as the welfare measure derived from the survey is affected by two important sources of bias. First, the LCS does not include the evaluation of own production, which i s expected to be particularly sizable in rural areas and among the poorest households. Second, underreporting o f income is well documented and is likely to lead to overestimating poverty levels, particularly in countries like Albania with a very active informal economy. The LSMS 2002, which i s the primary data source for this document, i s the first Albanian attempt to fill the current knowledge gap by collecting all the necessary information for a full consumption-based money metric at the national level. The LSMS 2002 is part o f a multi-year collaborative project amongst INSTAT, the World Bank and the Department for International Development (DFID) to strengthen the Government capacity to collect, analyze and disseminate statistical data for poverty monitoring and policy evaluation. This first round of the LSMS was conducted by INSTAT in the Spring of 2002 on a sample of 3,600 households and provides an indispensable baseline for monitoring progress in the Government poverty reduction strategy, as outlined in its NSSED. The 2001 pre-census listing provided a much-needed sampling frame. The sample was based on a two-stage cluster design, with 450 primary sampling units (PSUs) selected from the list of census enumeration areas (EAs).' In view of the dramatic and rapid changes that have occurred in the metropolitan area of Tirana, it was decided to update the pre-census listing of dwellings in the capital city. Following the drawing o f the PSUs, 8 households were then randomly selected in each.g The sampling frame was stratified in four regions, which roughly reflect a partition o f the country along agro-ecological as well as socio-economic lines. It is meant to be representative for Tirana, other urban and rural areas, as well as for Tirana and the three main agro-ecological/economic areas (Coastal, Central and Mountain)." The final dataset from the LSMS 2002 includes 1,000 households for the Coastal and Mountain strata, 999 for the Central stratum, and 600 inTirana. A total of 1,640 households are rural, and 1,959 are urban. Additional details on the survey methodology and fieldwork procedures can be found inthe Albania 2002 LSMS Basic Information Document on www.worldbank.org/lsms (forthcoming). As part o f the household survey system, an annual panel on a sub-sample o f LSMS households is scheduled to be conducted inthe Springof 2003 and 2004. Also, a full LSMS on a new cross-section is being planned for 2005 as part of the same project. As the size of the original census EAs varied substantially, in order to get PSUs of similar size, some aggregation andsplitting of the original EAs was performedbeforeselection. During the implementation of the survey there was a problem with the handling of the questionnaires for a householdthat therefore hadto be dropped from the sample. The final sample size is therefore of 3,599 households (16321 individuals) lo The Coastal stratum includes the districts of Lezhe, Kurbin, Kavaje, Mallakaster, Lushnje, Delvine, Sarande, Durres,Fier, andVlore The Centralstratum includes:Kucove, Skrapar, Kruje, Peqin, Gjirokaster, Permet, Tepelene, Devoll, Kolonje, Pogradec, Mirdite, Puke, Malesi e Madhe, Shkoder, Elbasan, Berat, Korce, Mat, and Tirana rural. Finally, the Mountain stratum includes: Kukes, Has, Tropoje, Bulqize, Diber, Gramsh, and Librazhd.Urban Tirana forms a separate stratum. 8 4. Structure of the Report The rest o f the report is organized as follows. Chapter I1provides an overview of poverty inAlbania, and contains a descriptive profile of the poor. Spatial dimensions of poverty, to the extent the data allow, are discussed in Chapter 111. Issues related to labor markets and poverty form the subject matter of Chapter IV. The human capital dimensions of poverty are discussed in Chapter V (health) and Chapter VI (education). The next two chapters contain an analysis of coping strategies: Chapter VI1examines the role o f social protection programs, while Chapter VI11 looks at internal and external migration as ways of coping with poverty inAlbania. Chapter IX concludes. All backgroundandtechnical papers are available withthe authors, andwill also beplacedonthe web for easy access. 9 CHAPTER 11:POVERTY INALBANIA 1. Introduction The multidimensional view o f poverty is so widely accepted in the literature on poverty that re- emphasizing it i s almost redundant. As defined in the World Bank's 2000/2001 World Development Report, poverty encompasses aspects o f deprivation over and beyond mere material consumption, and includes the psychological pain o f being poor, a sense o f vulnerability to external events, and powerlessness toward the institutions o f the state and society. Acceptance o f this view has implications for policymakers at the national and international level alike. At the national level the Government o f Albania has embraced a multidimensional approach to poverty reduction inthe NSSED, including a clear commitment to make progress towards the achievement o f the MDGs. In the international arena this i s reflected most prominently in the endorsement by the international community o f the Millennium Development Goals (MDGs), as set forth inthe United Nations' Millennium Declaration. Inline with this widespread consensus, assessment of poverty inAlbania is based on a multi-dimensional definition o f poverty, and besides income deprivation, poverty i s also defined interms o f inadequacies in a number o f non-income measures o f welfare such as education, health, empowerment, access to basic services and infrastructures. In addition, subjective measurements o f welfare are introduced to map people's perception o f their socio-economic conditions against more objective measures o f welfare. The rest o f this chapter i s organized as follows. Section 2 defines the poverty line and explores issues related to income poverty. Section 3 focuses on non-income dimensions o f poverty, while Section 4 contains a discussion on subjective poverty. Section 5 discusses the poverty profile, followed by a decomposition o f inequality in Section 6. Consumption patterns are described in Section 6, determinants o f poverty are explored in Section 7, and the chapter concludes with main findings and policy implications in Section 8. 2. Income poverty Since the economy o f Albania i s largely rural and informal, and income i s not readily and accurately measurable, income-based measures will provide distorted estimates o f poverty. For this reason, income poverty in Albania is estimated on the basis o f a consumption-based measure. An individual is considered poor if her level o f per capita expenditure falls below a minimum level necessary to meet her basic food and nonfood needs. This minimumlevel o f consumption i s denoted by the "poverty line" and represents the cut-off point separating the poor from the non-poor. Poverty lines can be set in several ways, and include absolute, relative and subjective poverty lines. Absolute poverty lines are defined in relation to a pre-determined threshold or anchored to a nutritional basket or to specific welfare levels, and have the benefit o f allowing for easy comparison across time and groups. Relative poverty lines reflect the degree o f destitution o f a household or individual in relation to the income o f the rest o f the population. And finally, subjective poverty lines incorporate peoples' perception o f their own welfare status. The poverty line selected ultimately depends onthe objectives o f the analysis." The calculation o f the absolute poverty line for purposes o f this assessment is based on an actual consumption bundle o f a sample reference population drawn from the survey, and converted to a monetary measure using the cost-of-basic-needs (CBN) methodology (Ravallion and Bidani, 1994). For the estimation o f the full poverty line, a food poverty line - or the cost o f obtaining a certain minimum Setting a poverty line is seldom uncontroversial, and is often the focus of heatedpublic debate, particularly when it is used as a reference threshold for entitlement programs. For distributions around the line, the choice may be particularly difficult, as even a small shift of the line can have significant impact on the poverty numbers. 10 amount of calories - i s first computed, and then adjusted to include essential nonfood items.The nonfood component i s calculated as the average non-food share o f those households that spend roughly the same amount for food as indicated by the food poverty line (see Annex Ifor details). The food line is computed using the food basket consumed by people in the second to fourth deciles.'* Taking into consideration the FA0 recommendations on the minimum calorie consumption according to age and sex, and adjusting these calorie requirementsto the population distribution in Albania according to the 2001 Census, per capita necessary calorie intake was estimated to be 2,288 calories per day. Calculated in this manner, the food poverty line (or extreme poverty line) is estimated at 3,047 Leks per capita per month. The full poverty line, estimated with allowances for basic nonfood items, i s estimated to be 4,891 Leks per capitaper month.13 Interms of poverty measures, the standarddecomposable indicators inthe Foster-Greer-Thorbecke (FGT) family are used. These include: (i) incidence of poverty, or headcount ratio, reflecting the percentage the o f population whose per capita consumption falls below the poverty line; (ii)the depth of poverty, measured by the poverty gap index, quantifying the average income shortfall of the poor in relation to the poverty line; and (iii) poverty gap squared, a measure o f severity of poverty. the Based on the measures discussed above, one-quarter o f the Albanian population, or close to 780,000 individuals, fall below the poverty line. Extreme poverty - as defined by the food poverty line - exists, and amounts to 4.7% of the population, or about 150,000 individuals, who are unable to meet basic food requirements. Consistent with previous studies, poverty i s found to be higher in rural areas, where the poverty headcount is 66% higher than in Tirana, and 50% higher than in other urban areas. Per capita consumption inrural areas, at 7,224 Leks, is about four-fifths of the consumption levels inurban areas.14 - Consumption-based inequality is moderate and in line with many other countries in the region, with a national Gini coefficient at 0.28. The estimate corroborates previous consumption-based inequality estimates, which are generally lower than income-based measures.N o significant differences exist across locations. The 90/10 ratio, another measure o f inequality, i s moderate at 3.6. l2 Deciles were calculated from the per capita consumption aggregate already adjusted for price differences. The first decile was excluded in order to avoid influencing the basket with patterns that risk being the result of some outliers. However, different reference populations, includingthe poorest 50%, producedvery similar results. l3 The 2002 LSMS collected all the information necessary to calculate the main components o f the consumption aggregate: food consumption (both purchased and consumed from own production), non-food expenses (clothing, household articles etc.), utilities (gas, telephone, electricity, etc), education, health, durables, and housing. Health andhousingexpenses are not includedinthe consumption aggregate, however, for reasons explainedinAnnex I. l4 The magnitude of the gap is smaller than estimated by the LCS 1998. Although drawing comparisons with previouspoverty estimates is enticing, it must be emphasized that, in addition to the obvious discrepancies due to differences in sampling and data collection methodologies, the welfare measurements o f the LCS are not comparable with the consumption-based measure used inthis assessment. 11 Figure 2.1, which graphs cumulative density functions by location, shows the relation between rural and urban poverty. Poverty rates in rural areas are consistently higher than urban rates, regardless o f the poverty line chosen. While first-order stochastic dominance between Tirana and the rest o f urban areas is not maintained, overlapping is minimal and the lines only cross at levels o f consumption above any reasonable poverty line. Figure2.1: CumulativeDensityFunctionsby location poverty line .B - .-m .E - - c 3 L1 n 0 c 0 ._ c .4 - 0 c E .2 - 0 - 20b0 40b0real eobo 1oobo per capita consumption 60bO The absolute poverty line as calculated above i s anchored to a specific consumption basket. In addition, absolute poverty lines can also be constructed in terms o f equivalence (in the sense o f purchasing power parity) to income o f US$2 and US$4 per day.15 Similarly, a relative poverty line can be constructed in terms o f median income o f the population, an approach recommended by Eurostat and widely used across Western European countries. It i s useful to examine the sensitivity o f the estimates o f poverty incidence to the various choices o fthe poverty line. As Table 2.2 shows, the food poverty line sets the lowest threshold, followed by the median-income relative poverty line, the US$2 PPP absolute poverty line, the 60 percentile income relative-poverty line, the absolute poverty line as computed above anchored to a nutritional threshold, and finally the US$4 PPP absolute poverty line. Note that very few people live below the food poverty line - valued at 3,047 Leks. 15 The $l/day PPP line, valued at approximatelyNL 1,90O/month, is not reported, as only a very smallpercentage o f individuals fall below this line. This is common practice in ECA countries, where the $2 and $4 PPP lines are more widely used. 12 The absolute poverty line, at NL 4,891, lies between the two absolute PPP lines used for international comparison purpose. Based on the median-income relative poverty line, 7 percent of the Albanian population consumes less than half of the country's median per capita expenditure. Using the 60 percent o f median consumption as relative cut-off point, the poverty headcount doubles to 13.5 percent. The concentration o f individuals around those points inthe distribution i s evident. Table 2.3 reports the sensitivity o f different poverty lines to shocks and windfalls that change income by an amount equivalent to 5- 10 percent o f current income, which gives an idea o f the number o f individuals that move in and out o f poverty as a result. The middle column (in bold) presents the headcount rates derived from setting each o f the poverty lines as proposed. Income i s allowed to drop by 5 percent and 10 percent inthe columns to the left o f the middle column, and increase by similar amounts incolumns to the right ofthe middlecolumn. The analysis suggests that a large number o f individuals are clustered around the poverty lines. Increasing the poverty line by 10 percent increases the percentage o f poor individuals by 25 to 50 percent, depending on the poverty line used. This i s true for negative shifts inthe poverty line as well. Inother words, a negative welfare shock (equivalent to an increase inthe poverty line, keepingthe distribution constant) and a positive shock (equivalent to a downward shift o f the line) produce more-than-proportionalchanges inpoverty. The importance o f adequate safety nets for this large number of individuals emerges as a public policy priority. Also, inview o fthe relatively low depth and severity o f poverty, small, well-targeted transfers can potentially produce sizable results in terms o f poverty alleviation. Finally, this type o f distribution warns against mechanically usingthe poverty line as a targeting threshold for public programs, as doing so could exclude potential beneficiaries who suffer from the very deprivations that the programs are trying to address. International comparison of poverty Although the $l/day PPP line i s widely used to measure extreme deprivation and i s also the MDG standard, it i s perhaps more appropriate for Albania, l6 like for other ECA countries, to use the $2/day line based on 1996 PPP values for absolute deprivation." l6 See, for instance, "Making Transition Work for Everyone: Poverty and Inequality in Europe and Central Asia", World Bank, 2000 l7 Although international comparisons for most countries are based on 1993 PPP values, 1996 PPP values are used to allow comparability with the ECA publication(referencedabove). 13 Albania Bulgaria Romania Macedonia Georgia Azerbaijan Survey year 2002 2001 1998 1996 1999 1999 $2 PPP 10.8 7.9 6.8 6.7 18.9 23.5 $4 PPP 59.3 31.9 44.5 43.9 54.2 64.2 Table 2.4 shows the poverty rates for both the $2/day and $4/day line in 1996 PPP for Albania as well as for several other countries in the region. The $2/day comparison puts Albania behind countries like Bulgaria and Romania, which are in the next wave o f EU accession countries, and ahead o f Caucasus countries like Georgia and Azerbaijan. The $4/day comparison, however, puts Albania behind Bulgaria, Romania, Macedonia and Georgia, and slightly ahead o f Azerbaijan. Like for any such comparison, however, it must be noted that while some degree o f cross-country comparability i s ensured by the use o f PPP values that reflect the different purchasing power in those countries, because o f the differences in methodologies used inthe estimation o f the welfare measures, the comparisons are somewhat weaker. 3. Non-income dimensions of poverty Before turning to a profile o f the poor, it is important to consider the various non-income dimensions o f poverty and deprivation that can potentially dominate the pure income dimension and result in a general sense o f ill-being (as opposed to well-being), even if money income i s above the poverty line, however defined. We invoke a discussion on this subject by identifying the key non-income dimensions o f poverty and establishing their relevance to Albania. The consensus and commitment among many countries and the international community to the achievement o f the Millennium Development Goals (MDGs) implicitly recognizes the substantial positive repercussions the attainment o f these goals would have on reduction o f poverty and welfare gains for millions o f households, and provide a natural starting point for this identification. Besides eradication of extreme poverty and hunger, MDGs focus on achieving universal primary education, promoting gender equality, improving child mortality, improving maternal health, combating HIV-AIDS, malaria and other diseases, and ensuring environmental sustainability (see Box). Based on the above, specific non-income dimensions o f poverty that merit further inquiry include (i)accesstobasic services; (ii)literacyandeducation; (iii)childmalnutrition; (iv) maternalhealth; (v) labor force participation; and (vi) empowerment and gender. These are discussed inturn. i.Access to basic services The overall picture that emerges from the analysis o f the 2002 LSMS data on access to basic services shows that non-income dimensions o f deprivation compound income poverty, and disparities are more pronounced than are readily apparent by looking at the income dimension alone. Inparticular, the analysis shows that the modernization o f the country which, in many respects, has been triggered by the transition reforms and the recent period o f steady economic growth, has benefited urban areas - particularly Tirana - at a much faster pace relative to the rural areas, which have been left behind. Inmost cases, the coverage o f basic infrastructure services is nearly universal in urban areas, but much less so in rural areas. This is not to say, though, that urban areas are free from problems- often even though access is available in theory, the service suffers so much from problems related with quality that they are virtually not available. 14 A case inpoint is access to water. Less than one in four rural dweller has access to runningwater inside the dwelling [i.e., - . piped water from - _ Fig. 2.2: Access to water , network connection), and less than one I in two has access to any runningwater, 100% inside or outside the dwelling. In Tirana, i 80% on the other hand, virtually everyone has running water (96.6 percent) and most (91.8 percent) have runningwater inside the dwelling. The situation i s similar in other urban areas. For the country as a I 20% whole, these figures are about 66 and 50 percent, respectively. 1 I Tirana Other urban Rural Total In terms o f water quality, however, the I URunning water inside .Running water outside ONo running water 1 situation o f Tirana appears to be worse than other areas, including other urban and rural (Table 2.5). Only one in four households in the capital city think that the water they use is o f good (drinking) quality, compared to 83 and 87 percent households respectively in other urban areas and inrural areas. Also, among those who do receive runningwater, nearly 30 percent o f the residents in other urban areas (and 26 percent in Tirana) report having had water for less than six hours each day. In the country as a whole, not even half the total number o f households with access to running water report having uninterrupted access to this service, and nearly half have water for 10 hours or less per day. The dismal conditions in which many rural Albanians live are reflected also in the figures on access to I improved sanitation. Only 40.8 percent Fig. 2.3: Sanitation o f rural dwellers have a toilet inside their dwelling, while 45.5 percent do not have 100% any access to a piped toilet. In Tirana, on 90% the other hand, 93.3 percent o f the 80% population lives in dwellings with at 70% least one toilet inside. Comparing across 60% 50% regions, the largely rural Mountain 40% region has the worst sanitation problems, 30% with 46.1 percent o f the population not 20% having access to a toilet piped to a 10% sewage network. 0% Tirana Other Urban Rural Total 1 Electricity is another case in point. While , 1 I WC inside the house WC outside, w/piping the coverage of the network is virtually I 10WC outside,wlout piping 21 ~ universal, delivery of the service is highly unreliable, and the situation is 15 worse in rural communities (Table 2.6). Only 13.8 percent o f Albanians report receiving electricity continuously (28 percent inTirana, 7 percent inrural areas) and 72.7 percent report daily interruptions (83 percent inrural areas, 56 inTirana). Among those reporting interruptions, electricity was not available for 8.6 hours on average on atypical day (9.4 inrural areas, 7.7 inurban areas, 5.6 in Tirana). Frequency of power supply interruption Tirana Urban Rural Total Never 28.3 21.7 6.7 13.8 Severaltimes amonth 6.3 8.7 3.4 5.3 Severaltimes aweek 9.8 11.1 6.4 8.3 Every day 55.6 58.4 83.4 72.7 Total 100.0 100.0 100.0 100.0 Box 2.1: MillenniumDevelopment Goals The MDGs-which comprise 8 goals, 18 targets and 48 indicators-reflect the international commitment to cut by half the proportion o f people in extreme poverty by 2015, improve education and health, and preserve the environment. By subscribing to the MillenniumDeclaration, the Government of Albania has taken on the commitment of regularly reporting on the status of Albanian efforts to achieve the MDGs. The National Strategy for Social and Economic Development (NSSED) has already internalizedmany of these goals as targets and indicators for progress. Regular monitoring, evaluation, and reporting on the MDGs also needs to be incorporated into existing and planned M&E systems, as well as on other sub- strategies at both the central and local levels. A recent publication prepared by the Human Development Promotion Center (HDPC) for the United Nations in Albania takes stock of where the country stands in relation to meeting7 MDG goals and 11 of its targets (MDG8 i s the development of global partnerships): The MillenniumDevelopmentGoalsin Albania GOAL HDPC, Albania OTHER SOURCES* MDG1:Eradicate extreme poverty and hunger Probably Inadequate datato predict MDG2: Achieve universal primary education Probably Unlikely to be achieved MDG3: Promote gender equality and empower women Possibly Likelyto be achieved MDG4: Reduce child mortality Probably Unlikely to be achieved MDG5: Improve maternalhealth Possibly Indicator dataimproving MDG6: Combat HIV/AIDS, malaria and other diseases Possibly Unlikely to be achieved MDG7: Ensure environmental sustainability Unlikely Too hardto tell *From Business Plan For Supporting The MillenniumDevelopment Goals InEurope and Central Asia", " ECSHD, August 2002 The report also includesperformance summaries (see second column intable above) inwhich, based on current status and past trends, the authors attempt to quantify the likelihood that Albania will meet each goal on time. The authors are of the opinion that "there is a good chance that Albania can achieve the MDGs", but only if "improved policies, stronger institutions and additional resources" are put in place. However, according to other sources, as reported in the column to the far right, achieving these goals appears less easily within Albania's reach. It should also be noted that there are a number of lacunae in the availability of proper indicators to measure and track progress. One case in point is the measurement of poverty: without the existence o f reliable baseline estimation, it was unrealistic until now to commit to a specific target. 16 Table 2.7 reports five indicators of non-income poverty that collectively constitute the Unmet Basic Needs (UBN): water and sanitation, housing conditions (as perceived by the household), energy supply, crowding o f the dwelling, and education of the household head. A household is defined as UBN-poor when two or more of these basic needs are unmet, and is defined as extreme UBN-poor when three or more are unmet. Tirana Urban Rural Total 1. Inadequatewater and sanitation (*) 0.5 2.6 28.6 17.5 2. Inadequatehousing(**) 8.5 6.3 16.5 12.5 3. Inadequateenergy supply (***) 1.7 9.0 18.1 13.5 4. Crowding (more than 3 persons/room) 10.3 15.6 18.6 16.7 5. Education(hhheadwl Drhnarv or less) 34.7 47.0 74.8 61.2 Poor (two or more UBN) 11.5 16.6 47.2 33.8 ExtremePoor (three or more UBN) 2.3 3.2 18.3 11.9 Non poor (one or no UBN) 88.5 83.4 52.9 66.2 The ordering o f the rates of UBNpoverty by location does not differ from the money-metric poverty rate presentedinthe previous section. The differences between rural and urban areas, however, are much more pronounced. Rural areas fare worse, often by a large margin, than urban areas. UBN poverty is three times greater in rural areas than in urban areas. The difference in extreme UBN poverty i s even more pronounced, with a six-fold gap betweenTirana and rural areas. These indicators, while important in their own right as basic needs, also have longer-term effects on poverty. The lack o f good water and sanitation may not only cause illhealth, but also impair the ability to generate income and build human capital (reflected, for example, in the ability to learn or to attend school). The lack o f a reliable supply of electricity hinders the profitability (or outright feasibility) of productive investments, lowering economic growth, often with long lasting effects. ii. Literacy, EducationalAttainment and Enrolment Rates Education attainment o f a people, in addition to being a well-established indicator of welfare, i s also a policy tool that can break the cycle of poverty for many. By international standards, literacy levels in Albania are high, particularly among the younger, urban population (Table 2.8), but fall substantially among rural elders, with only about halfreporting to be able to read and write with confidence. Table2. 8: Literacy by agegroup and location Tirana I Otherurban I Rural Total 17 Significant differences exist in educational attainment between urban and rural populations, and while less than one-forth o f the rural population has completed secondary school, about two thirds o f the adult population in urban areas holds a secondary or vocational school diploma (Table 2.9). The rural-urban difference is most striking at the university and post-graduate levels: in rural areas only 1 adult in 40 hold a university degree, compared to 1 in 5 in Tirana, and 1 in 10 in other urban areas. Table 2. 9: Educational attainment Average schooling in the age group 21 years and above is 8.5 years, with adults in rural areas on average completing 3.5 year less o f formal education compared with their counterparts in Tirana, and about 2 years less than adults in other urban areas. Net primary school enrolment is quite high- at 93.3 percent - with no large differences between urban and rural areas. Gross primary enrolment rates are also high at 99.8 percent. These figures, however, likely hide considerable differences in attendance and in quality o f education. Evidence from qualitative work indicates that rural children attend school for only 2-3 hours a day, and experience absent and/or inadequately qualified teachers, lack o f teaching material and poor infrastructure.'* Secondary net school enrolment rates provide an even more stark picture o f the state o f the educational system inAlbania, with less than forty percent o f children inthe appropriate age group enrolling in school nationwide (compared to 43.9 in 1998). The gap in enrolment rates between rural and urban areas i s most evident in post-primary education, with less than 3 out o f 10 secondary school-age children enrolled in rural areas compared to 7 in Tirana and 6 in other urban areas (Table 2.10). Urban Rural Total Total Urban Tirana Other Urban Net enrolment 1998 2002 2002 2002 1998 2002 1998 2002 Primary 97.2 92.9 92.2 93.2 93.4 93.5 94.6 93.3 Secondary 71.2 62.2 70.8 59.5 30.2 24.9 43.9 38.7 Primary 99.2 1 97.5 99.6 100.2 I I 99.8 Secondary II70.3 I 81.5 II 66.8 II 28.3 I I 43.9 l8For analogousfindings conceming the unevenquality of education, see also UNICEF (2000), Assessment of Social and Economic Conditions of Districts in Albania. 18 Of the children not enrolled in school, more than half reported having little interest in studying, while about another quarter gave the need to work as the reason for not going to school. About two thirds o f this group o f respondents are involved in agricultural labor. iii. Child malnutrition Malnutrition refers to an unhealthy physiological condition linked to imbalances, excesses or deficiencies in energy, protein, or other nutrients (e.g. iodine, vitamin A, iron), and is frequently a part o f the vicious cycle that also includes poverty and disease. Inadequate energy intake or absorption leads to forms o f malnutrition that, particularly for children, are usually defined as follows: Wasting (Low weight-for- height), indicative o f a recent and severe weight loss, often associated with acute starvation and/or severe disease; Stunting (Low height-for-age), indicative o f the failure to reach linear growth potential as a result of sub-optimal long-run health and/or nutritional conditions; and Underweight (Low weight-for-age), indicative o f low body mass relative to age, resulting from such factors as low levels of food intakes, past episodes o f under-nutrition, and poor health conditions. Table 2.11 presents Z-score indicators (which express the anthropometrics value as the number o f standard deviations below or above the median value o f a reference population) for wasting, stunting and underweight, The cut-off points used in this assessment are -2 Standard Deviations (SD) from the reference population to define moderate malnutrition, and -3 SD for severe malnutrition. Stunting Wasting Underweight IBV location Rural 34.9 18.6 11.8 2.9 15.5 1.7 Urban 31.4 19.3 8.3 1.2 11.3 0.8 Male 34.4 19.4 12.2 2.8 17.4 1.9 Female 32.9 18.1 8.5 1.7 9.6 0.6 Total 33.7 17.3 10.6 2.3 14.0 1.4 As the above indicates, one child in every seven in Albania is severely stunted, while one in three is moderately stunted. The percentage o f moderately wasted children i s lower, at about 10 percent." In line with other indicators, rural areas generally fare worse than urban areas inall three measures. Although the differences are not large, boys appear more likely to be undernourished than girls. Age-wise, children in the 24-35 month age groups exhibit the highest levels o f stuntingamong all children below 5 years. iv. Maternal health/antenatal care Virtually every woman i s assisted at delivery by a skilled health personnel (i.e., doctor or trained midwife). The vast majority o f women also receive antenatal care by skilled health personnel during pregnancy. Rural-urban differences are acute, however, and whereas about 10 percent o f rural women report not receiving any form o f medical assistance during pregnancy, only 2 percent o f the women in urban areas have the same experience (Table 2.12). An additional 2.8 percent o f women in rural areas report assistance from unskilled personnel only, compared to 2 percent in urban areas. Regionally, the l 9A total o f 946 children are includedin the analysis, out of 1,342 measurementattempts. Many observationswere excludedbasedon the default thresholds suggestedby the software `Anthro.' 19 situation i s worst in the remote North-Eastern region o f the country, where 17.6 percent o f women have no access to antenatal assistance, and an additional 4 percent relies on unskilled personnel. Among those who do get care, women in rural areas make 3 antenatal visits to a provider compared to more than 5 visits by women in urban areas. Skilled Unskilled No care Urban 95.9 2.0 2.1 Ofwhich: Tirana 93.9 6.1 0 Other Urban 96.7 0.3 2.9 Rural 87.8 2.8 9.4 Tirana 93.9 6.1 0 Coastal 92.1 0 7.9 Central 92.6 2.8 4.7 Mountain 78.4 4.0 17.6 Total 90.6 2.5 6.9 v. Empowerment and Gender Promoting gender equality and empowering women is one o f the Millennium Development Goals. One indicator used to measure progress toward this goal i s the enrolment rate o f girls and boys in primary and secondary education. The LSMS survey data does not show any significant 90 gender differences in enrolment rates among 80 Albanian school children at either the primary school 70 or secondary level (Figure 2.4). Literacy rates o f 60 individuals age 15-24 are around 98 percent, and 50 there appear to be no significant gender differences 40 on this count as well. 30 20 Another good indicator o f gender disparity i s the 10 share o f women in wage employment in the non- 0 agricultural sector. An analysis o f the labor force Male Female Total participation shows significant gender differences, DPrimary enrolment .Secondary enrolment 1 with women's participation being about two thirds that o f men's. Irrespective o f location, the share o f women holding full time jobs i s much lower, and only 23 percent o f women report working full time, compared to 45 percent men who report working full time (Table 2.13). The different unemployment rates suggests that a large number o f women participating only marginally in the labor force, and are either underemployed in farming activities or are discouraged workers. 20 * Individuals **** Counts who work an average of 35hrs or moreper week are considered to befull time workers. * Counts discouragedseasonal/laid off workers as unemployed. individuals who work less than I 5 hoursper week in agriculture as unemployed. Compared with urban areas, the share of working women in total labor force is substantially higher and the gender gap narrower inrural areas, again reflectingwomen's participationto family farming as unpaid part-time laborers. Relaxing the "one-hour criterion" definition of employment, unemployment rates among women in rural areas wok out to 19.4 percent, significantly higher than 15.3 percent unemployment among men (Table 2.14). *Individuals ** who work an average of 35hrs or moreper week are considered to befull time workers. ***Counts discouraged/seasonal/laid off workers as unemployed. Counts individuals who work less than 15 hoursper week in agriculture as unemployed. vi. Access to assets Land ownership is very widespread in Albania, following the rapid "privatization" o f the early 1990s. Having access to land, however, does not appear to be a sufficient condition for escaping poverty; in fact, as Table 2.15 shows, land ownership is more widespread among the poor (91 percent of poor rural households own land) than among the non-poor (87.5 percent). Only in urban areas i s land ownership more common among the non-poor households (7.4) relative the poor (6.0 percent). The averagefarm size inAlbania is very small, a mere 0.72 hectares for rural areas according to the 2002 LSMS. Absolute differences between poor households (0.6 hectares on average) and non-poor households (0.76 hectares) are not striking.Inper capita terms, however, the average size of land holdings among the rural non-poor is 1,812 square meters, 74 percent more than 1,039 square meters, the average size of land holdings among the poor. 21 Rural Urban Total Not Poor Total Not Poor Total Not Poor Total Poor Poor Poor Percent of Householdsthat OwnAgricultural Land 87.5 91.0 88.3 7.8 5.9 7.6 48.9 62.4 51.5 Livestock i s another important asset for rural people in Albania, and accounts for a sizeable share of total agricultural production. More than 86 percent of households engaged in agriculture own some livestock, with more poor households (92.7 percent) owning livestock relative to the non-poor (84.6 percent). The poor, however, own a smaller number of heads compared to the non-poor. As far as durable assets are concerned, there are huge differences in total value between urban and rural areas, with individuals in urban households controlling approximately 45,000 Leks worth of durables compared to only 27,000 Leks in rural households (Table 2.17). Tirana Other urban Rural Total Total per capitavalue of householddurables (Leks) 42,071 46,810 26,923 34,680 Difference between the poor and the non-poor households in the ownership of durable assets is striking, with the extreme poor individuals owning, invalue, only one eighthof the assets controlled by non-poor, and about one half of the average among the poor (Table 2.18). Non-poor Poor I Extremepoor I Total Total per capitavalue of householddurables (Leks) 43,008 11,374 5,100 34,680 4. Subjective poverty The LSMS questionnaire contains a module on subjective poverty, and several questions on various subjective measures and perception of poverty were administered to household heads. Responses to the subjective questions present a rather gloomy picture, with 9 household heads in 10 reporting to be not satisfied with their current financial situation, with about half in this group reporting being highly unsatisfied (Figure 2.5). As Figure 2.6 shows, dissatisfaction i s high even with respect to basic needs requirements. 22 Fig2.5:Satisfactionwithament6 r m d situation 1 Fig. 2.6:Householdbasic needs requirenents 100% 80% 60% B c I C ! 40% 20% Tirana Uban iaJral Tdd 0% Tirana urban Rural Total OMore than adequate .lust adequate DLerr than adequate Table 2.19 reports average consumption levels according to the household-head self-assessment of financial condition. The relationshipbetween objective and subjective status i s monotonic and of the right sign, regardless o f location. The per capita expenditure level in households with a financially dissatisfied head i s less than half of the average expenditure of those households in which the head i s fully satisfied with his current financial conditions. Total 9403 I 100.0 I 8468 I100.0 I 7212 I 100.0 I 7801 I100.0 Among the subjective indicators collected, the LSMS also included a modified Cantril scale question, which referred explicitly to peoples' evaluation of their economic welfare rather than satisfaction with life in general.20Respondentswere askedto place themselveson the rungs of a 10-stepladder, the bottom one signifying the poorest and the highest one the richest. Overall, 26.2 percent of the respondents classified themselves in the bottom two rungs of economic welfare, while only 6.9 percent classified themselves as being in the top three rungs. Most households (23.7 percent placed themselves in the third lowest rung, followed by 21.3 percent in the fourth lowest rung and 16.1 percent in the fifth rung. If the bottom two rungs are classified as beingpoor, the subjective poverty measure o f 26.2 percent comes remarkably close to the objective measure. However, it's not clear ifthe two groups are the same. 2o Specifically, the question was: "Imagine a 10 step ladder where on the bottom, the first step, stand the poorest people, and on the highest step, the tenth, stand the rich.On which step are you today?" 23 Overall, there appears to be a moderate correspondence between objective and subjective measures. The rank correlation between the per capita consumption variable and a 7-rung scale2' o f the subjective measure is 0.49 (Table 2.20). Among those rated as "objective" poor, only about half perceive themselves as being poor. A similar percentage of individuals who perceive themselves as poor, are poor according to the objective money metric measure. Among the non-poor, 1 individual in 5 perceives herself as poor. Overall, it appears that people do not equate the concept o f poverty exclusively with income. Subjectivepoor Objective poor Non-poor Poor Total Non-poor 81.9 18.1 100 82.8 51.6 74.6 Poor 50.1 49.9 100 17.2 48.4 25.4 Total 73.4 26.2 100 100 100 100 Responses to subjective questions contained in the LSMS questionnaire also provide some indication o f trends in living standards in Albania in the face o f sustained economic growth. The responses show that only 1 in 10 household heads report worsening in personal financial situation in last 3 years, while one- third o f all households report improvements.At the same time, more than 20 percent o f households think that their life has worsened in general, and more than 40 percent report health care as their primary concern. The non-poor seem, by and large, to have a positive (or at least non-negative) perception o f the changes in their economic well-being over the past three years, while among the poor a large portion (a majority among the extreme poor) feels their situation has been deteriorating. Few o f the poor think their situation has improved inthe past three years. About two third o f the extreme poor feel that their situation has infact worsened, and the remaining one third feel it has not changed. It's difficult to come to any robust conclusions based on the above, though it does appear overall that Albania is a country in recovery, income-wise, but the poor feel excluded from this improvement. At the same time, more Albanians appear to be experiencing deterioration in their living conditions insofar as non-income dimensions o f daily living are concerned. 5. A profile of poverty i.Thedemographicsofthepoor Average household size inAlbania i s 4.3 members per household. The average household size among the poor i s 5.7, o f which 2 members are below the age o f 15, compared with a household size o f 4 among the non-poor, o f which 1 i s below the age o f 15. The dependency ratio is consistently higher among the poor, irrespective o f location. Household heads in poor households are on average 3.6 years younger than their non-poor counterparts, and possess only 6 years o f schooling, compared to 8 years among the non-poor. 21 Given the limited number of respondentswho classified themselves inthe top four rungs, we collapsed them into one single group. 24 Table 2. 21: Main characteristicsof the Door Mean number of workers includes workers 15 and over only. Meanyears of school of household head assumes that the household head completed the maximumyears possible in the second highest level of schooling completed. Dependency ratio=(number of children age 0-14 $.number of elders age 60 and over) / number of household members age 15-59 Poverty rates are highest among large households with 7 or more members, with rates above 50 percent regardlessof location (Figure 2.7). Householdswith 4 membersor less representabout 43 percent of the population, but constitute only 15 percent of the poor. About 40 percent of the poor live in large householdsof 7 or more persons (while only representing 17 percent of the population). I ~ Figure 2.7: Poverty rates by household size and location I 60 5 40 50 Ill 85 30 a 20 10 0 I 1 member 2 members 3 members 4 members 5 members 6 members 7 or more members Poverty incidence among younger people is above the national average, and is highest among rural children. In rural areas, 4 out of every 10 children under the age of 5 live in poverty. Almost half o f the 25 poor in Albania are below the age o f 21. Elders in Tirana exhibit some o f the lowest rates, constituting a headcount o f 12 percent.22 Poverty headcount Tirana Other Urban Rural Total % of population Poverty share ii.Poverty,dwellingconditionsandaccesstoservices There appears to be a strong relation between non-income dimensions o f deprivation and income poverty. Even after controlling for urban-rural differences, the poor benefit less from access to public and private services. Poor individuals are twice as likely than their better-off counterparts to have no access to adequate sanitary facilities, and are significantly less likely to have access to running water. They also live in more crowded dwellings: the difference in the crowding index between poor and non-poor is staggering, with almost half o f the poor sharing a room with two or more household members, versus only 20 percent among the non-poor. As expected, power supply shortages, instead, cut across economic lines, affecting poor and non-poor individuals almost equally. Among the poor, eight individuals in 10 have no access to a private telephone. Sanitation WC inside 92.0 48.3 68.0 80.7 22.9 41.4 61.2 WC outside, wl piping 5.6 13.1 9.7 10.4 14.4 13.1 10.6 WC outside, wlout piping 2.2 38.4 22.1 8.6 62.4 45.2 28.0 Access to telephone Both fixed andmobile 24.5 1.4 11.8 3.9 0.4 1.6 9.2 Fixed line only 27.6 1.7 13.4 17.1 0.0 5.5 11.3 22 Note that demographic profiles are generally sensitive to the equivalence scale assumptions made in the construction of the welfare measure. Results of the sensitivity analysis performed (see Annex II),however, show that the poverty rankings are robust to different equivalence scale assumptions. 26 Mobile only 25.1 43.3 35.1 22.3 10.6 14.3 29.8 No telephone 22.8 ~~-~~ 53.6 39.7 56.7 89.0 78.6 46.6 Crowding (persons per room) Less than 1 8.1 6.1 7.0 0.8 0.5 0.6 5.4 1to 3 76.2 69.9 72.6 51.0 50.7 50.8 67.2 3 and over 15.7 24.0 20.3 48.2 48.7 48.6 27.5 Distance from health center 10minutes or less 60.5 44.0 51.4 49.0 29.7 35.8 47.4 11-29minutes 36.2 36.8 36.6 48.6 39.7 42.5 38.1 30-59 minutes 1.9 7.0 4.7 1.5 9.6 7.0 5.3 One hour or more 1.4 12.2 7.4 1.0 21.1 14.7 9.2 BOX2.2: No payphone? There is Vodafone! The booming mobile telephone business, in the face o f a limited, malfunctioning land system, i s all too evident in Albania. Take a stroll in Tirana and it will become clear how Albanians o f all ages are increasingly relying on and sporting the latest technology in mobile phones. As the LSMS data shows, almost two thirds o f the better-off population o f Tirana has access to at least one mobile phone in their household, and about the same proportion have access to a fixed line at home. The situation is somewhat different for the poor: four out o ften poor residents o f Tirana have no access to a telephone, although still about one third have access to a mobile phone in their household. In rural areas, a higher percentage o f both poor and non-poor have access to a mobile phone than to a land line, most likely reflecting the low coverage o f land services in more remote areas. The difference between poor and non-poor in rural areas is sizable, with the poor significantly lagging behind in their access to telephone services, being this fixed or mobile. Urban Albanians spend an average o f 5 percent o f their monthly budget in telephone calls. Even in rural areas, the budget share going to telephone expenses is sizable, at 3 percent. However, in view o f the limited access to private phones in these areas, local telephone providers are the most likely beneficiaries. While mobile phone seem indeed to provide an alternative to fixed lines for better-off rural residents, the prohibitive costs o f the service makes it improbable for it to also represent a viable alternative for the poor, and the rural poor in particular. For those who can't afford a mobile phone, a phone card and a payphone is still the most likely alternative to stay intouch. iii.Poverty andeducation The strong link between poverty and low educational levels is well established, and i s confirmed by the analysis of the LSMS data. Primary school enrolment rates are lower among the poor, and lowest among the extreme poor. The pattern is much more pronounced for enrolment rates in secondary school. Among 27 the poor and extreme poor, only about 2 children in 10 enroll in secondary school, versus 5 in 10 among the non-poor. Extremepoor Poor Non-poor Total Net enrolment Primary 88.6 91.6 94.1 93.3 Secondarv 19.8 19.4 46.7 38.7 Gross enrolment Primary 90.9 97.1 101.1 99.8 Secondary 21.0 21.6 53.1 43.9 The following graph gives a more visual representation o fthe growing gap inage-specific enrolment rates for the poor and non-poor. Although the trends in secondary school enrolment rates are a matter of concern across the board, irrespective of socio-economic status, the magnitude o f the dropout rates between the ages of 14 and 18 among the poor are particularly striking. Fig. 2.8: Enrolment and Poverty I I I 120 88 100 d F .C % 20 ,IW 0 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 i Age 1--tNonuoor Poor The incidence of poverty among adults holding a primary school diploma or less is double the poverty rate among individuals with a vocational diploma or higher. More than 80 percent of the poor did not complete secondary school or hold a vocational school diploma. 28 Table 2. 25: Povertv Measures and Shares bv Educational attainmentand Location Educational attainment for people 21 and older Poverty measures Yo of Poverty Tirana Other Urban Rural Total population share No schooling 19.0 20.5 30.7 26.6 3.5 4.1 Primarv 26.7 26.1 30.8 29.3 61.6 79.4 Secondary 10.5 14.3 19.5 15.6 15.8 9.7 Vocational 2y 14.2 14.4 17.1 15.0 1.5 0.8 Vocational 4y 10.9 8.2 16.5 12.0 10.7 5.2 University and higher 2.9 2.6 3.8 2.9 6.9 0.7 Total 15.9 18.8 28.1 23.7 100.0 100.0 Only 20 percent o fthe poor continued their studies beyond the primary school, while among the non-poor the percentage is above 40 percent. The gap in educational attainment o f the poor in relation to the non- poor is even more evident in Tirana, where more than two thirds o f non-poor adults hold a post-primary diploma (and 25 percent have a university degree), compared with only one third among the poor (and only 3.5 percent are university graduates). Although still substantial, the differences between poor and non-poor in other urban areas, and in rural areas in particular, are less pronounced, perhaps reflecting differences inreturns to education between Tirana and the rest o f the country. Fig.2.9: Educational attainment by economic status and location ~ 100% 90% 80% 70% 'HUniversityand higher 60% UVocational 50% 0Secondary 40% 0Primary0yrs ~ 30% HPrimary4 yrs 10NoSchooling 20% 10% 0% Nonpoor 1 Poor Nonpoor 1 Poor Nonpoor 1 Poor ~ ~ Tirana Other Urban Rural I Total ~ In addition to the gap between poor and non-poor, gender appears to add an additional wedge in educational attainments. At equal economic status, adult males are better educated than their female counterparts. Almost half o f non-poor male hold a post primary school diploma (11 percent have a university degree), compared to one-third o f the non-poor female adults (6 percent have a university degree). Among the poor also, the share o f men with higher education is higher compared to women. 29 I Fig2.10: Educationalattainment ofthe poor by gender 100% 80% 60% UVocational 0Secondary 40% Primary 8 yrs Primary 4 yrs 20% - No schooling 0Yo Nonpoor 1 Poor Nonpoor Nonpoor Poor ~ I Male I Female Total Achieving the MDG goal on universal primary school enrolment i s within the country's reach. Much more challenging, but perhaps bearing the most long-term impact, i s to increase secondary school enrolment and vocational training. Post-primary education has a significant impact on poverty reduction, and this should continue to be an important element ofthe government strategy. Stunting Wasting Underweight Moderate Severe Moderate Severe Moderate Severe Poor 36.0 17.2 11.4 3.3 17.6 2.5 Nonpoor 32.5 19.9 10.3 1.8 12.2 0.7 Malnutrition i s also found to be somewhat more common among poor under 5 children, than among non poor. The differences are however not large, and are statistically significant only for moderate stunting and underweight. iv.Povertyand Employment Irrespective of how employment is defined, the poor exhibit substantially higher unemployment rates when compared to the non-poor. The unemployment rate among the poor based on the standard ILO definition - at 14.2 percent - is almost double the rate among the non-poor. Among the extreme poor the unemployment rate is virtually triple. Using the relaxed definitions, unemployment rates among the extreme poor is above 30 percent. A significantly higher proportion of marginal workers - e.g. discouraged, seasonaland laid off workers -can be found among the poor. 30 Table 2. 27: Labor Force Participationfor Poor and Non-poor (Age >14) I Non-Door I Poor IExtremeDoor I Total Total Working 54.3 52.5 43.1 53.9 Working Full Time* 34.3 28.9 20.0 33.1 Working PartTime 20.0 23.6 23.0 20.8 Unemployed 5.O 8.7 13.4 5.9 Discouraged, seasonaland laidoffworkers 3.6 5.3 6.4 4.0 UnemploymentRate 8.5 14.2 23.7 9.8 UnemploymentRate 2** 13.7 20.9 31.5 15.4 UnemploymentRate 3 *** 17.7 24.0 33.1 19.1 The incidence and severity of poverty are much higher in households where the household head is unemployed. In each o f the three areas for which the LSMS is representative (Tirana, Other Urban, and Rural) the poverty headcount i s almost twice as large for these households as compared to households where the head has a job. A similar difference can be noted in the poverty gaps in both rural and other urban areas. In Tirana, the difference in the severity of poverty between the two groups is even more pronounced, the gap beingmore than three times larger ifthe household head is unemployed. EmploymentStatus of Povertymeasures HouseholdHead Tirana Other Urban Rural Total Headcount Unemployed * 31.7 28.6 I 51.0 I 36.2 Employed 16.4 15.7 I 28.3 I 24.1 Povertygap Unemployed * 9.8 6.0 12.1 8.5 Employed 3.1 3.5 6.5 5.5 The numbero fpeople employed per household does not seem to be linkedto poverty inthe same manner across different locations. Outside o f Tirana one source o f income seems sufficient to reduce significantly the likelihood ofbeinginpoverty, whereas inTirana at least two sources of income are neededto obtain a similar result. In rural areas, as additional employment i s likely to conceal larger underemployment in family farms, no straightforward relationship exists between the number of employed people in the household and poverty. 31 Number of employedpeople in Location household Tirana I Other Urban 1 Rural Total 1 0 20.6 26.3 39.7 29.3 1 19.5 18.6 26.3 21.6 2 or more 14.5 17.7 29.6 26.3 Total 17.8 20.1 29.6 25.4 v. Povertyand incomesources Table 2.30 shows the different composition of income of poor and non-poor households. The main source of income for poor households is agricultural business (37%), followed by wage employment (27%). For non-poor households dependent work i s the single most important source (3 5%), followed by agricultural (25%) and transfers (14%). Non-agricultural business provides only 6% of poor families' income, slightly less than transfers (8%) and "Ndihme Ekonomike" (7%). Compared with poor households, non-poor households derive a significantly higher share of their income from transfers. For both groups, pensions provide about 14% o f the total. Likewise, no difference can be noted in the share o f income provided by unemployment benefits (less than lYo), maternity, social care and other minor public transfers and other sources (both less than 1%). - E: m 28 % + $ $ % - 5 2 % E m & = Q ) -2 m Q) 0 % 2 i?i m 3 ; E E 2 6 5 5 BL s 3.5 % L . c 3E: . c z z0. Y y m = o 5 0 2" 51 s +2 z g & 5 0 Non-Poor 34.6 24.8 8.5 14.4 2.0 13.7 0.4 1.3 0.5 Poor 26.8 36.7 5.7 7.9 7.4 14.1 0.5 0.6 0.3 Total 32.6 27.8 7.8 12.7 3.4 13.8 0.4 1.1 0.4 InTable 2.31, families are classified according to the main source of income, Le., the one that provides more than 50% of the More than one-third of the families participating in the LSMS get most of their income from wage employment (this share is much higher in urban areas, particularly in Tirana - 64%). In this group, the headcount index of poverty i s about 21%, varying from 19% in urban areas to 24% in rural ones. As expected, poverty is most severe for households relying on social assistance ("Ndihme Ekonomike", unemployment benefits, maternity, social care and other public transfers), where the headcount index is 46%, and is higher in urban areas (48%) than inrural ones (43%). Poverty is also severe for households relying mostly on agricultural businesses (34%, all rural poverty) and among rural pensioners (32%). Among those relying on pensions, rural households are remarkably poorer, with a headcount index of poverty reaching 32%, compared to 22% in urban areas. The incidence of poverty is lower in families 23This implies that about 5% of the households are not considered in the table, as no single source reaches this threshold. 32 depending on non-agricultural business (14%) and on transfers (17%). Poverty rates among households relying on transfers are higher in urban areas (21%) than in rural areas (15%). I Urban I Rural I Total Headcount Headcount Poverty N'Ob'' IPoverty Index N.Obs. Paverty Index N.Obs. Index lother --nla-- 3 --nla- --nla-- 6. DecomposingIncomeInequality The process o f economic transition has often raised concerns about its potential negative effects on inequality. Mounting evidence exists in ECA countries in support o f such view, and o f how increasing inequality may ultimately hinder the growth process and undermine poverty alleviation. Much o f the increase in inequality in transition economies can be linked to the collapse o f guaranteed employment and the dramatic drop in the relative or absolute income o f many unskilled workers. In this respect, many would and have argued that an increase in inequality can actually be a welcome outcome, as it reflects an evolving incentive structure that rewards individual productivity and differences in individual human capital. However, regardless o fthe view taken, increasing inequality is becoming a deep social problem in the ECA region, and i s often linked to deteriorating social cohesion and poor governance. Inequality i s generally decomposed along two lines: (a) within and between population subgroups; and (b) by income source. While a Theil coefficient is used for the former type o f analysis, the latter decomposition is based on the Gini coefficient. The aim o f the proposed decomposition is to explore the nature o f the association between the different components o f income and inequality. Inparticular, we are interested in knowing whether transfers, both public and private, have played a stabilizing role and lead to a reduction in inequality. As reported earlier in the chapter, consumption inequality inAlbania is relatively moderate, with a Gini coefficient o f around 0.3. However, income inequality i s significantly higher, with a Gini of per-capita total income o f 0.58. Results o f the decomposition are reported in Tables 2.32 and 2.33, which report, for each income component k, the shares in total income, sk, the concentration coefficient, and the component contribution to total inequality, Pk. The coefficient E gives the marginal contribution o f each k source to total inequality. When the concentration coefficient of source k i s greater than the overall Gini coefficient, an increase in income source k (holding everything else constant) will increase inequality. Table 2.33 shows that inequality i s (strongly) positively correlated with non-agricultural business income. The relative concentration coefficient i s significantly higher than the overall Gini coefficient (0.99 versus 0.58), the percentage contribution to overall inequality i s high (Pk i s equal to 43%) and the marginal contribution is positive. Dependent work explains a considerable portion o f inequality -with a percentage contribution above 30% - but its negative marginal contribution makes wage income inequality reducing. 33 The contributions to inequality o f all remaining sources are significantly lower, and their marginal contributions remain negative, reflecting their potentially inequality-reducing role. Among these sources, at the margin, pensions have the strongest equalizing effect (although its overall contribution i s relatively small), while private transfers have a neutral marginal impact on inequality: an increase in the income share from this latter source would have an insignificant effect on overall inequality. An increase in the share o f income from social assistancewould have a moderate marginal impact on overall inequality but, giventhe small contribution ofthe component to overall inequality, only limitedequalizing potential. Table2.32: Gini decomvosition bv incomesource Gini Share of incomeMean income (Leks) Gk Pk Ek Total income 0.579 9088 Dependentwork 0.735 0.354 3218 0.514 31.428 -0.040 Anricultural business 0.820 0.147 1333 0.461 11.681 -0.030 Non-agriculturalbusiness 1.095 0.251 2281 0.987 42.760 0.177 Transfers 0.899 0.107 968 0.565 10.380 -0.003 Pensions 0.755 0.113 1031 0.248 4.852 -0.065 Social assistance 0.902 0.019 174 0.037 0.122 -0.018 Income data, particularly from sources such as self-employment and businesses, are traditionally fraught with measurement problems. In order to assess the sensitivity o f the results to potential problems in this component, and in view of the unusually high contribution to inequality of this source, this component is set to zero and total income redefinedaccordingly. Table 2.33 shows the related results. Inthis scenario, overall inequality i s lower, and the overall Gini coefficient is 0.50. The marginal contribution of dependent work and transfers is now positive. On the contrary, and in line with the results in Table 2.33, both pensions and social assistance-with negative marginal contributions -continuetohaveanequalizingrole, althoughboth-giventhelowP-haveverylittlepotentialtomake a dent on overall inequality. The overall results point to potentially favorable - albeit small - re-distributive effects of fiscal policies and expanding social assistance programs to reduce inequality. However, given the limited resource envelope o f social spending and the very low contribution of this component to overall income inequality, it is highly unlikely that the government will be able to affect the distribution by increasing the social assistancebudget. A more viable action for the government would be to reduce poverty and inequality by improvingthe targetingeffectiveness of existing social programs. Table 2. 33: Gini decomDosition bv incomesource fnon-apricultural income assumed to be zero) Share o f income Mean Income (Leks) Gk Pk Ek Total income 6807 Dependent work 0.473 3218 0.572 53.693 0.064 Agricultural business 0.196 1333 0.506 19.666 0.001 Non-agriculturalbusiness 0.000 0 Transfers 0.142 968 0.628 17.709 0.035 Pensions 0.152 1031 0.294 8.851 -0.063 Social assistance 0.025 174 0.068 0.343 -0.022 34 7. Consumptionpatterns Table 2.34 reports budget shares, by location, of the main expenditure components. Contrary to what is observed in other transition economies, food shares are still remarkably high at almost 63 percent, with individuals inrural areas allocating up to two thirds o f their budget to food.24 Utility expenses account for about 12 percent o f total budget, with significantly higher shares in urban areas, and in Tirana in particular. The larger utility shares in urban areas derive primarily from higher electricity and (mobile) phone bills, while over 40% o f the utility expenses in rural areas relate to firewood and fuel usage. The share o f the budget going to education expenditures i s similar to many other countries, and it i s highestin urban areas other than Tirana, reflectingthe higher costs o f higher education for lack o f local facilities. As expected, a large share o f food consumption in rural areas (31 percent) is derived from own production, highlightingthe important role played by family farming in ruralhousehold economies. Tirana Other urban Rural Total Food 58.5 59.1 66.3 62.8 Nonfood 24.8 24.0 21.2 22.6 Utilities 14.6 13.7 10.8 12.3 Education 2.1 3.2 1.7 2.3 Total 100.0 100.0 100.0 100.0 Eating out in restaurants and consumption o f drinks and coffee in bars i s widespread in Albania, particularly in urban centers, and accounts for up to 8 percent o f people's food expenses in Tirana, and 6 percent in other urban locations. Nationwide, Albanians allocate 4.5 percent o f their budget to eating out. Tirana Other urban Rural Total Purchased 91.0 90.3 65.9 76.9 Non-purchased 1.o 3.7 31.2 18.6 Eatenout 8.0 6.0 2.9 4.5 Total 100.0 100.0 100.0 100.0 Non-Poor Poor Total Food 62.0 66.7 62.8 Nonfood 23.3 17.3 22.6 Utilities 12.5 13.5 12.3 Education 2.2 2.6 2.3 Total 100 100 100 Food consumption patterns reveal very similar dietary habits across locations. The only significant difference relates to the higher share allocated to flour and cereals in rural areas, partly offset by the higher consumption o f bread and pastries in urban areas. Households in urban areas spend more o f their 24 Itmustbe notedthat, the budget shares are obviously affected by the inability to include some ofthe components (likerent and health) inthe aggregate. 35 monthly food budget in eating out. Finally, expenditures for fruit in urban areas is double the rural shares, likely reflectingthe higher costs ofperishables inurban areas. Table 2. 37: Food exuenditure uatterns bv location Tirana Other urban Rural Total Breadandpastries 9.1 9.7 5.2 7.1 Cereals, flour andpasta 4.9 6.5 12.5 9.6 Meat 21.2 20.9 19.3 20.0 Fish 1.5 1.3 1.1 1.2 Milkand diary products 18.9 18.2 23.7 21.3 Oil 6.5 7.1 7.3 7.1 Fruits 6.4 5.8 2.9 4.3 Vegetables andlegumes 13.2 14.9 14.9 14.7 Conservedand frozenvegetables 1.9 1.9 2.0 1.9 Sugar and confectionery 3.4 3.5 3.9 3.7 Condiments and spices 0.6 0.5 0.5 0.5 Non alcoholic beverages,tea and coffee 4.3 3.5 3.6 3.6 Miscellaneous 0.2 0.2 0.3 0.3 Foodeaten outside 8.0 6.0 2.9 4.5 100.0 100.0 100.0 100.0 Mean food exuenditure uer cauita (new Leks) 5,290 5,015 4,782 4,9 10 8. Determinantsof poverty This section presentsthe results of a multivariate analysis ofthe determinants and correlates of poverty. A Probit model is estimated, usinga two-stage procedure that allows taking into account the LSMS sample design. In this way, stratification, weights and clusters are accounted for. The dependent variable is dichotomous and assumes value equal to one when per capita consumption lies below the poverty line. The problems o f endogeneity are not specifically examined, and to this extent the coefficients on the independentvariables have to be interpretedwith caution. Giventhe different characteristics of urban and rural poverty, two separate regressions for urban and rural areas are estimated. For lack of variability in urban areas, either because no urban household or all o f them have these characteristics, some variables are excluded form the urban The results of the multivariate analysis confirm much of what has been presented so far based on descriptive statistics. Education i s negatively and strongly correlated with poverty. Both in rural and urban areas, families with higher shares of members with secondary and higher education are less likely to be poor. As expected, the effect of higher education i s larger than that of secondary education. Poverty i s more likely to be found in large households: the probability o f a household being poor increases with household size. Families with a higher share of children are also more likely to be poor. Employment status also matters, and when a higher share of family members i s unemployed the household i s significantly more likely to be poor. Having access to telephone services (either fixed phone at home or a mobile phone) is negatively correlated with poverty. The effect i s larger in magnitude for mobilephones. 25 Readers interestedinthe full model and estimationresults are referredto the backgroundpaper for details. 36 While the above correlates are significant across geographical areas, urban and rural poverty are different and affected in different ways by other household characteristics. In urban areas, for instance, receiving Ndihme Ekonomike is positively associated with poverty, suggesting that the program at least partially reaches its intended beneficiaries in urban areas. The ownership o f household assets such as appliances and other durables (e.g., televisions, cars) is negatively correlated with urban poverty. Also, higher crowding - defined by number o f rooms per household member - in urban dwelling i s strongly correlated with poverty. Inrural areas, having a household member engaged independent work or being self-employed in a non- agricultural business reduces the probability o f being poor. Conversely, households receiving private transfers or rural pensions are more likely to be poor. Ownership o f agricultural assets i s negatively correlated with poverty. Other variables that are statistically significant and negatively correlated with poverty are size o f land, number o f cows, number o f small animals (poultry and rabbits), and agricultural durables. Having access to adequate sanitary facilities i s also negatively correlated with poverty. Among community characteristics considered, households living in a community with sewage are less likely to be income poor. In line with the descriptive analysis and with previous findings, after controlling for other features, female-headed households are equally likely o f being poor than their male-headed counterparts. Surprisingly, single female-headed households inrural areas are actually less likely to be poor. Although the descriptive analysis reveal strong spatial dimensions o f poverty, the multivariate analysis shows that, under equal conditions, households in the rural North-East are equally likely to be poor than their rural counterparts in the other regions. This is also true for the urban models. Finally, living in a newer, brick dwelling does not appear to be an indication o f lower poverty. In summary, the multivariate analysis confirms many o f the previous results and emphasizes policy- relevant areas for poverty reduction. Although causality should be inferred from the results only with caution, the importance o f factors such as improved education, particularly at post-primary levels, and non-farm employment opportunities, particularly in rural areas, are evident. In addition, the number o f highly significant correlates confirms the multidimensional nature o f poverty, and shows, for instance, that limited access to physical and human capital, both at the household and community level, goes hand- in-hand with income deprivation. 9. Concludingremarksand policy implications The poverty line presented in this chapter i s the first absolute poverty line based on nationally representative household level expenditure data ever estimated for Albania. Like any poverty line, it is based on a number o f assumptions and methodological choices, and as such i s open to debate and discussion. The welfare measure adopted for this purpose is based on the Cost o f Basic Needs (CBN) approach, i.e., a food poverty line based on the cost of the minimum calorie requirements augmented to make an allowance for non-food basic necessities. This resulted in the choice o f an absolute poverty line of 4,891 Leks per capita per month. Based on this measure, one-quarter o f the Albanian population, or close to 780,000 individuals, fall below the poverty line. Extreme poverty amounts to less than 5% o f the population, who are unable to meet basic food requirements. A large number o f individuals are clustered around the poverty lines. Increasing the poverty line by 10 percent increases the percentage o f poor individuals by 25 to 50 percent, depending on the poverty line used. This is true for negative shifts in the poverty line as well. In other words, small negative and positive shocks will produce more-than-proportional changes in poverty. Poverty appears to have a marked spatial and regional dimension in Albania. Rural areas and the Mountain region are found to be consistently poorer, according to all definitions o f poverty. From an international perspective, the 37 data presented confirm the picture o f Albania being one o f the poorest countries in Europe. The PPP poverty rates presented (2 and 4 dollars per day) are higher than those o f neighboring countries, and put Albania more or less in line with Caucasus countries like Georgia, and slightly ahead o f Azerbaijan. Despite growing concerns about widespread inequality in transition economies, inequality i s found to be moderate, with a consumption-based Gini index o f 0.28. From the point o f view o f the demographic profile o f poverty, the LSMS data show that larger, younger households are the most likely to be poor. Some demographic criteria can therefore be used to achieve a better targeting o f poverty and social programs. Households relying heavily on social assistance (and in rural areas also on pensions) are found to be disproportionately represented among the poor, suggesting that there is scope for improving the impact o f the current programs. These findings have obvious implications for the design o f policies and programs to reduce poverty. The importance o f adequate safety nets for this large number o f individuals emerges as a public policy priority. Also, in view of the relatively low depth and severity o f poverty, small, well-targeted transfers can potentially produce sizable results in terms o f poverty alleviation. Finally, the distribution o f poverty inAlbania warns against mechanicallyusingthe poverty line as a targetingthreshold for public programs, as doing so could exclude potential beneficiaries who suffer from the very deprivations that the programs are trying to address. Analysis o f the LSMS data shows that non-income dimensions o f deprivation compound income poverty, and living conditions disparities are more pronounced than are readily apparent by looking at the income dimension alone. Modernization o f the country has benefited urban areas, particularly Tirana, at a much faster pace relative to the rural areas, which have been left behind. In most cases, the coverage o f basic infrastructure services i s nearly universal in urban areas, but much less so in rural areas, though the urban areas face serious quality issues. Needless to say, the non-income dimensions o f poverty should receive no less attention than the income dimensions. As part o f the NSSED, the Government has pledged renewed effort towards making marked improvements to its public service infrastructure, particularly in relation to delivery to the poorest. It is likely that for many households social exclusion, the lack o f access to basic services and infrastructure, and to secondary and higher education, are factors hindering their ability to lift themselves above the poverty line. This further suggests that poverty-reduction policies would need to focus on rural and remote regions where the dismal condition o f basic services is most perceptible. Farm assets in Albania are fairly equally distributed, following a bottom-up approach to decollectivization inthe early 1990s.Giventhe highshare o fthe population engaged inagriculture, however, farm size is, in many cases, too small to be a viable commercial option or simply one that allows a family to escape poverty. Thus, even though land ownership is very widespread in Albania, the average farm size i s very small and access to land is not a sufficient condition by itself for escaping poverty. The current flow o f people out of agriculture and the rural sector i s therefore likely to continue in the medium and longer term, posing pressing problems for social and labor policies in urban as well as in rural areas. This calls for more effective labor policies to ensure more stable, non-farming jobs in rural areas. Since the agricultural sector provides the most substantial share o f income for the rural poor, addressing the factors hindering the development o f this sector have to be an essential component o f any policy mix aimed at reducing rural poverty. There exists a strong link between poverty and low educational levels. Primary school enrolment rates are lower among the poor, and lowest among the extreme poor. The pattern i s much more pronounced for enrolment rates in secondary school. The incidence o f poverty among adults holding a primary school diploma or less i s double the poverty rate among individuals with a vocational diploma or higher. More 38 than 80 percent o f the poor have not completed secondary school or hold a vocational school diploma. As is common in transition countries in Eastern Europe, education indicators in Albania are fairly good when compared to countries with similar levels of poverty. Achieving universal primary school enrolment is within the country's reach. Much more challenging, but perhaps bearing the most impact, is the need to increase secondary school enrolment and vocational training. The importance o f post-primary education cannot be overstressed, given its significant bearing on poverty reduction. Only 33 percent o f the individuals 15 years o f age and older reported holding a full time job, while another 21 percent reported having done some work on a part-time basis (31 percent in rural areas), bringingthe total of working people to 53.9 percent. About one halfofruralworkers only work part-time, and a large portion of them report working only a few hours a week. Irrespective of how employment is defined, the poor exhibit substantially higher unemployment rates when compared to the non-poor. The unemployment rate among the poor based on the standard ILO definition (at 14.2 percent) i s almost double the rate among the non-poor. Among the extreme poor the unemployment rate i s virtually triple. Using the relaxed definitions, unemployment rates among the extreme poor is above 30 percent. A significantly higher proportion of marginalworkers (those classified as discouraged, seasonal and laid off workers) are found among the poor. In sum, the modernization of the country that has in many respects been triggered by the transition reforms and the recent period of steady economic growth, is certainly contributing to improve overall living standards in Albania. The benefits, however, appear to be accruing mostly to urban areas, and particularly Tirana, while rural and more remote areas are being left behind. Several non-income dimensions of deprivation appear to be appalling given the overall level of development o f the country. Without direct, focused and well-targeted policy action and support from the Government, the possibility for many of the rural poor, and for the uneducated urban poor, to escape poverty by connecting to the economic growth process seems remote. 39 CHAPTER 111: SPATIAL DIMENSIONS OF POVERTY 1. Overview Albania has a total area of 28,750 square km, of which some 699,000 hectares is classified as agricultural land (24 percent of the total area), 1,027,000 hectares as forest (36 percent) and 446,000 hectares as pasture (15 percent). The remaining 25 percent is classified as other, which includes urban areas, about 135,000 hectares o f lakes and waterways, and unused rocky and mountain land. Albania can be divided into three major agro-ecological zones. The lowlands consist of the coastal plain along the Adriatic Sea and are the most productive agricultural areas of the country. Irrigation possibilities do exist and a wide range o f crops can be produced. The more extensive, hilly areas are mostly planted with fruit trees, olives, and grapes. The mountainous zone contains remote valleys and high mountain peaks, and has harsh winters. Crops can be grown inthe valleys, while the higher zones are mostly forest or livestock pasture. Overall, pasture quality i s poor and over-grazing is common. The country is very mountainous, with altitudes rangingfrom sea level along the Adriatic Sea inthe west to 2 751 m at Mt. Korabi in Diber District in the Northeast. Only 16 percent of the land lies below 100 meters, 55 percent falls between 100 meters and 1,000 meters and 29 percent i s above 1,000 meters. Accordingly, the land used for agriculture is often quite sloping, with only about 44 percent of the agricultural land having a slope of less than 5 percent, 37.5 percent between5 percent and 25 percent, and 18.5 percent above 25 percent. Albania can be divided into three major agro-ecological zones based on climate and topography. These zones are: Coast Zone. The lowland zone, which consists of the coastal plains along the Adriatic Sea, with altitudes ranging from sea level to 200 m. The southern sub-zone has a relatively dry Mediterranean climate with hot summers averaging 26 C and mild winters averaging 10 C, while the northern sub-zone is 2-3 C cooler with a greater incidence o f frosts and cold winds in winter. Rainfall ranges from 800-1,300 mm in the south and increases up to 2,000 mm in the north. About 80 percent of the total annual precipitation occurs from October to March, resulting in a need for irrigation for summer crop production. Overall, these conditions allow production o f a wide range of crops such as cereals, forages, vegetables and grapes, as well as citrus inthe most southerly part o f the zone. Central Zone. The transitional hilly (sub-mountainous) zone, ranging from about 100to 900 m, consisting primarily of hills stretching from north to south between the coastal plains and the mountains. This zone i s typically 2-4 C cooler with generally higher rainfall than in the coastal plains. Olives, grapes, temperate fruit trees, maize and wheat are grown, and there are extensive areas of low and shrub forest. The transitional zone is sometimes considered to include the relatively extensive and agriculturally significant Korce basin in south-east Albania, located at over 800 m in the driest part of the country with an average annual precipitationo f 790 mm and an annual averagetemperature of 10.5 C. Crops such as wheat, beans, sugar beets and potatoes are important in this area, and temperate fruit trees such as apples and plums may also be grown. Mountainous Zone. The mountainous zone consists o f intermountain valleys and high mountain peaks, with mild summers and cold winters with heavy frosts and significant snowfall. Summer temperatures rarely exceed 25 C, while minimum temperatures in the winter may fall as low as -20 C. The annual precipitation averages about 900 mm in the south to 1,500 mm in the north, and up to 3,000 mm high in the northern mountains. Crops such as maize, forages, summer vegetables and winter wheat may be grown in the valleys, with barley and potatoes at higheraltitudes. Temperate fruit trees such as apples and plums may also be grown. At highaltitudes, the zone consists mostly o f forests and pasture for livestock. 40 2. Poverty in the regions The LSMS sample was drawn to be representative for the three agro-climatic regions, since they reflect very different socio-economic conditions and potential. The analysis of the data reveals that households in the most remote districts in the Mountain region in the north and northeast of the country fare worst in terms o f poverty (Table 3.1). Almost half of resident of this area are poor, and more than a fifth in this group live in extreme poverty. Also, the depth o f poverty in this area is much more pronounced than in any of the other regions, with a poverty gap index of over 11 percent. Average consumption in those mountainous regions i s two-thirds o f consumption levels in Tirana, and about 20-30 percent lower than the rest of the country. Poverty rates and depth in all other regions are around or below the national averages. Mean consumption per capita(Leks) 9,054 8,424 7,499 6,173 7,801 Gini 0.3 1 0.29 0.28 0.28 0.29 A look at the income cumulative density functions by region confirms that the mountain region is the poorest, regardless of the poverty line used. The Central region fares better than the Mountain region but worse than either Tirana or the Coastal region. Stochastic dominance is maintained by Tirana over the Coastal region over a realistic range ofthe poverty line. Fig3.1:Income cumulative density functions by region poverty line 2000 I 4000 I real per capita6000 I BOO0 I 10000 I consumption 41 Inequality levels are moderate and stable across regions and the Gini coefficient inthe regions is between 0.28 and 0.29, with the exception o f Tirana, where inequality i s higher and the Gini is 0.3 1. 3. RuralPoverty As has already been shown in the previous chapter, 66 percent o f the poorest quintile and 61 percent o f the second quintile are rural households. In all, 34.8 percent o f the rural population is poor (poor 29.6 percent and extreme poor 5.2 percent). A significant amount o f consumption (more than 30 percent) in rural areas derives from own-production, which indicates the importance of small-scale agricultural production for food security and maintenance o f minimum consumption levels. Considering that a lot o f the labor earnings and non-farm agricultural incomes are currently based on potentially unstable economic sectors (i.e., construction), this becomes even more relevant. The LSMS data reveal that the rural areas depend largely on agriculture as their main source o f income (44 percent), followed by employment (19 percent), public transfers (17.8 percent), private transfers (12.6 percent) and non-agricultural production (5.9 percent). Comparedto urban areas, the share o f all non-farm income sources i s lower in rural areas. Urban households depend on public and private transfers for 41.5 percent o f their income, while in rural areas only 30.4 percent o f income comes from these sources. Ifthe rural income sources are split up in consumption quintiles, it becomes evident that the difference between a poor rural household and a non-poor rural household lies mainly in its level o f private transfers. The highest quintile relies to a quarter o f its income on private transfers, while less than ten percent o f the income comes from private transfers in the poorest quintile.26The share o f agriculture in income declines from over 50 percent in the lowest quintile to 33.8 percent in the highest quintile. However, in value terms the highest quintile receives more than double o f income from agriculture per months than the poorest quintile. Table 3.2 shows a breakdown o f income share by urbanhural as well as by rural quintiles. Income Source Urban Rural Rural Quintiles 1 2 3 4 5 Employment 46.4 19.0 16.6 21.2 23.1 17.6 16.6 Agriculture 1.9 44.3 52.4 45.6 42.7 42.6 33.8 Animal Products 0.2 8.7 8.3 7.9 7.9 7.5 12.6 Livestock 0.1 7.9 11.7 7.4 9.3 5.9 3.5 Crops 1.6 27.7 32.4 30.3 25.5 29.2 17.7 NonAg. Production 9.0 5.9 2.6 8.5 6.0 7.8 4.9 Private Transfers (total) 16.5 12.6 8.8 8.5 12.3 13.1 24.2 Public Transfers (total) 25.6 17.8 19.5 16.1 15.3 17.8 20.2 Other Income I 0.7 I 0.4 I 0.1 I 0.0 I 0.6 I 1.1 I 0.3 I On average, income generated in rural areas is only two thirds o f the income that the urbanpopulation has available per month. As expected, income from business and employment are three times higher in urban areas. Private transfers matter more in rural areas. While the amount o f monthly private transfers per 26This observationholds also true for urban areas where the highest quintilerelies on private transfers for around40 percent of their income (urban quintiles are not shown inthe Table). 42 quintile in urban areas does not differ too much, the poorest rural quintile receives around a fifth of the transfers o f the richest quintile. Income from agriculture is also significantly larger in the higher consumption quintiles, mainly deriving from crop and animal product sales. Rural households in the highest quintile also differ significantly from the rest in terms o f their non-farm income availability (ten times larger between quintile 1and quintile 5, but still more than three times larger between quintile 4 and quintile 5). hncome Source Urban Rural RuralOuintiles I 1 1 2 1 3 1 4 1 5 1 Employment 17,904 6,574 4,232 7,260 7,866 6,935 7,242 Agriculture 717 12,172 7,262 9,858 13,395 16,017 17,089 Animal Products 70 1,501 860 1,523 1,348 1,946 2,124 Livestock 68 1,289 1,365 1,360 1,343 1,290 1,017 lCrops I 580 I 9,382 I 5,037 I 6,976 I 10,704 112,781I13,948 I Non-Agricultural. Production 14,922 2,952 633 1,305 2,372 2,785 9,573 PrivateTransfers (total) 2,527 4,027 1,706 2,565 4,386 4,835 8,206 Public Transfers (total) 4,701 3,006 2,630 2,901 2,704 3,252 3,800 L r - i n c o m e I 526 1 105 I 22 I 5 I 241 I 73 I 243 I Total Income 41,298 I28,833 I 22,636 I 29,644 I 38,175 133,732 I70,480 Almost 90 percent of the rural population own and use agricultural land. Poorer households own and use the land to a slightly larger percentage (Table 3.4). Rent o f agricultural land is undertaken rarely. Considering that the average land size owned by a household i s rather small, on an average around 0.72 hectares, this indicates that at this stage there i s very little room for enlarging agricultural operations. Many people rely on agricultural production for subsistence and do not seem to be ready to give this up. The average land size owned by a household is 0.76 hectares for non-poor households and 0.6 hectares for poor households. and Rent Agricultural Land Table 3.5 provides a further breakdown of the land holding sizes in Albania. More than 50 percent o f all households in Albania own less than 0.5 hectares o f land. In rural areas, more than 60 percent own less than 0.8 hectares of land. These are very small average land sizes and lower than the average in other Eastern European countries. Only around 3 percent of all households own land areas larger than 2.5 hectares. Three-fourth of the households living in mountainous areas have less than 0.5 hectares at hand. 43 They also, however, indicate that very little structural reform has taken place duringthe last ten years and holdings have become rather smaller than larger during this period. As the lower section of Table 3.5 indicates, land size by itself does not determine whether a household is poor or non-poor. However, one has to consider that these figures relate to the total sample, thus also including urban households with small land holdings. The data show that households with more than 1 hectares o f land tend to be less poor. Ihv ouintile > 2.5 hectares 3.78 I 1.17 I 3.45 1.49 I 0.52 3.17 III II 1 2 3 1 4 5 Less than 0.2 hectare4 26.7 27.8 25.9 27.5 26.5 0.2 -0.48 hectares 25.81 16.93 18.82 13.82 12.86 0.48-0.8 hectares II 14.72 IIII 15.78 IIII1 19.1 1II 18.46 IIIII 11.54 IIIII 0.8 1.28 hectares - 20.19 19.47 18.74 19.92 20.42 1.28 -2.5 hectares 11.2 18.74 12.39 17.05 18.99 I >2.5 hectares I 1.36 I 1.3 I 5.08 I 3.23 I 6.16 I 4. Ruraland Mountain Areas -Pocketsof SeverePoverty Albania's rural areas, particularly in the Mountain region in the north, are amongst the poorest areas of Albania. Compoundingthe problems for those Albanians who live inthese areas is the harsh climate and rugged, mountainous geography. The majority of residents make their living through subsistence farming, which also provides a significant source of the rural families' daily food intake. However, most do not have enough land to provide adequate sustenance or variety, and most rely on a small set of staple crops and products for their meals. "...Inmost cases, thefood of my children is potatoes soup. I bake the bread myself because it is cheaper. My children have by now got used to thefact that they can go to sleep withoutfood, while my great pain is to see them not askfor bread even when they are hungry. When I got married I had no idea my life would be like this..." (a woman in the Bushat commune)27 `'T h i s and other quotes in the section are taken from interviews with families in the communes of Bushat and Postribe outside of Shkodra, in the remote, barren highlands of Albania, situated on the lake that bears its name. Both communes have 11 villages each, and maintain a total populationof about twenty-threethousand inhabitants. The results and conclusions drawn from these interviews are by no means representative, either of the village or regionfrom where they are drawn or of Albania as a whole. 44 As in other areas o f the country, one o f the most obvious manifestations o f poverty in the rural areas is a lack o f steady employment; in general, people inthese areas desire work, but are unable to find enough to provide any sort o f comfortable lifestyle. "There are nojobs in Bushat. You cannot even clean housesfor the others because people are not that rich to afordpaying for a house cleaning lady. In a nutshell, we see nofuture. Wedo not know what will happen to us, and our children. The only faith we have left is that in God." (a woman fiom Bushat commune) "Iwork all day long at home, and till that little land we have where we haveplantedfew vegetables enough to ensure our survival. There is no mention ofjobs at all. There are nojobs, and we have nowhere to seek. )'(amother of 6 children, Bushat commune) A vast majority o f buildings in rural areas are extremely old, and as such many o f the difficulties associated with maintaining a healthy living situation are related to the upkeep o f crumbling buildings. Many structures are abandoned or dilapidated, and stand unused. Rural residents also face significant difficulties with the administration and maintenance o f basic infrastructure and government services. The country's roadways, particularly in the mountainous areas, are in a state o f severe disrepair. This often results in entire areas o f Albania being cut o f f from rest o f the country, including from government administration and from basic services. Sanitation facilities are very bad, and most families do not have sanitary toilet facilities. In the absence o f running water and appropriate sewage systems, toilets are usually located outside the house. "...We are six members in this house. ...All the waste water pipes have been blocked. The water stinks. The WC has not functioned in two years, and we have to go out in the openfields. ...One thing is clear though: we have no money to buy our food, while the commune says that we have to look after the building ourselves since now we are the owners. We do not know what this means,..'' (a mother of 4 children, Bushat commune). Electricity i s another case in point. While the coverage o f the network i s virtually universal, the delivery o f the service is highly unreliable, with the situation being by far worse in rural communities. The vast majority o f people in the rural Mountain areas have only intermittent access to electricity, and most o f time, even when the electricity i s on, the voltage i s far too low to make any use o f it. Inmany cases, particularly in the northern regions, the lack o f any administrative presence has also been accompanied by the return to prominence o f j s , which are traditional administrative structures built around extended family connections. In many villages, this traditional approach has completely replaced the state as the primary administlative mechanism for local citizens. "The quality of the social services in the village is closely linked with the decentralization level of the local government.Actually, these services are notfunctioning well and we do not have any competenceto make improvements. The opinions of the local authorities are not considered relating the quality of servicesprovided by the teachers,healthpersonnel or other people involved in such services. Even in the cases when explicitly the local authorities express negative opinions about those they are totally neglected. In fact we are closer and more accountable to the communities compared to diferent ministriesfiom which they are appointed and to which they give response. Whenthe community members are not satisfied with the quality of services provided they come to us because they think that we can change the situation and they don't and can't go to the ministry to complain. '' (a member of the commune council, Bushat) 45 The returnto prominence of various traditional structures has also revitalizedblood feuds. Inmany areas in the north, blood feuds have decimated entire families, and left remaining members confined to their households in fear, effectively preventing them from attending school, purchasing goods, and securing employment. "...I comeJi.om Tropojabecauseof the bloodfeud. My husbandgot killed twoyears ago,andmy have sister's husbandfive years ago. Ihave two children, but Ialso have the children of my sister since she re- married, We are 10 members who live under the same roofl and have no land and no social assistance. My 8year old son and my 6year old daughter do not go to school since we are shut in the housefor fear of vendetta." (awomanfiom Bushat commune) "...My husband was killedfor vendettdbloodfeud 8 years ago. I live with two elderly parents andfour children. My 2 daughters have left because they got married. When the land plots were being allocated, we also benefited according to the legislation, but then we got entangled in disputes with the ex-land owners and my husbandwas killed. None of us works, and we have no land and no other income to make a living. Thecommunehelped me get ajob as a cleaning lady in the school, but a decision of the Ministry turned my life upside down since it said that the schools would be cleaned by the pupils themselves, and therefore there would be no needfor cleaning ladies. Actually, this is not possible because the children are very young, and often have no clothes, and lack nutrition, and the village has no water. The school principal kept me, and eachpupil pays money for me every month, which is enough to buy me bread. I feel like a beggar, and ashamed?om the people because they have themselves no money. This is not a way to live,just receiving aid. We are not even entitled to Ndihme Ekonomike because on paper we are land owners..." (a mother of 4 children, Bushat commune) The educational situation is in dire straits inthe rural areas. Teachers are leaving these outlying areas to move to urban centers, and the remaining students often face overcrowding problems because of too few teachers. In some cases, the situation i s made even worse because rural schools are sparsely located, and children have to walk long distances. In inclement weather conditions, which are common in the north and the mountainous regions, this ultimately results in poor school attendance. Lack o f fbnds for education leaves many schools without even basic supplies for children. In some cases the physical structure of the school itself i s in poor shape. Even the local school administrators admit that the education system faces daunting challenges. "The commune has seven 8th grade schools, one secondary school, and four primary schools. A classroom has on average 30-40 students, but there are also cases when this number reaches up to 45-50. 4 schools were reconstructed with state funds; one school was re-constructed with funds of an international organization. The rest of the schools are in very bad shape, there arefew benches, no glass on the windows, and the doors are broken." (EducationInspector, Bushat commune) The heath situation is similarly desperate. Inthe commune o f Bushat, out of 11villages of the commune, only 4 have health centers, and there i s only one maternity home, with 4 midwives and no gynecologist. Before 1990, there used to be health centers inall the villages, while now they have either been destroyed or are usedas housing by the homeless. Rural families also report having to pay a "gift" to the doctor to ensure decent health services. Though basic care is supposed to be free in Albania, incidents o f doctors or health staff requiring an informal payment are quite common. "The doctors look at your hands. If you do not pay, they won 't even look at you. It is now becoming almost legal to give the doctor Lek 200for a check-up, and the nurse Lek 100for an injection. Then, there 46 are the medicines. You can only imagine what our pension can buy us." (male, around 60, Bushat commune) The conditions and problems illustrated above have created in many rural residents a sense o f helplessness. Faced with living in brutal conditions, many have chosen to leave in favor of the opportunities presented by urban and peri-urban living. Though most do not necessarily know what to expect from their migration, the risk is worth it, given the alternative o f staying and struggling as they have. "My husband is dead.Ihave three children. My daughter suffers fiom a chronic disease.Iam an invalid, and my fingers have been cut. I led a dog's life there, since I had no land, received no agriculture produce since we lived in the mountain, and it would be futile to plant anything there. I received an invalid's pension of Lek 6,000per month. We had a house, but the rain thatfell would all come inside, and we could not afford to make any repair whatsoever. Our village was totally isolated. Wehad to walk for 4 hours in order to reach the closest health center. There was no schoolfor our children. Thepeople of the village started leaving by the dozen. Finally, we also decided to leave since we could not live alone on the top of the mountain. Here Ipay a rent of Lek 4,000 per month. We buy food with Lek 2,000 per month, and there is nobody to help us, because our neighbors can hardly sustain themselves. Wereceived some humanitarian aidfor some time, but that is no longer the case. My sick daughter has to undergo surgery, but Icannotpayfor it. I have to waitfor my two other children to grow a bit more, so that they can leave and may befind ajob. It was good that we left the village anyway, since there even if one died there would be no one to bury him. But even here it is not easy... (a mother of three who recently migrated to the Bushat commune) 5. Povertyin peri-urbanareas The peri-urban areas lie on the outskirts o f both urban centers, and are often situated on or near sources o f fresh water (rivers and streams, wells). They are usually composed largely o f informal settlements and structures, and often lack connections to basic utilities (water, electricity, waste disposal). Most residents of the peri-urban areas are recent arrivals from other areas in Albania who have migrated to the urban centers with the hope o f employment and a better life for themselves and their families. As such, the population o f the peri-urban areas i s both fluid and difficult to gauge. Often times these residents migrated from other areas o f Albania en masse and ended up in communities that mirror the communities intheir previous location, and are comprisedprimarilyo fextended families. *' " ...We are 7families that have come to Tiranafiom the same neighborhood in the village. We countfor more than 50people, and we even are relatives. Men discussed about leaving the village several times, but were not really decided since they had hopes the situation in the village would change. There was also talk about the construction of a water supply, a road and a health center. Only later did we understand that nobody was going to do anything about us, and the more we stayed in the village the '*Our knowledge of poverty and living inperi-urban areas is extremely limited as no adequate quantitative study is available to look at it in a scientific manner. The LSMS data is not representativeat the peri-urban level, and thus does not allow for meaningful statistical analysis of this phenomenon. However, in view of the rapid demographic change the country is experiencing and the increasing burden rural migrant place on overstretched municipal budgets of urban centers, it is important to understand and study peri-urban areas in more depth. This section is based on interviews o f families in the peri-urban areas of Tirana and Dunes, and the quotes provide some glimpses of poverty. Note, however, that these quotes are not meant to be representativeof all the peri-urban areas. 47 worse it would be. We sold our cattle, and lost the money in thepyramid schemes.And with that we also lost hope. Like us, in groups, everybody was leaving, and most of us would settle in Tirana. We had some source of security since we had 15 men who knew how to build houses, and that at leastprovidedfor our shelter. We also had received newsfiom those who had settled in Tirana earlier that although they lived in bad conditions, it could not be compared to the life we had lived back in our village. That is why we came, andjoined our relatives and acquaintanceshere..,)'(an old manfiom Kamza, outside of Tirana) Some have found hope in their new lives, and have managedto succeed in their move. But others have found only the same kinds of hardshipsthat they faced elsewhere in Albania.As inthe ruralareas, one of the most obvious manifestations of poverty is a lack of employment opportunities. Most peri-urban dwellers rank unemployment as the most significant or one of the most significant problems that their families face. Inmost cases at least one family member is able to find employment, but this employment i s oftentemporary and/or intermittent.Almost no families have a memberwho i s employedregularlyand full-time. "We have come+om Puka to Tirana. We came with the hopefor a better life, but we are disappointed. We havefound nojobs, although my husband and my son go to the city everyday on the lookfor work. Our house construction is not yet complete since the moneyfiom the sale of our previous house and the savings were not enough, and what we had we have already spent. We buy the waterfor washing, cooking and drinking. There is no waster water system and no toilets. But we cannot complain, since nobody forced us to come here. When our relatives in Puka ask us how we arefaring, we basically shut up, and tell them we have everything... (a manfiom Kamza) " " ...Ihave been living herefor sevenyears, and together with my son we work at whatever we can get hold of to maintain ourfamily, composedof II members. My wife and my daughter-in-law look after the house since none of them works. Weset out early everyday,and search the construction sites in the city tofind a job, because construction is our profession. The job that lasted the longest was construction in a condominiumfor 4 months. There is no mention of a contract since they always tell us that they do not need labor, but they employ us nonetheless since they takepity on us. This is the samefor everybody. (a manfiom Kamza) After employment, food problems would likely rank second amongst peri-urbanresidents. Most families have significant difficulty in feeding their families. Almost all already purchase the most inexpensive food items available, and most pay little or no attention to providing any variety of diet for themselves and family. Most purchase food on credit without interest, which is usually done via an informal list system recordedby the shop owner. Most houses in the peri-urbanareas are quite new, and often homemade.In some cases these houses are solid, concrete structures; in many others, however, the new residences are essentially shacks, put together with scavenged materials and wood. Most buildings in the peri-urbanareas around Tirana are illegalconstructions. "...This building used to be a stable. Now there are 12families that live here. We have used whatever came at hand to divide theplace: boards,plastic, cartoons, and steel sheets. Instead of the window glass we useplastic sheet. There is no water, no WC,and no sewage system. Wefill the water with big bottles, whilefor the WC we use some small barracks, which we have built ourselves on top of the channel that goes by here. The channel is really in the open and all the households' discharge their wastewater there. We live only I O meters away@om the channel, and our children literally play there because they do not know where to stay. Wehave nopermit. And why wouldwe need that? Thecows which used to stay in this building before us coming here had better conditions than we do; they, had water, sewage and electricity..."(a manfiom Kamza commune) 48 Along with the varied quality o f housing, however, one must also take into account that many Albanian families live in close quarters with a large number o f their immediate and extended relatives. It i s also important to remember that poor households rarely consist o f more than a couple o f rooms. It i s not uncommon to find households o f upwards o f 8-10 people. A huge problem in peri-urban areas is the lack o f basic services. Chief among these is the government's inability to provided clean running water to many household^.^^ Many families have to get their water from a nearby well, or from a source o f running water not located near their residence. Because o f a corresponding lack o f waste removal systems in these peri-urban areas, many families obtain their water from sources near where they dispose o f their waste (both human waste and trash). Many o f the same problems hold for the delivery o f basic electrical services as well. Many residences are connected to electric wires, but most indicate that service is intermittent (at best), and that the power provided i s insufficient to run basic household appliances. Families moving to urban areas have only served to overcrowd the already insufficient school systems. Class size has increased dramatically in the peri-urban areas, and many schools have started running on multiple rotations (one set o f classes half o f the day, a completely different set the other half) in an effort to relieve stress on teachers and classrooms. This has resulted in many students coming to school for only a few hours each day. Access to health services in peri-urban areas is very limited, and in the peri-urban areas o f Tirana there i s still no network o f health centers that would be in line with the number o f inhabitants. 6. Povertyand InequalityMaps Given the small sample size, it is not possible from the LSMS to analyze the spatial dimension o f poverty at finer levels o f geographical disaggregation. An attempt to obtain rates at prefecture level reveals the lack o f precision o f the estimates, particularly for the smallest prefectures. Although the ranking o f prefectures using the LSMS data seems plausible, particularly along the tails, accurate data will be available only after completion o f a detailed and comprehensive spatial analysis that estimates full poverty and inequality maps at the district and municipality levels. Such a poverty and inequality mapping analysis i s currently being carried out as part o f a collaborative effort involving MOLSA, INSTAT and The World Bank. Based on the methodology described in Elbers, Lanjouw and Lanjouw (2002), census and survey information are being combined to produce finely disaggregated maps o f poverty and inequality at the levels o f the district, municipality and prefectures. This work is ongoing, and the final findings are being discussed with the Government o f Albania. This section includes some o f the preliminary findings inthe forms o f maps andtables3' Inorder to producepoverty and inequality maps, income or consumption expenditure data is needed such that it is representative at low levels o f aggregation or is o f sufficient size at low levels o f aggregation to yield statistically reliable estimates. Household budget surveys or Living Standard Measurement surveys covering income and consumption usually used to calculate distributional measures are rarely o f such a sufficient size, whereas census or other large sample surveys sufficiently large enough to allow disaggregation have little or no information regarding monetary variables. This makes the case for combining the above using information on monetary variables from one set o f surveys (LSMS) and disaggregated units from another set o f surveys (Census). The basic idea is to estimate a linear regression 29Some progresshas been made inrecent years to improve the basic water infrastructure inthe peri-urban areas of Tirana (layingpipesand connectingmainpipesto houses), but it has notyet become filly functional. 30The poverty mapping, usedas a backgroundpaper for this section, was financedby the Albanian Ministry of Labor and Social Affairs. Supportand assistance from INSTAT andMOLSS is gratehlly acknowledged. 49 model with local variance components using the information from the smaller and richer LSMS data sample and including some aggregate information from the Population and Housing Census or other sources available for all the statistical units in the sample @e., from the General Census of Agricultural Holdings). The estimated distribution o f the dependent variable in the regression model (monetary variable) i s then used to generate the distribution for any sub-population in the census conditional to the sub-population's observed characteristics. From the estimated distribution of the monetary variable in the census data set or in any o f its sub-populations, estimates are made of a set of poverty measuresbased on the Foster-Green-Thorbecke indexes, the Sen index and an absolute poverty line calculated using the information contained in the rich sample survey, as well as a set of inequality measures based on the Gini coefficient, the Gini coefficient of the poor and two general entropy (GE) measures. These indices are calculated for the whole o f Albania and disaggregated at the level of the four strata used in sampling the LSMS, the six strata for which the linear regression models were estimated, 12 Prefectures, 36 Districts, 384 Communes, and the 11Mini-municipalities comprising the city of Tirana. Table 3.6 reports poverty and inequality measures and bootstrapping errors for the whole of Albania, and disaggregated by the four regions and by ruralhrban type for the Coastal and Central region^.^' The region o f Tirana i s better off interms o f per capita consumptionand percentage of individuals below the poverty line (head count) compared to the others, while the Mountain area has the highest proportion of poor individuals. More than one-third of the population i s poor in the rural areas in both the Coastal and Central regions. According to the four poverty indices considered the Mountain region is the worst off. The region of Tirana shows higher inequality inthe distribution of per capita consumption (according to the Gini coefficient index and the two General Entropy indices used) relative to the other regions. Table3. 6:Poverty and inequality indices (33) FGT(1) FGT(2)I Gini Gini-poor Sen 1 Con ALBANIA 6.96 2.48 II 12.38 15.29 14.28 II GE(0) II GE(1) 1I 15.05 7,569.67 COAST II HeadCount IIII 28.60 I 1.28 0.44 0.19 0.52 0.27 10.40 0.53 0.77 120.21 26.64 6.48 III1 2.32 I129.54 II 12.40 14.83 16.36 17.68 8,148.48I 1.94 0.65 0.28 I131.57 1 1.15 0.40 10.58 1.24 1.88 249.18 CENTRAL I 29.49 7.00 2.43 I 11.94 15.36 12.16 12.43 7,177.76 2.32 0.76 0.33 0.54 0.46 10.71 0.51 0.52 222.95 I MOUNTAIN 40.85 10.98 4.20 I127.35 II 127.40 13.56 19.43 12.25 12.41 6,181.78 1.60 IIIII 0.63 IIIII 0.31 I 0.55 I 0.34 10.67III11II0.51 IIIIIII0.52 IIIIII1 120.69 TIRANA 18.01 4.11 1.42 29.35 11.88 2.80 14.18 14.54 8,981.39 1.09 0.38 0.17 0.63 0.52 0.30 0.65 0.70 140.85 COASTurban 15.63 3.80 1.40 30.25 12.81 2.54 15.34 15.19 9,935.96 1.84 0.61 0.28 0.94 0.78 0.47 1.04 1.00 467.70 COASTrural 34.84 8.47 3.00 28.87 12.23 6.89 13.97 16.54 6,816.09 3.22 1.09 0.46 1.74 0.48 1.09 1.94 3.87 283.67 CENTRALurban 19.48 4.08 1.29 26.34 10.59 2.79 11.28 11.42 8,168.94 1.56 0.48 0.20 0.64 0.50 0.39 0.56 0.58 163.15 CENTRALrural 34.41 8.43 2.99 27.02 12.25 6.80 11.87 12.23 6,689.88 3.45 1.13 0.48 0.86 0.53 1.14 0.80 0.85 327.31 Table 3.7 reports the poverty and inequality measures calculated at the Prefecture level. The Prefecture of Vlore has the highest per capita consumption andthe lowest percentage o f poor people (16.57%), whereas 31 For any given location, the means constitute the point estimates, while the standard deviations are the bootstrappingstandard errors ofthese estimates. 50 according to the Gini coefficient consumption is very concentrated (33.61%). On the other hand, the Prefecture of Diber is the worst off, with per capita consumptionof only 6,211 Leks per month, and the highest percentageof poor individuals (42.1%). 1.38 0.48 0.22 0.96 0.40 0.42 1.oo 1.14 259.03 4: ELBASAN 31.84 7.54 2.61 26.60 11.89 5.91 11.48 11.74 6.852.57 1.19 0.39 0.17 0.46 0.3 1 0.33 0.46 0.52 134.69 12: VLORE 18.26 4.14 1.42 33.52 11.72 2.82 18.63 20.93 9,817.49 1.58 0.47 0.19 2.12 0.45 0.37 2.55 4.42 502.02 Tables 3.8 and 3.9 report poverty and inequality measures disaggregatedat District andMini-municipality (for the City of Tirana) level. Table 3. 8: Poverty and inequality indices by DISTRICT (?A) 0.77 0.31 0.57 0.47 0.70 0.50 0.51 I207.01 2: BULCUIZE 1 59.46 I 18.80 II8.02 II26.48 II15.63 II18.76 II 11.47 II11.52 14.872.82 2.68 1.41 0.78 0.65 0.54 1.74 0.58 0.58 157.85 3: DELVINE 14.50 3.17 1.05 42.84 11.19 2.05 30.83 35.36 13,400.59 2.34 0.67 0.27 6.16 0.79 0.49 9.40 12.44 2,113.08 4: DEVOLL 27.89 6.09 1.96 26.21 10.70 4.57 11.05 11.39 7,203.81 4.16 1.26 0.49 0.71 0.65 1.16 0.61 0.64 369.79 51 5: DIBER 38.14 10.07 3.79 27.98 13.25 8.40 12.74 12.95 6,472.22 2.01 0.79 0.38 0.70 0.48 0.82 0.66 0.66 161.23 6: DUWS 21.78 5.38 1.96 31.60 12.71 3.80 16.50 16.76 8.996.73 io: GJIROKASTER 18.33 3.93 1.27 28.55 10.85 2.65 13.32 13.82 8,783.22 1.94 0.55 0.22 0.91 0.59 0.44 0.90 1.10 323.58 11: HAS 47.52 13.56 5.37 27.97 14.16 12.30 12.73 13.07 5,770.89 3.13 1.38 0.71 1.02 0.68 1.58 0.95 1.03 228.69 12: KAVAJE 27.98 6.76 2.40 29.52 12.22 5.10 14.18 14.55 7.689.25 52 53 Figure 3.2: Poverty Headcount by Prefecture Figure 3.3: Per Capita Consumption by District Ratio of the number of poor persons to the total populatio (in%) Per Capita ConsumpBon (in lek per month) n 1 5 - 2 5 04500-6500 - 3 5 -255 = 6500 ~35 -- 7500 4 = 7500 8500 8500 - 13500 Figure 3.4: Poverty Headcount by Communesand Municipality The data presented above intables can also be depicted inthe form of poverty maps, three o f which are shown here. The map on the top-left (Figure 3.2) shows the poverty headcount by prefecture, while the map above (Figure 3.3) shows the distribution of per capita consumption by district. The map on the left (Figure 3.4) shows poverty headcount by communes and municipality. This analysis is particularly informative for the purposes of targeting, since the location and concentration o f the poor become immediately obvious from an examination of the tables and the associatedmaps. When finalized, the complete version of the poverty mapping will also contain inequality maps showing pockets of low and high inequality of income as measured by the two Gini coefficients and the two entropy measures. 54 CHAPTERIV:LABORMARKETDEVELOPMENTSAND POVERTY 1. Overview Macroeconomic reforms and restructuring of the early nineties diminishedlabor demand in Albania, and like in many other transitional economies, the country experienced falling labor force participation(LFP) rates. Mass privatization of state-run enterprises and increased competition for a reduced number ofjobs also pressed many people to withdraw from the labor force precipitately. According to official statistics, the national LFP rate was the lowest in the year 2000, when only two-thirds of the working age population inthe country was active inthe labor market. Male LFP rates inAlbania have beenhistorically higher than female Figure 4.1: Albania labor force participation rates, 1993-2001 participation rates, and the transition years affected more ~ female workers than male workers. In the year 1998, for instance, in the aftermath of the pyramid schemes crisis, the male LFP rate was 85 percent, whereas just 55 percent o f 1993 1994 1995 1996 1997 1998 1999 2000 2001 1 working age women participated in the labor Imarket. Source: INSTAT,Albania Nevertheless, Albania has one of the highest LFP rates among the Central and Eastern European countries. In the year 2001, just the Czech Republic and Slovakia had higher participation rates than Albania. The gender gap inlabor participation inAlbania has been relatively large and is comparable with the labor gap inBosnia and Herzegovina, Macedonia, Romania, Croatia andthe FRof Yugoslavia. High levels of labor participation in Albania, like in other transition economies, are a legacy of the full employment policy of the socialist period, especially for female workers who had many economic and social incentives to stay active in the labor force. Another force supporting high LFP rates in Albania is the structure of the economy, which being largely agricultural, provides employment, even if mostly part- time, to a considerable segment of the population. Agricultural employment in Albania increased even more in the nineties, when the government implemented land reforms creating a large number of new landholders. The LSMS data shows that 60 percent of the working age population in Albania is active in the labor market, and the LFP rate is 71 percent for men and 50 percent for women.32The data show that LFP rate among individuals with more than secondary education was 74 percent, 7 percent higher than among individuals with secondary education, and 18 percent higher than among individuals with less than secondary education. 32The discrepancy betweenthe official statistics and survey LFP rates might be due to different age groups usedfor calculation. The INSTAT currently uses ages 15 to 59 for men and 15 to 54 for women as the official working age, while this study uses 14 years of age and above for both men and women. See the labor market definitions box for details. 55 Source: Albania LSMS, 2002 Men comprised 56 percent of the labor force and, reflectingthe demographics o f the country, one-third o f the labor force consisted of workers of ages between 15 and 30 years old. The composition o f the labor force shows that 62 percent of those active inthe labor force had less than secondary education, andjust 10 percent had post-secondary education. The labor force is mostly rural (65 percent), which mirrors the economic structure of the country. Rural areas have higher LFP rates (68 percent) compared with urban areas (48 percent), and among four geographical regions, Tirana has the lowest LFP rate (48 percent). loutofthe LaborForce 37.5 I 36.9 I 51.8 I 51.6 1 31.6 I 40.3 Source: Albania LSMS, 2002 In2002, one-third ofthe working age populationwas out ofthe labor force. A majority ofthose out ofthe labor force resided in urban areas (58 percent) and half were under 40 years of age. Housewives (25 percent), retirees33(21 percent), and students (19 percent) together made up the core of those out of the labor force. A relatively large group o f individuals reported being handicapped or disabled (10 percent) and there was a sizeable share of discouraged workers (10 percent) among those out of the labor force. These shares of retirees and housewives were significantly lower than those in 1996,34suggesting that more of the older people were staying active, and more housewives were probably returningto the labor force. On the other hand, an increased share of students in this group could indicate that more of the younger people were delaying employment for further education. 33The current retirementage in Albania is 54 for women and 59 for men, but it is set to increase to 59 for women and 64 for men.To insure comparabilitywith other CEE countries as well as with previous labor studies inAlbania, populationof ages 15 to 64 years old is considered as working age populationfor this study. 34Accordingto Rashid, M. (200l), the 1996 out-of-labor-force segment consisted mostly of retirees (41 percent), housewives(38 percent), students (12 percent), and discouragedworkers (9 percent). Note that the 1996 survey did not cover Tirana. 56 BOX 4.1: LABORMARKET DEFINITIONS The definitions below are basedon the 2002 AlbaniaLSMS questionnaire and/or follow commonlyused ILO labormarketdefinitions: Working age population: population of ages 14 years old and above is considered as working age populationfor this study. (Note that the official statistics publishedby INSTAT defines the working age as 15-54years for women and 15-59years for men.) Employedor working:anyonewho workedat all inthe last sevendays precedingthe survey and anyone with a permanent job who has not worked for the following reasons: own illness, maternity leave, a household member sick, holidays, education or training, temporary work load reduction and strike or suspension. Unemployed:an individual who has not worked in the last seven days preceding the survey, but has looked for a job in the last four weeks. An individual is also considered unemployed if she/he has not worked in the last seven days, has not lookedfor ajob in the last seven days, but has found ajob which will start later. Labor force: Sumtotal ofemployed andunemployed Out of labor force: an individualis considered out of labor force is she is not employed and not looking for ajob. Labor force participationrate (LFP): LaborForcedividedby WorkingAge Population Employmentrate: Employeddividedby WorkingAge Population Unemploymentrate: Unemployeddividedby LaborForce Number ofworking hours per month: Product of hoursworkedper week and number of weeks worked per year, dividedby 12, the number of months inayear. Full-timeemployment: Employmentis consideredfull-time ifan individualworks 35-t hours perweek. Part-time employment: Employmentis considered part-time if an individualworks less than 35 hours perweek. 2. Employment Macroeconomic reforms and restructuringof the early nineties ledto a sharp fall in employment rates in Albania, and though male employment almost recovered by the end of 1990sto its pre-transitionlevels, Figure 4.2: Albania employment rates, 1991-2001 1female employment 'remained very low. At 60 percent in the year 2001, 80 Albania's emp1o)ment rate I 70 was high by Central and I 60 1 Eastern European norms, 1 50 All and was comparable with :t?:;:ent ~ 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 rates in Bulgaria, and ~ 0-Male0-Female-National average Romania. I Source: INSTAT,Albania Reduced employment rates of the nineties were a result of the falling public sector employment. The numberofjobs inthe public sector decreaseddramaticallyfrom 850,000 in 1991to 189,000 in2001. This drop in the public sector employment was mostly due to mass privatization of state-owned enterprises. Liberalization of prices, end of state subsidies, lack of capital, and foreign competition - measures 57 undertaken to make the national economy competitive in the long run - severely weakened newly privatized companies and led them to bankruptcy. The industrial sector - extraction of minerals, metallurgy, equipment, chemicals, paper andtextiles-were hit the most. Reducedrevenuesdecreasedthe numberofjobs in government structures as well. 1991 1993 1995 1997 1999 2001 Total 1,404,091 1,045,918 1,137,829 1,107,677 1,065,104 1,063,093 Public sector 850.091 375.338 275.887 226.295 201.429 188.965 -Budgetary nla 182,299 159,619 145,319 128,452 120,114 -Nonbudgetary nla 193,039 116,268 80,976 72,977 68,85 1 IPrivate sector (non-agriculture) 30,000 80,277 111,942 120,382 102,675 113,128 Private sector (agriculture) I 524,000 I 590,303 I 750,000 I 761,000 I 761,000 I 761,000 Source: INSTAT, Albania Employmentinthe privatesector, on the other hand, increased significantlyduringthe same period, with number of jobs in the non-agriculturalprivate sector increasing almost four times. Private agricultural employment remainedthe largest across sectors, andthe number ofjobs in agriculture increasedby more than 200,000 (or 45 percent) between 1991and2001. Inthe year 2001, more than 80 percent of total employment was providedby the private sector, with an absolute majority ofjobs beingin agriculture(71 percent). Aside from agriculture,most ofthejobs in the year 2001were in the social sectors (health and education), services (commerce, hotels and restaurants), manufacturingindustries (mostlyfood processingandtextile), andenergy andtransport. The survey basedemployment rate in2002 of 54 percentwas less than the official employment rate o f 60 percent reportedfor the year 2001. Employment rate was considerably higher among men (64 percent) than women (45 percent), and individualswith post-secondaryeducation hada muchhigher employment rate (70 percent) than those with less education. Percent of total employed 30.4 49.0 45.5 26.7 24.9 38.5 Total Employed 64.1 44.8 50.9 57.1 70.1 53.9 Sixty two percent ofthe employedhadfull-timejobs and 39 percenthadpart-timejobs, and a muchlarger share of men (70 percent) than women (51 percent) had full-time employment. More individuals with post-secondary education (75 percent) had full time jobs, compared with 73 percent of workers with secondary education and 55 percent of workers with less than secondary education. The employmentrate in Tirana was only 39 percent, and was the lowest among all regions in Albania. On the other hand, 87 percent of jobs in Tirana were full-time. The coastal area also had a large share of full-time jobs (68 percent). Employmentinthe mostly agriculturalcentraland mountainareas was largelypart-time. 58 Percent of total employed 32.5 I 44.0 I 51.4 I 13.1 I 23.8 I 46.3 I 38.5 TotalEmployed I 54.0 I 57.0 I 57.7 I 38.8 1 37.3 I 66.2 I 53.9 Source: Albania LSMS, 2002 Laborallocationacross sectors showsthat about 56 percent oftotal employment was inagriculture,which i s less than the share reported by official statistics (72 percent). The share of non-agriculturalemployment (34 percent), on the other hand, was larger than the share reported in the official data (20 percent), and most of non-agricultural jobs were in services (9.5 percent), health and education (6.9 percent), construction(6.8 percent), energy andtransport (4.3 percent), andpublicadministration(4.3 percent). Self-employmentwas the most common type of employment, and in the year 2002, almost two-third of the employedwere either self-employedor worked as an unpaidworker in a householdfarm or non-farm enterprise, and only one-third were wage-employed. An absolute majority of self-employmentwas in agriculture. Wage-employed The share of wage employed (35 percent) was significantly higher than the share reported in 1996.35 Wage-employment was almost evenly divided between the private and public sectors, which together providedabout 98 percent of paidjobs, andNGOs and publicwork programsprovidedthe remainingtwo percent of paidjobs. Jobs ingovernment structures (40 percentoftotal wage-employment) dominatedthe public sector employment, while state-owned enterprises provided a substantially smaller share of total employment (8 percentoftotal wage-employment). Morethan 10years I 41.4 9.9 6.2 9.2 25.7 Total 100 100 100 100 100 Source: Albania LSMS, 2002 Amongthe wage employed, employment in government structures remainedthe most secure, with almost 80 percent of employees havingtheir jobs for at least three years and 60 percent beingemployed for six 35 Accordingto Rashid(2001), the share ofwage-employedin 1996was 23 percent.The 1996LSMS didnot cover Tirana, where wage employment is dominant. 59 years or more. Jobs were less secure in the state-owned enterprises, where, on average, ajob lasted three years or less. Half o fjobs in private companies and NGOs were created in the last five years. An analysis o f employment within sectors shows that most o f the recent jobs were created in the private sector. The survey data indicate that private firms and individual employers provided more than 50 percent o f the jobs created inthe last five years preceding the survey. More than 80 percent o f wage employment was full-time. Within sectors, state-owned enterprises offered the largest share o f full-time jobs (95 percent o f total employment with SOE). Private companies also provided mostly full-time employment (91 percent), as did NGOs (88 percent). The share o f part-time jobs was the highest amongjobs offered by government structures (20 percent) and individual employers (26 percent). Source: Albania LSMS, 2002 TheInformal Sector According to the ILO classification, the informal economy i s characterized as consisting o f units engaged inthe production o f good or services with the primary objective o f generating employment and incomes to the persons concerned. These units typically operate at a low level o f organization, with little or no division between labor and capital as factors o f production and on a small scale. Labor relations - where they exist - are based mostly on casual employment, kinship or personal and social relations rather than contractual arrangements with formal g~arantees.~~ According to the ILO, the informal sector can be regarded as a group o f operational unitsthat form part o f the household sector as household enterprises or unincorporated enterprises owned by households. Based on this definition, the informal sector consists mostly o f own-account workers and non-farm household enterprise^.^^ Regardless o f relatively well-defined concepts, it i s very difficult to quantify the informal economy. Depending on particular analytical purposes and on the basis o f the data collected, researchers use different criteria to characterize the informal sector in different economies. One o f the most frequently 36 According to ILO, production units of the informal sector have the characteristic features of household enterprises. The fixed and other assets used do not belong to the production units as such but to their owners. The units as such cannot engage intransactionsor enter into contracts with other units, and not incur liabilities on their own behalf. The owners have to raise the necessary finance at their own risk and are personally liable for debts or obligations incurredinthe production process.Expenditure for production is often indistinguishable from household expenditure. Similarly, capital goods such as building or vehicles are used indistinguishably for business and householdpurposes. 37 In addition to the ILO concept, the economic literature commonly uses other operational definitions of the informal sector. One of the widely used definitions is based on the status of a household enterprise, which being informal, is often not registered with local authorities, and as such remains outside the official institutional framework. Another common approach is to use the labor force status of the individual as the criterion for differentiating the informal from the formal sector. 60 faced dilemmas i s whether or not to include the agricultural self-employment as part o f the informal sector, and many decide not do so because the agricultural self-employment does not require formal registration and licensing, and all farming activities enjoy legal status. Following ILO guidelines, the informal sector i s defined here as own-account workers and workers in non-farm household enterprises. The informal sector in Albania i s analyzed using individual employment data and data on non-farm household enterprises. The employment data shows that about 10 percent o f the employed in Albania worked in household non-farm enterprises or as own-account workers. The size of the informal sector i s much larger in urban areas (24 percent) than in rural areas (5 percent). About 15 percent o f the employed inTirana are own-account workers or work innon-farm household enterprises. Almost three quarters o f individuals in the informal sector were employed for at least three years, and about 10 percent o f them were engaged in the informal sector for more than 10 years. More than 90 percent o f the people inthe informal sector started their own business as a way o f searching for ajob. Sector of employment Tirana Other Urban Rural Total Household farm 0.0 4.7 78.5 54.9 Own account or non-farm householdenterprise 15.0 23.8 5.4 10.4 Wage employment outside household 85.0 71.5 16.1 34.7 Total 100.0 100.0 100.0 100.0 Seventy five percent o f the self-employedwere men and 25 percent o f them were women. All age groups and all educational levels are present among the self-employed. The share o f self-employed with post- secondary education i s 11 percent, which is slightly higher than that o f the total labor force (10 percent). -~ ~- 7 Among the self-employed, Figure4.3: Duration of employment in the informal sector (aspercentof total employment ) I more men than women have post-secondary education 40% (12 percent versus 9 30% percent). Sixty four percent 20% o f the self-employed reside 10% 0% in Tirana and other urban ~ less than 7 7-12 months 1-2 yrs 3-5 yrs 6-10 yrs more than 10 areas, confirming, once months years again, that informal J activities are more common inurban areas.38 The nonfarm household enterprisesdata show that about 15 percent o f households in Albania had a non- farm enterprise. Both datasets, individual employment data and data on non-farm household enterprises, show that reselling goods (49 percent), taxi and freight transportation (15 percent), manufacturing (7 percent), construction (6 percent), agro-business and food processing (6 percent), and restaurants (5 percent) were the most common among household businesses. While agro-business, food processing and manufacturing are more common inrural areas, there are more restaurants, services, and health and social work places in urban areas. Sales, transport, and construction, on the hand, are quite common in both, urban and rural, areas. 38Recallthat the definition usedhere does not include agricultural self-employment as part o f the informal sector. 61 Other services (Legal, marketing, consulting, etc.) 3.4 1.7 1.0 1.6 Health and social work 3.5 1.4 0.6 1.4 Entertainment 0.0 1.9 1.0 1.3 Other 11.7 46.0 3.8 5.9 More than 80 percent o f non-farm enterprises had been in business at least two years, and 40 percent o f them were more than five years old. Forty percent o f household enterprises were home-based or had a fixed location, but more than 30 percent o f them had no fixed location and functioned on streets. Single households independently owned 94 percent o f non-farm enterprises, while the remaining six percent had more than one household as owner. Less than a half o f these enterprises were registered at a local tax office, and just about 55 percent o f them kept some form o f accounting. Seventy seven percent o f businesses hadjust one employee, mostly the household head himself, while the rest also employed other household and non-household members. 3. Unemployment Highrates of unemployment inAlbania, like inmany other Central and Eastern European countries, have been a result o f mass privatization o f inefficient state-owned enterprises carried out by the government as a part o f macroeconomic reforms and restructuring. Dramatically reduced public sector employment led many workers to look for jobs in the private or informal sectors, or to totally leave the labor force. At 17 percent in the year 2000, the last year for which the regional data i s available, the unemployment rate in Albania was still very high and was comparable with rates in Serbia and Montenegro (26 percent), Croatia (21percent), Slovakia (19 percent), and Bulgaria (18 percent). Since its peak in 1993, when more than 22 percent o f the labor force in Albania was out of work, the unemployment rate has fallen to the mid-teens and in the 2001, the last year for which official data is available, the unemployment rate was 14.5 percent. Unemployment in Albania has been historically higherfor women than men. 1991 1993 1995 1997 1999 2000 2001 Total registeredunemployment 140 301 171 194 240 215 181 - Male 65 159 91 110 130 113 96 -Female 75 142 80 84 110 102 85 Unemployment rate, percent 9.1 22.3 13.1 14.9 18.4 16.8 14.5 -Male 7.8 20.9 11.7 13.9 16.4 15.0 13.2 - Female 10.5 24.2 15.0 16.6 21.4 19.3 16.5 62 1993 1995 1997 1999 2001 Total registeredunemployment 100 100 100 100 100 -Male 53 53 56 54 53 -Female 47 47 44 46 47 -Withsecondaryeducation 45 49 48 49 48 -Withuniversityeducation 3 2 3 3 2 Fromtotal registeredunemployment: - 15-19vears old - 21-34 15 12 12 13 7 years old 49 46 46 46 47 - 35 years oldandover 36 42 42 42 46 The data for 1993-2001 indicate that more men register as unemployed than women. Inthe year 2001, there were just two percent of individuals with higher education among the registered unemployment, suggesting that university education is most likely to help escape unemployment. Primary or secondary education, on the other hand, does not guarantee such opportunity. In 1993-2001, the share of 35 year olds and over among the unemployed increased by 10 percent, while the share of 15 to 19 year olds decreasedfrom 15 to 7 percent.Inthe year 2001, 93 percent ofthe unemployedwere in primaryworking age. The LSMS 2002 data show that the unemployment rate in Albania in 2002 was slightly below 10 per~ent.~'Consideringdiscouraged, seasonal and laid off workers as unemployed, the unemployment rate becomes 15 percent. Unemploymentis higher among women than men, with this gender gap increasing when more relaxed definitions of unemployment are applied. According to the survey data, unemployment is higher among individuals with secondary education than among individualswith less than or morethan secondary education. Table 4. 12:Laborforce and unemploymentby gender and education (percent ofworking age /UnemploymentRate3** 17.0 I 21.9 I 21.3 5.6 19.1 * Counts ** discouraged/seasonal/luid off workers us unemployed. Counts individuals who work less than 15 hoursper week in agriculture as unemployed. At the rate of 20 percent, the unemployment rate in Tirana in the year 2002 was twice higher than the national average, and was the highest among the regions. According to the survey, almost every fifth adult of working age in Tirana was unemployed. Overall, the unemployment rate in urban areas (23 percent) was almost seventimes higherthanthe rate inrural areas (3 percent). 39Usingthe standardILOdefinition 63 Region Location Urban (except Coastal Central Mountain Tirana Tirana) Rural Total Labor ForceParticipationRate 58.9 62.5 63.1 48.2 48.4 68.4 59.7 Out ofthe Labor Force 41.1 37.5 36.9 51.8 51.6 31.6 40.3 UnemploymentRate 8.4 8.9 8.6 19.6 22.9 3.2 9.8 UnemploymentRate2** 14.5 14.3 13.9 25.2 32.2 6.7 15.4 UnemploymentRate 3 *** 15.1 21.6 19.0 19.6 23.9 17.2 19.1 More men than women are unemployed, and more than a half of the unemployed have just primary education.The central region makes up morethan 40 percent of the unemployed, and about 70 percent of the unemployedare 25-64 years old. I Figure 4.4: Composition o f t h e unemployed Source: Albania LSMS, 2002 On average, an unemployedpersonspends 14 months looking for ajob, and the length ofjob search does not significantly deviate across gender and education. The job search lasts longer in the mountain and central regions, though even in Tirana the unemployed spend, on average, one year, looking for a job. It takes longer for an older worker to find ajob than for her younger competitor. On average, the 15-25 year olds find ajob within less than ayear, but for workers of 46 years old and above thejob search might take up to 18 months. Just a quarter of the currently unemployed4' were engaged in income generating activities in the last 12 months precedingthe survey. Long-term unemployment is higher among women than men, with almost 90 percent of currently unemployedwomen being out of work for at least a year. Just 45 percent of the unemployedare registeredat the local labor office, and informal networking through friends and relatives i s the most commonway of looking for ajob. 40 Current, or short-term,unemployment is basedonthe labor market status withinthe last seven days precedingthe survey. 64 Box 4.2: Why is the unemploymentrate so highinTirana? According to the survey data, Tirana has the lowest labor force participation rate and the highest unemployment rate among the four geographical areas in Albania. This case of a capital city having a higher unemployment rate than other parts of the country is unusual for Central and Eastern Europe, becauseexperience inthe regionshows that capitalcities usuallyprovidemore employment opportunities and, consequently, have lowerunemploymentrates. One explanationof the high unemployment rate in Tirana might be in the type of employment that the city provides.Historically, the government of Albania invested heavily in the capitalcity building many industrial enterprises in and around Tirana. By the late eighties, Tirana was providing employment to large numbers of people in government structures and state-owned industrial enterprises. Market economy reforms and privatizationof the early nineties led to mass lay-offs from newly privatized, but bankrupt, industrialenterprises, and reducedpublic revenues inturn brought a sharp reductionin public sector budgetary jobs. Having had a large share of industrious and public administrationemployment, Tiranaexperienced a sharp rise inunemployment. Another explanation for high unemployment rates in Tirana lies in the massive internal migration in Albania. Economichardships of the early nineties forcedmany people from poor rural areas to move to Tirana in search of better employment opportunities.Even if employment in Tirana was already scarce, prospects of future employment opportunities, private sector developments, and foreign aid and investment attractedthousands into the capital city. Even if officially the number of populationin Tirana is less than 500,000, the city outskirtsshelter about 300,000 morepeople. High unemployment rates are, in general, common for urban areas in Albania. As mentioned before, people living in rural areas have access to land, and, regardless of how small a land plot can be, it provides some earning opportunities. The poor living in urban areas, however, do not have even such opportunity andhaveto search for employment outsidethe household. 4. Wage Developments High inflation swept away gains in nominal wage in the early 1990s and pushed the real wage into the negative zone. The sharpest I Figure 4.5: Real wage developments, 1993-2001 declinewas in 1992, when thereal wage decreased 25 percent. In 1994 and 1995, the real wage increased at the rate of 25 percent annually. The pyramid scheme crisis and high inflation of 1997, however, retracted most of the gains, and the real wage dropped 17 percent. The post-crisis 1 HChange of monthly averagewage Change o f CPI 0 Real wage growth macroeconomic stabilization and low, singe-digit, inflation rates of recent years have allowed the real wage to grow again, andsince 1998 it has been growingmorethan 10 percentannually. Between 1997 and 2000, transport and communications had the highest nominal wage increase across industries.In the year 2000, an average monthly wage in transport and communicationwas the highest 65 across industries, followed by wages in manufacturing and services. Even ifthe wage increase in services in 1997-2001 was the secondhighest acrossindustries, earnings inthis industrywere unsteady. Industry 1997 (Leks) 1997 1998 1999 2000 2000 (Leks) Manufacturing 9,411 100 115 130 141 13,230 Construction 8.340 100 127 131 150 12.489 lTransDortandcommunication I 9,350 I 100 I 126 I 155 I 174 I 16,225 I Trade 8,819 100 109 124 123 IO, 889 Services 7,814 100 152 137 167 13,012 Analysis of the LSMS 2002 data shows that in 2002, on average, NGOs and humanitariaddonor organizations paidthe highestmonthly wage and state-owned enterprises paid the lowest monthly wage in Albania. Employment inthe private sector (with a private company or an individual employer) brought a higher wage than employment in the public sector (with government structures or state-owned enterprises). Across all sectors, wages in Tirana were higher than in any other region. Outside Tirana, wages inthe coastal region, on average, were higherthan intwo other regions. Source: Albania LSMS, 2002 According to the survey data, an individual with post-secondary education earned 23,991 Leks a month, which was 37 percent more than an average earning with secondary education and 44 percent more than an average earning with less than secondary education. The wage difference was especially high inNGOs and private companies and was low in state-owned enterprises. Table 4. 16:Average monthly wage by sector and education (Leks) Source: Albania LSMS,2002 41 Out of three people in the survey, who reported being engaged in public work programs in Tirana, two were technical professionals and one was inadministration. 66 A more detailed analysis of average monthly wages across industries shows that jobs in banking, international organizations or foreign forms, sea and coastal transport, and fishing paid highest wages in Albania. Employment in financial markets, information technology firms, and NGOs also led to high earnings. Compared to the national average, wages in manufacturing industries, services, health and social work, and education were very low. Employment with private households paid the lowest wage in the country. Source: Albania LSMS, 2002 On average, men earned higher wages than women across all sectors, except inNGOs and humanitarian or donor organizations. The female earning disadvantage is most visible in the private sector. Within the public sector, on average, government structures also paid men more than women, while female and male earnings at state-owned enterprises were mostly equal. 67 1 Figure4.6: Averagemonthly wage by sector andgender (inLeks) I I I State-ownedenterprise Government Nationalaverage I Individual employer Privatecompany I NGO or humanitarianorganization I 10,000 15,000 20,000 25,000 30,000 ~ Source: Albania LSMS, 2002 Returnson education:OLS estimatesof earnings function The OLS estimates o f earnings functions help determine how different factors, such as education, gender, age, region, sector and industry o f employment, affect wages. The explanatory variables used to explain wage differences are education, gender, age, region, sector o f employment, industry, and tenure. The regression results show no increasing returns to education for the national economy. In the public sector alone, those with college education earn more, but there are no increasing returns to education in the private sector at all. In contrast to advanced CEE economies, where structural changes o f the transition years led to increasing returns to education, the Albanian economy still heavily relies on low-skill labor and does not seem to reward education. Interms o f demographic characteristics, the regression results show that age is not an important factor in determining wages. Women have a clear disadvantage in earnings, and all else being equal, a female worker earns on average, 20 percent less than a male worker. Regional differences are marked, and jobs inTirana pay higher wages than any other area inthe country. Regardless o fthe sector of employment, a worker in Tirana, on average, earns 20 percent more than a worker in the coastal area, 30 percent more than a worker inthe central area, and almost 40 percent more than a worker inthe mountain area. Sector of employment Education Public (Leks) Private (Leks) Wage differential (percent) Less than secondary 13,455 17,890 -0.33 Secondary 14,636 20,404 -0.39 More than secondary 20,661 35,302 -0.71 Total 16,685 20,380 -0.22 Workers inthe private sector are paid almost 30 percent more than inthe public sector. The private-public wage differential varies by educational attainment, and individuals with post-secondary education in the private sector have the largest wage gain (71 percent). Higher earnings make employment in the private sector more attractive, and, in turn, draw more people into this sector. Long-term work experience in the public sector does not bring significant earning benefits. Private sector jobs, however, do reward work experience - a private sector employee with three or more years o f working experience at the current job earns on average 25 percent more than a new hire. 68 5. Labor Market and Poverty Employment i s a key determinant o f poverty and, generally speaking, providing more employment opportunities and job creation are seen as ways to fight poverty. The LSMS data supports this links between the labor market and poverty in Albania: the poverty rate associated with the labor force status shows that the incidence o f poverty was the highest among the unemployed (33 percent), with every third unemployed person being poor. The poverty rate among the employed and those out o f the labor force was almost equal at about 22 percent, suggesting that the working population i s not guaranteed against poverty. On the other hand, this situation implies that those out o f the labor force rely on income from outside the formal labor market, such as earnings in the informal sector and remittances from household members working abroad, and might have other coping strategies to fight poverty. The employment status o f the household head is crucial for avoiding poverty. The survey data shows that the incidence o f poverty was much higher in households where the household head was unemployed. On average, the poverty rate was almost twice as highfor these households as comparedto households where the household head had ajob. Albanian households are usually large and most o f them have more than two wage earners. Poverty incidence among households with two or more working members (24.5 percent) i s lower than among the households with one (27.2 percent) or none (28.2 percent) working member, but this rate is still very high suggesting that multiple wages are not a guarantee against poverty. Household employment ratio42 Share o f people43 Poverty incidence Zero 7.0 28.2 One-third or less 14.5 27.2 More than one-third 78.5 24.5 Total 100.0 25.4 Labor market characteristics of thepoor The poor were less often employed or out o f the labor force and much more frequently unemployed than the non-poor. The unemployed accounted for 9 percent o f the poor and only 5 percent o f the non-poor. Slightly more than 50 percent o f the poor and non-poor were employed, but the earning capacity o f the poor was noticeably weak because most o f them were employed in low-productive and low-earning industries. An almost twice the proportion o f the poor (34 percent) than the non-poor (18 percent) were low-paid, andjust about five percent o fthe poor earned highwages comparedwith 16 percent o f the non- poor. More than seventy percent o f the poor worked in agriculture, compared with 50 percent o f the non- poor, and aside from agriculture, the poor were more frequently employed in construction, while the non- poor worked mostly inretail, public administration, and the social sector. 42The ratio i s defined as the number of working membersdivided by the size of the household 43Percentage of persons in householdswith given employment ratio 69 Mining industry 2.1 0.7 1.8 Energy sector 1.5 0.9 1.4 Textile manufacturing 1.2 1.5 1.3 Food industry 1.4 0.3 1.2 Other 11.8 4.9 10.3 Across all sectors, the poor earn much lower wages than the non-poor, with an average monthly salary o f the poor being 46 percent less than what the non-poor earn. The private sector pays higher wages than the public sector, but the wage differential between the poor and the non-poor i s almost equally large in both sectors. State-owned enterprises pay the lowest wage across sectors, and these enterprises have the smallest poverty wage differential as well. Employer Non-poor Poor Total PercentDifference Government structures 17,700 12,197 17,242 -55.0 Private company 22,158 16,845 21,309 -53.1 State owned enterprise 14,215 12,828 13,903 -13.9 Individualemployer 20,389 15,422 19,368 -49.7 ITotal [ 19,333 I 14,745 [ 18,680 I -45.9 Source: 2002 Albania LSMS Employment as one of the key determinants ofpoverty The results o f a multivariate analysis confirm that along with household size, education, gender and location, the employment status o f household members is also a key determinant of poverty. These results show that when a larger share o f family members i s unemployed, the household is significantly more 44Eaming categories are calculatedjust for wage employed. 70 likely to be poor. In rural areas, on the other hand, having a household member engaged in dependent work or being self-employed in a non-agricultural business reduced the probability o f being poor. Poverty and the Informal Sector In 2002, across all employment groups in Albania, the informal sector workers had the lowest poverty rate o f 11 percent. The incidence o f poverty among the informal sector workers was twice as low as the national average rate for the employed (22 percent), and almost three times lower than the poverty rate among the employed inagriculture (29 percent). 6. ConcludingRemarksand Policy Implications Albania enjoys a high labor force participation rate (66 percent), but the levels o f skills are low, and almost two-thirds o f the labor force havingjust basic or less than basic education. Almost two-third o f the working population i s either self-employed or work as unpaid workers in a household farm or non-farm enterprise. Most self-employment i s in agriculture, and only one-third o f the employed earn wages. About 10 percent o f the employed in Albania work in the informal sector, and this share being about 24 percent in urban areas. There are more men than women in the informal sector. There is evidence o f long-term employment in the informal sector, with almost 75 percent o f individuals being engaged in informal activities for more than three years, and about 10 percent working in the informal sector for more than 10 years. About 15 percent o f households in Albania have a non-farm enterprise, owned independently by that household (94 percent), or shared by two or more households (6 percent). Less than a half o f all entities are registered with a local tax office, about 55 percent have some sort o f accounting, and most o f them employ just one or two persons. Reselling goods, providing transportation services, manufacturing, construction, food processing, and restaurants are among most common types o f informal sector activities inAlbania. Unemployment remains widespread, and the unemployment rate is especially high in urban areas. Long- term unemployment reigns - it takes, on average, more than one year to find ajob. Tirana has the lowest labor force participation and the highestunemployment rate compared to other geographical areas in the country. After sharp declines o f the early and late-nineties, real wages in Albania are growing again. Over the 1999-2001 period, real wage increased, on average, by 10 percent annually. Still, wages in Albania remain very low - in the year 2002, an average monthly salary was 18,680 Leks (about $US 133), and almost 50 percent o f the wage employed had earnings o f less than $US 100 per month. L o w pay i s mainly associated with low educational attainment, and also with working in low paying industrieswith insecure jobs, such as services, agriculture, and the social sector. Women earn significantly less than men with equal qualifications. Private sector earnings are higher that earnings in the public sector, and the wage differential increases with educational attainment. Employment is one of the key determinants o f poverty - the probability o f being poor i s significantly higher when a household member i s unemployed. Inrural households, on the other hand, having a wage- employed member or a member who is self-employed in a non-agricultural business reduces the probability o f being poor. The poor are more often unemployed than the non-poor, and the poverty rate among the unemployed i s much higher than among the employed and those out o f the labor force. The incidence o f poverty among those out o f the labor force i s equal to the incidence among the employed. Yet, working does not prevent poverty - many o f the poor are employed, but their earning capacity i s weaker than that o f the non-poor because the poor are employed in low productivity industries, such as 71 agricultureand construction,and across all sectorsthe poor, on average, earn 46 percent less than the non- poor. With the poverty rate of 11 percent, the informal sector is not immune from poverty.Nevertheless, the rate of poverty in the informal sector is lower than that for the wage employed (14 percent), and almostthree times lowerthanthe povertyrateamongthe self-employedinagriculture(29 percent). The predominantlabor market and povertyissue in Albania is how to deal with the big excess supply of labor and the limited market mechanismsto originate adjustments. The excess supply of labor is a result o f high birth rates and the lack of demand for labor to absorb the supply. The government i s no longer able to maintainhigh levelsof labor demand, and the newly created market mechanismsare too weak to offset employmentlosses. In addition, wages are low and there is not much room for price-adjustments. The excess labor supply has brought high rates of unemployment, hidden unemployment, and a largely subsistence-based informal sector. As the result, large segments of the labor force, such as younger people, new graduates, women, andpeopleof pre-pensionage do not participateinthe labormarket. 72 CHAPTERV: HEALTH 1. Overview The health care sector in Albania is mostly publicly owned, and the Ministry o f Health (MOH) i s the major provider o f services. The private sector i s limitedmostly to the distribution and commercialization o f pharmaceuticals and to dental care services, though the last decade has seen a surge o f private outpatient practices and diagnostic centers in urban areas, especially Tirana. The MOH owns and administers all health care services through its district directorates, with the exception o f the Tirana region where a pilot program is now in place that gives all administrationresponsibilitiesfor primary health care and public health to the newly created Tirana Regional Health Authority (TRHA). The health system is mostly funded through budget allocations, although in 1995 the Health Insurance Institute (HII) was established to secure an additional source o f finance to the system. The HI1 i s funded through payroll taxes and by budget transfers for the vulnerable groups o f the population but since a large percentage o f the Albanian population is involved in informal activities, this scheme does not represent a significant source o f funding. Primary Health Care (PHC) is provided by health care centers and health posts in rural areas; in addition to health centers, PHC in urban areas i s also provided by large polyclinics offering specialized outpatient care. In 1997, the government developed a PHC policy that aims at securing one health post per village and a health care center per commune, and to create PHC teams ledby family physicians. The health care centers are usually staffed by three primary care doctors and by nursing staff. Inthe last few years there has been an intensified effort aimed at renovating and updating many PHC facilities, which lacked maintenance and were heavily damaged after the upheavals o f 1991 and 1997. Table 5. 1:Health CareServices 1990 1999 Outpatient care Number of facilities 3302 2242 - Health centers 1046 567 -Ambulances 2196 1624 -Polvclinics 60 51 Inpatient Care Numberofhospitals 160 51 -ofwhichrural 86 0 Numberofhospitalsper 100,000 inhabitants 4.87 1S O Numberofhospitalbeds 13228 10207 NumberofhomitalbedsDer 100.000 inhabitants 402.6 302.5 Source: INSTAT 2002 Secondary and tertiary care i s offered by public hospitals and clinics. Inthe last decade, there has been a large decrease in the number o f hospitals and hospital beds in the country. Although many of the institutions that closed were poorly equipped, their closure adversely affected access to these services. Albania has one of the lowest hospital beds per inhabitants in all Europe, and the number o f beds per 100,000 inhabitants i s half the average o f the EU countries, and less than half the average o f the CEE countries. 73 2. Statusof Population Health Despite its low per-capita income level and limited health services, life expectancy in Albania is 72 years for men and 78 for women, one o f the highest life expectancies in the area and higher than the average o f the CEE and o f all Europe. Diet and lifestyle are thought to be behind these positive indicator^.^^ However, there are large regional variations in life expectancy, and areas in the poor northern regions have significantly lower life expectancy at birth. In contrast with the life expectancy figures, infant and maternal mortality rates in Albania - at 28 per 1000 live births and 26 per 100,000 live birthsrespectively -are amongthe highestintheregion. Infantmortality isthoughtto beparticularly highinruralareas, where most o fthe poor people live. Source: WHO/EuropeHealth For All (HFA) data set *:Data correspond to theyears 1990and 1999. Albania's death rate today (5 per 1,000 inhabitants, according to European Observatory, 2003) i s very similar to that o f the year 1990, but there have been significant changes in the causes o f death since. Circulatory diseases are still the most important cause o f mortality in Albania, but infectious, respiratory, and digestive system diseases have decreased in importance as causes o f mortality, while deaths due to injuries and neoplasm have markedly increased. Disease 1990 1999 Infectious & parasitic diseases 10.80 2.22 Neoplasm 66.70 75.50 Endocrinal, metabolic, nutritional 5.30 3.82 Blood. blood-forming organ diseases 3.00 1.10 IMental disorders I 0.00 I 1.87 I Circulatory system diseases 204.00 204.69 Respiratory systemdiseases 94.30 3 1.30 45Comparing Albania's lifestyle indicators with those of the average of the EU countries usingthe HFA data set shows that although the percentage of daily smokers in Albania is above CEE averages, alcohol consumption per capita is one of the lowest in Europe, and the quantity of fruits and vegetablesavailable per personis one of highest. 74 Source: INSTAT 2002figuresper 100,000 inhabitants) Self-ReportedHealth Status The LSMS 2002 collected information on morbidity, which is summarized in Table 5.446About 16% of the Albanians report havinga chronic condition, and among those reporting a chronic illness the majority are 50 years and older (see Appendix for age adjusted self-reported morbidity). This first measure of self- reported morbidity also varies across gender and region of residence. More women report having a chronic condition than men, and people outside the Tirana region are more likely to report a chronic 40% ofthe people), bone (40%), the nervous system (lo%), andthe respiratory system (9%). illness. The chronic conditions reported most were diseases of the circulatory system (reported by about Male Female Tirana Other urban Rural Not poor Poor Ext. poor Total Chronic 0.140 0.183 0.135 0.173 0.162 0.172 0.133 0.111 0.162 (0.005) (0.006) (0.009) (0.009) (0.007) (0.006) (0.009) (0.016) (0.005) Illness 0.127 0.147 0.039 0.124 0.164 0.130 0.158 0.188 0.137 (0.007) (0.008) (0.007) (0.010) (0.010) (0.006) (0.013) (0.031) (0.007) Data from the LSMS show that close to 14% of the total population reports having an illness or injury in the four weeks before the interview. As in the case of chronic conditions, there are strong variations in self-reported morbidity across gender, age, place of residence, and poverty status. For the total sample, the difference across gender is not significant, but women of childbearing age and older are more likely to report an illness than men. In the case o f women in childbearing age, this result could be reflecting discomforts related to pregnancy. There are also significant differences across regions o f residence, and the proportion of people living in Tirana reporting an illness is significantly lower than the proportion of people living elsewhere. In contrast to the case of chronic diseases, the correlation between reporting an illness in the last four weeks and levels o f education i s negative. This is not surprising as there is ample evidence suggesting that people with higher levels o f education in general have better health outcomes. Finally, the proportion of poor people reporting an illness i s significantly higherthan the proportion of the non-poor reporting any illness or injury. The majority o f the people reporting an illness (close to 80%) 46 Measures of self-reported morbidity shouldbe treated with caution, as they can be positively correlatedwith the income and education level of the individual. For instance, people with higher levels of educationon average are more likely to visit a health care service where a chronic disease can be diagnosed, and are more likely to recognize symptoms of a disease that will promptthem to get a health check up. Similarly, the more educated are more likely to know whether or not they have a chronic condition. It is for this reason that more non-poor and more educated report havinga chronic healthproblemthan poor and less educatedpeople do. 75 reported having either cold or flu in the last four weeks. Among the poor, this percentage was equal to 84%, while among the non-poor it was 76%. Similarly, 86% o f people living in the mountainous region of Albania reported having cold or flu, compared to 72% o f people living inTirana or the Coastal area. 3. Health Insurance The health insurance scheme offered by the Health Insurance Institute (HII) i s still a very limited scheme, not only in terms o f the population covered, but also in terms o f the number o f services it covers. The scheme, established in 1995, was implemented in stages. Inthe first stage, the resources collected by the Institute were used to pay primary care physician salaries and to cover essential pharmaceuticals. In the second stage, starting in 1997, the scheme covered family doctor and general practitioner (GP) services, which were offered free o f charge to those covered, and a subsidy on almost 300 pharmaceuticals. At the beginning o f 2001, another 30 products were added to the list o f subsidized drugs; and in the Tirana region, the Health Insurance Fund extended coverage to all outpatient care services. Inthe Tirana region, entitlements now include outpatients specialized care along side the services o f the GP and family doctor.47 Incontrast to the rest o f the country, the HI1fund pays employees o fthe policlinics and healthcenters in the Tirana Region. Expenses for primary health care services are also financed by HII, inaccordance with an agreement between the TRHA and the HII. This agreement defines the health services package that HI1will purchase for every policlinic or health center. The health services package includes the expenses for the payment o f the general practitioner, specialty physicians, nurses, other personnel employed in the policlinics and inthe health centers, and other current expenses. According to the Law on Health Insurance (Law No. 7870, 1994), all economically active individuals in Albania, whether they are employees, employers, self-employed, or unpaid family workers are obliged to contribute to the scheme and obtain a `license.' The state i s responsible for the insurance contributions o f the economically inactive. The state pays the insurance contributions o f children, full time students, retirees, the disabled, the unemployed, pregnant women, and citizens under compulsory military service. However, data from LSMS 2002 show that only about 39% o f the total population reports having health insurance license. As can be seen in Figure 4.1, the proportion of people with a health license increases with income and varies greatly between rural and urban areas. People in urban areas, especially Tirana, are much more likely to have a license than people living in rural areas. The differences are also large across regions, and while 63% o f people living in Tirana report having insurance, only 17% o f people living in the Mountain region report having health insurance. Finally, children and people older than 50 years o f age are also more likely to have a health license than others, with close to 40% o f children 0-14 years old and more than 50% of people older than 50 years old (60% o f those older than 70) having a health license. 47EuropeanObservatory, 2003. 76 FigureS 1:Proportionof people with ahealth license I 0.8 iI 0 6 '0 I 0 4 0 2 1 2 3 4 5 ConsumptionQuintiles I Tirana 0Urban Rural i Sub-groups o f the population whose insurance contributions are paid by the State are more likely to have a health license than otherwise. For instance, children and older people are the most likely to have a license, since the State pays their contribution. However, it's important to note that although these groups are the most likely to have a license, not all o f them do. More than 80% o f children are full time students, and yet less than 40% o f them report having a health license. This suggests either not all eligible to be covered by the State know about it, or that only those with strong incentives to get it (such as people with illness that will benefit the most from the pharmaceutical discounts) have the health insurance license. Further, the structure o f employment makes it very difficult for the government to enforce the payment of health insurance contributions for people engaged in agriculture or any informal activity. About 55% o f the employed work in a household farm, another 10% in a non-farm enterprise, and only about 35% are involved in wage employment. Not unsurprisingly,the economically active are the least likely to have a health license. Among the active population, the unemployed (the only group for which the State covers the insurance contribution) are the most likely to have a health license, while the laid-off workers and the workers involved in agriculture for less than 15 hours a week are the least likely to have it. 4. Health Care Utilization Outpatient care Close to 17% o f Albanians reported a visit to a primary care provider in the recall period o f four weeks. Among those that reported a chronic illness, 57% seek outpatient care (Le., visits to a public ambulatory, private doctor, nurse, midwife or paramedic, and visits to traditional healers or "popular" doctors), and among those reporting an illness or injury, 58% seek outpatient care. On average, there are large variations in outpatient care utilization across population groups. There are large differences in utilization between people living in rural and urban areas, differences that persist even when age differences are taken into account. For instance, while in Tirana three in every four people seek outpatient care when ill, in rural areas only one o f every two do. Similarly, non-poor people are more likely to seek care when ill (61.6%) compared with the poor (48.8%). Also, and as expected, people with a health license are more likely to visit an outpatient care provider when illthan people who do not have a health license. As mentioned before, having a license decreases the price o f prescribed drugs and eliminates any payment for GP and family doctor services. There is anecdotal evidence suggesting that given the low quality of these services most o f the people that visit public primary care providers do so mostly to obtain drug prescriptions. Under these circumstances, people who might not be able to afford the full price o f a prescription drug might be discouraged to seek care. 77 Table 5. 5:Outpatient care utilization (percentof those ilr) IMale I Female I TiranaI Urban I Rural INot Poor 1Poor \Nolicense I License I11last 4 weeks Ambulatory I 50.0 I 52.1 I 68.6 I 56.0 I 48.4 I 54.9 I 42.2 I 45.2 I 61.4 Private Doctor 5.2 5.1 12.8 7.7 3.8 6.4 1.9 4.3 6.5 Nurseiparamedic 12.2 13.4 1.0 10.0 14.5 13.1 12.3 13.4 11.9 Popular doctor 0.2 1.o 0.0 0.0 1.o 0.6 0.7 0.8 0.4 Seek care 56.8 58.8 74.4 63.0 55.1 61.6 48.8 52.4 67.2 Total Ambulatory 12.7 16.0 10.7 15.4 14.6 15.3 11.6 I 11.2 19.4 Private Doctor 1.4 1.9 1.o 2.4 1.4 1.8 1.1 I 1.2 2.4 Nurseiparamedic 2.6 3.5 0.5 2.4 3.9 3.1 3.1 3.2 2.9 Popular doctor 0.2 0.4 0.0 0.2 0.4 0.3 0.3 0.4 0.2 Seek care 14.7 18.6 11.6 17.7 17.2 17.6 14.1 13.5 21.8 Public ambulatories receive the majority o f people looking for outpatient care. Data from the L S M S show that 85% o f people that visited a primary care provider went to a public ambulatory (72% only visited an ambulatory). Table 5.5 shows how the percentage o f people visiting an ambulatory service when illvaries across place o f residence, poverty status, and ownership o f a health license. People in Tirana and other urban areas are more likely to go to a public ambulatory for care than people in rural areas. This i s not surprisingas in many rural areas primary health care is offered by health posts staffed only with a nurse or midwife. Similarly, people with a health license are more likely to seek care in a public ambulatory than people without a license. Although doctors might not charge a fee to patients without a health license, only doctors can provide drug prescription and the license entitles patients to exemptions or discounts in the price o f pharmaceuticals covered by the health insurance. As health insurance in the entire country only covers the services o f general practitioners and the drugs they prescribe, the difference inutilization rate betweenthose having a health license and those that do not is only significant for those who went to a public ambulatory, not for those who visited a private doctor or a nurse. The participation of the private sector in health care in Albania i s very limited, and it is mostly concentrated in dental and pharmaceutical service^.^' However, in the last year there has been an increase in the number o f private diagnostic centers and outpatient specialized centers located mostly in urban areas. Not surprisingly, the proportion o f people visiting a private doctor when illi s only about 5%. The majority o f people using these services live in urban areas, mostly in the Tirana region where the percentage o f people visiting private providers i s almost twice as large as the percentage in other urban areas, and three times as large as in rural areas. As these services are on average more expensive than those offered by the public sector, poor people are on average three times less likely to visit a private doctor than non-poor people. One in every eight Albanians visits a nurse, a paramedic, or midwife when ill.Most o f the people using these services live in rural areas or in urban areas outside Tirana. Similarly, people living in the Central region are more likely to use these providers than people outside this region. Finally, not that many people seek the services o f a popular doctor when ill.Less than 1% o f people do, and most o f them live in the Mountain region. 48EuropeanObservatory, 2003. 78 These differences in the type o f outpatient care provider across regions can partially be explained by the differences in availability o f these providers across region. At the beginning o f the transition, many doctors left rural and remote areas lured by more lucrative opportunities in the cities, especially Tirana. The introduction o f new payments for the GPs, which included a capitation system with higher weights for rural and isolated populations, succeeded in attracting some GPs to rural areas but not to more remote areas (IHSD, 2000), resulting in large variations in the number o f health care personnel across regions o f the country (Table 5.6). Source: Estimates by Maeda, based on MOH data Differences in utilization across age groups are largest among children and the elderly (Table 5.7). While 66% o f the non-poor children seek care when ill,only 50% of poor children do. Differences in outpatient care utilization are largest among the elderly, and while about 70% o f the elderly non-poor receive outpatient care when ill,only 41% o f the elderly living in poor households do. 50-69 0.672 0.483 0.640 (0.033) (0.066) (0.030) b=70 I 0.720 I 0.415 I 0.636 I (0.050) (0.086) (0.050) Maternalhealthand Antenatal Care Albania's maternal mortality rate i s one o f the largest in the region: it is more than 3 times the rate prevalent in Serbia and Montenegro and more than two times the rate in the European Union. For this reason, one o f the main health objectives o f the Albanian Government National Strategy for Socio- Economic Development i sto decrease maternal mortality. 79 The Albanian LSMS does not have information on maternal mortality; however, it has ample information on women's access to health care while pregnant and for delivery. Even though almost all women who gave birth in the last three years before the interview received pre-natal care (93%), some income and regional variations in the proportion of women receiving this health service exist. Poor women, women living in the Mountain region, and women living in rural areas are less likely to receive care relative to others while pregnant. Similarly, among those receiving antenatal care, there are significant differences in the number o f consultations women seek. Women in Tirana, on average, seek more than twice the amount o f pre-natal consultations compared to women living inthe Mountain region or in rural areas. Table5. 8:Antenatal carebv Dovertv status and repions /Not-Poor( Poor Coastal Central Mountain Tirana lother Urban1 Rural I Prop. Receiving Pre-Natal care 1 0.96 0.87 0.92 0.95 0.83 1.00 0.97 0.91 (0.02) (0.01) (0.02) h m b e r of visits II I II II II II I I (0.01) 1(0.03)1 (0.02) (0.03) I(O.00) I 4.4 1 3 . 9 1 4.1 3.9 2.8 1 7 . 7 4.9 3.4 I (0.18) 1(0.22)1 (0.27)III(0.22) III(0.12) I(0.55) III (0.29) III(0.15) 1 Source: Albania LSMS 2002; Standard errors inparenthesis There are also differences in the percentage o f women receiving care at birth and in the type o f provider assisting in delivery. Even though almost all women were attended by a skilled provider (98%), women in urban areas and non-poor women were more likely to receive care from a doctor than from a midwife. Source: Albania LSMS2002 Non-prescribeddrugs About 30% people buy non-prescribed drugs when ill.There are no income differences in the percentage o f people buyingthese pharmaceuticals, but there are large regional differences. Albanians living in the mountainous areas o f the northeast part o f the country are more likely to buy non-prescribed drugs when illregardless of their age (43% buy non-prescribed drugs). Incontrast, people living in Tirana are significantly less likely to purchase non-prescribed drugs at any age than people living in any other region in Albania (17%). These results are not surprisingas people inthe Mountain areas are the least likely to go to a health service when ill,while people inthe Tirana region are the most likely to do so. Hospitalization Data from LSMS 2002 show that about 4% o f the total population reports having a hospitalization in the year before the interview. There are no differences in hospitalization rates across poverty status once age is taken into account. However, regional differences exist, and people inthe Tirana region have the lowest rate o f hospitalization inthe previous year (2.5) comparedto other urban areas (4.7) and rural areas (4.6). 80 5. ExpendituresonHealth Care Public Expenditureson Health Public expenditure on health amounted to 18,022 million Leks on health care in the year 2002.4' This amount represents about 3% o f total GDP, one o f the lowest shares in the Balkans, and about one-half o f the CEE average.50Public expenditure on health is not only low when compared to other countries in the region, it i s also low when compared to Albania's expenditure a decade ago. As share o f total government expenditure, public expenditure on health was 9.8% in 2002, slightly up from 8.8% in 2000. In contrast, Croatia and Slovenia spent 13.7 and 15.6 percent o f total government expenditure on health while the EU average was about 13 percent (1998). 1991 2000 2001 2002 Total public spendingiGDP 62.2 25.9 27.5 27.9 Public spendingon health/GDP 4.8* 2.3 2 2.7 Spendingon health as YOof total public spending 8.8 7.4 9.8 This limited public spending on health care has implied an increased participation o f the private sector in financing health care. Health care services are increasingly being paid out-of-pocket for both formal "co- payments" and informal payments. Using the LSMS data, combined with administrative data on public expenditure, public spending on health in the year 2002 is estimated to constitute about 30% o f total expenditure on health care. In contrast, the share o f public spending on health in Greece i s 51% o f total health spending, and in Slovenia it i s about 86%.51 Private Expenditures on Health The sharp decrease in public resources earmarked for health, has resulted in a large increase in private expenditure. Estimates from the Albanian LSMS indicates that household expenditure on health care is quite large, representing between 1.8 and 3.4 percent o f total GDP, and about 70% o f the total health e~penditure.~~ This trend is worrisome as most o f this expenditure is out-of-pocket, which has resulted in income inequalities in access to health care services and in a lack o f financial protection o f the population against catastrophic health expenditure. Outpatient Care Although there is no difference in the amounts paid by the poor and the non-poor for outpatient care services, there are large differences across regions. People living in the Tirana region on average pay significantly less for their outpatient care visits, including treatment, lab test, gifts, medicines, and transportation to outpatient care services compared to people living outside Tirana. However, total expenditure in Tirana is not significantly different from total expenditure in outpatient care o f people 49Source: Ministry of Finance, Medium Term Expenditure Framework (MTEF). 50In line with its commitment to improve access and quality of health care services, the Albanian authorities have started to increase expenditure on health care. The Government of Albania 2003-2005 MTEF has projected an increase inpublic resources earmarkedfor health from 2.7% to 3.1% of GDP by 2005. 5'Data for percentage of public spendingas percentageof total health spendingfor Greece and Slovenia comes from the WHOiEurope Health for All database. 52This percentage includes transportation costs. The upper bound was calculated by multiplying expenditure data with recall periods of a month by 12. The lower bound was calculated by calibrating data from the LSMS with known administrative data to correct for possible seasonality in health care expenditure that the survey does not capture. A description of the methodology usedto obtain the lower bound can be found inthe appendix. 81 living in urban areas o fthe Central region, but it is still significantly lower than either rural or urban areas o f any other region inthe country or rural areas o f the central region. There are several reasons why people in Tirana spend less on outpatient care. Since 2001 a pilot program has been in place inthe Tirana region, under which the whole primary care system inthe region has been reorganized and i s now under the direction o f the new Tirana Health Authority. The financing o f the whole system i s now the responsibility o f the Health Insurance Institute, which among other things pays for the salaries o f all doctors, nurses, midwives and paramedics. In contrast, in the rest o f the country the Health Insurance Institute i s only responsible for paying the salaries o f family doctors and GPs. The personnel paid by the insurance institute receive on average higher salaries than those paid by MOH. Also, and again only inthe Tirana region, health insurance covers all outpatient care services and not only the services provided by the GP or family doctors. Thus, the difference in outpatient care expenditure between Tirana and other regions is due to the lower percentage o f people paying for outpatient services inTirana. Percentage paying Coastal 40.7 39.4 93.5 32.3 30.2 97.9 Central 55.6 21.9 84.7 25.4 32.5 94.9 Mountain 1 57.6 I 27.7 91.8 I 37.9 57.3 98.3 Tirana I 25.8 9.5 82.2 I 22.2 5.0 87.5 Medicines constitute the major expenditure for people seeking outpatient care in all regions o f the country and represents close to 55% o f total expenditure in outpatient care visits. Almost everybody pays for the medicines prescribed (88% o f the total), although people with a health license receive discounts. Expenditure on non-prescription drugs in the four weeks before the interview was about 340 Leeks (about 10% o f the food poverty line). There i s no difference in total expenditure on non-prescribed drugs across income levels or regions o f residence. Expenditure on treatment, lab tests, and transportation each represent about 11-12% o f the total expenditure. Adjusting for place o f residence, there are no significant differences in outpatient care expenditure between those having a health license and those who do not, suggesting that health practitioners probably do not differentiate between the insured and the uninsured when charging fees. HosDital Care Table 5.12 shows average expenditure in hospital care among individuals reporting a hospital stay in the year before the interview. Hospitalizations are rare events and the estimates presented in the table are not very precise but they do give a rough idea o f the pattern o f expenditure across poverty status and across regions, Poor people on average spend about 30% less on hospital care than the non-poor (the difference is not statistically significant, however), both in rural and urban areas. 82 Total expenditure on hospital care can be considerable, especially for the poor. Thus, while the average per capita monthly consumption of a poor family is about 3,800 Leks, the average monthly expenditure on hospital care among poor people reporting a hospital stay in the last year is about 1,300 Leks, or close to 34% of the total per capita monthly expenditure. In the Tirana region, this situation is even more dramatic as average hospital expenditures equal 40% of per capita monthly expenditures of a poor family. Source: Albania LSMS 2002; Standarderrors inparenthesis Informal payments Informal payments are unofficial payments to a health care provider for services, which are supposed to be provided at no charge to the patient. Although in some Europeancountries there is a common practice for patients to offer unsolicited gratuity payments to health care providers, there is ample evidence that a large proportion of the non-formal payments today are not voluntary, and are either requested or expected (Lewis, 2000). Informal payments can constitute a barrier to access health care services and can also distort government efforts to improve equity and efficiency inthe health care system. As seen in Table 5.11, about 30% of those visiting an outpatient care provider in the last four weeks report making a gift to the doctor, nurse, or the health care service. There are large regional differences in the percentage of people reporting making these payments. For instance, in Tirana only 10% of people seeking outpatient care gave a "gift" to the provider, while almost 40% of people in the Coastal region did. Not only are people in Tirana less likely to pay, on average they pay significantly less than people living inother regions. Coastal Central Mountain Tirana 576.4 587.5 722.0 337.5 (84.7) (105.3) (109.1) (52.2) Informal payments in outpatient care represent about 11% of the total expenditure, in contrast to hospital care, where they represent close to 25% o f the total expenditure. Among people giving "gifts" during a hospital stay (60% of all cases), 43% said the gift was requested or expected. In the case o f outpatient care, about 40% o f the people who went to the public ambulatories said they were required or expected, while 25% o f those who went to a nurse mentioned such requirements. The problem o f informal payments in hospital care seems to be a large one (all hospital services are supposedly free), and more important than the problem in outpatient care facilities. The conditions o f the public hospital facilities are in general very poor, some even lacking basic services such as electricity, medicines, and medical equipment.Inthe words o f an Albanian elder women, interviewedinthe study of perceptions on public health care (MOH, 2000): "When my mother was in the hospital, I had to clean up 83 startingfiom the table cloth to the glasses. The toilet conditions were bad. Ifyou don'tpay you are likely to get nothing. You have to give money to almost evevbody, to the door-keeper, the nurse, sanitary and the medic... "The low remuneration o f the personnel has further accentuated this problem, and probably contributes to the proliferation o f informal payments. Box 5.1: Informal paymentsfor health care,peri-urban areas of TiranaandDurres Although my children are too youngfor school, both were immunized and receive regular check-ups at the local State hospital. Gifts of 200 to 500 new Leks are required for simple check-ups, but more complex treatments cost much more. For example, when my last child was born, wepaid IO,000 Leks in gifts and otherfees. -35yearoldmanintheoutskirtsofDurres Normally we need to pay 500 Lek in exchange for treatment. When my oldest daughter was born (in Kukas), the doctor indicated that he expected a gift, but Irefused to pay because the child was a girl and I had ordered a boy! Although all my children were immunized, my two youngest children were not born in a hospital. Female family members helped Drita (wife) with the births. When a family member needs heath care, we go to the State hospital in Durres. -Ramiz,32year oldheadof householdinswampareaof Durres Both my children were born at the State hospital. At each birth my family paid between 5,000 and 10,000 Leks in gifts to the doctors. We usually pay about 500 Leks for regular visits. Iam not pleased with the quality o f the service Ireceive at the hospital, even after giving the doctors gifts. Even after Ipay for the drugs that the doctors prescribe, the drugs are often expired. - Shukri, 25years, head of household, swamp area of Durres Fatjon s daughter is sick with a recurrent cough and she+equently needs to visit the doctor at the clinic. Thedoctors do not askfor money, but if we do not offer a gift, we will constantly be told to come backfor more checks. The doctors simply waste their time until a gift is made. When Angela (grand-daughter) needs to visit the doctor, the other children in the household collect cans and sell them to recycling to raise the moneyfor the doctor's gift. The last time they visited the doctor, theypaid 1,300 Leks worth oj medication on credit. Doctors cared more under communism,or at leastpretended to care. -AbdulhouseholdinBreguILumiareaof Babru,peri-urbanareaof Tirana From Bezhani and Tracy, 2002: Households open interviews in peri-urban areas of Tirana and Durres peri-urban areas. Theseinterview excerpts are not meant to be representativeof the country or the areas where they were done, but an indication ofproblems faced by some of their inhabitants. HouseholdExpenditureson Health On average Albanians spend almost 700 Leks per capita per month on health care, which is about 18% o f the average per capita consumption o f a poor family. However, there are large differences in the amount households spend across consumption quintiles. Households belonging to the poorest 20% o f the consumption distribution on average spend about half what households in the richest quintile spend on health. As percentage o f total consumption, households in the poorest quintile o f the income distribution spend almost 10% o f their per capita monthly consumption on health (almost 13% a proxy o f per capita permanent income i s considered); in contrast, the richest families spend on average around 5% o f their total expenditure on health care. As hospitalizations are rare events, most o f the expenditure on health goes to outpatient care visits. Medicines represent a large share o fthe total health expenditure and, as was mentioned before in the case o f outpatient care, medicines represent close to 55% o f total expenditure, while the percentage is 17% inthe case o f inpatient care. But there are also many people purchasingdrugs without prescription, and thus expenditure in medicines represents a significant share o f total household expenditure. On average, 3.5% o f total household per capita expenditure goes to the purchase o f drugs whose production and distribution i s almost completely in the private sector. Households in the poorest 84 end o f the income distribution on average pays twice as much in medicines (as share of total expenditure) than in households at the richest end ofthe di~tribution.~~ Table 5. 14:Per capita monthly health expenditure and share of health expenditure on total Quintiles I I1 I11 IV V Total Per capitahealth expenditure 449.8 544.4 618.5 691.4 892.9 672.9 Source: Albania LSMS 2002 aTotal health expenditure was added to total per capita household expenditure to compute these shares. However, the consumption quintile distribution does not include health expenditure Includes medicines prescribed in outpatient care services, hospitalizations and non-prescribed medicines 6. CatastrophicHealthExpenditureand Poverty Table 5.15 shows the inequalities of the health system infinancially protecting households against catastrophic health expenditures (expenditures constituting over 10% of total income). On average 26% of people report health expenditures that are equal or above 10% of their total income, and about 9% report health expenditures representing 25% or more of their total income. People inthe poorest quintiles are more likely to incur catastrophic health expenditures than their richest counterparts. Source: Albania LSMS 2002; Total income refers to total consumption excluding health expenditure. 53The LSMS 2001 was collected between April and June of 2002. This survey has a comprehensive module on health, which includes questions on outpatient and inpatient service utilization and expenditure. As many people cannot remember exactly the number of visits they do to certain health services, especially preventive ones, the recall period for most of the LSMS questions related to outpatient visits was four weeks. By doing this, the questionnaire avoids recall errors, but it does not take into account seasonality inhealth care utilization. As such the monthly data presentedhere is only representative of the months when the data was collected. 85 Following Wagstaff and van Doorslaer (2003), a headcount for catastrophic health expenditure, pE,as well as an index that weights this measure by giving more emphasis to the poorest households, are computed. A measure of the catastrophic payment excess, po,which measures how much the health expenditure surpasses the share of income we define as catastrophic, is also calculated. These measures provide summary statistics indicating whether or not the poor are more affected by catastrophic health expenditures, and whether their excess payments are larger than richer households. These indexes confirm that not only are the poorest households more likely to incur catastrophic health expenditure, they are also more likely to have larger excess in health payments over the threshold (the weighted measures WEand WOare larger than the means, pEand po). Threshold PE CE WE >lo% 0.25632 -0.13371 0.290592 >25% 0.08977 -0.17031 0.105063 CO WO >lo% P O 0.021712 -0.03133 0.022392 >25% 0.047343 -0.08104 0.05118 The measures of catastrophic health expenditure tell us about the percentage of people that have health expenditure in excess of a pre-established threshold, but they do not tell us about whether or not the person with such expenditures falls into poverty, or in case she is already poor, she falls deeper into poverty. Following Wagstaff and van Doorslaer (2003), poverty measures before and after health expenditure are calculated, which yield this information. A summary of these measures i s presented in Table 5.17. Pre-payment Post-payment Headcount 25.4 33.9 Normalized gap 5.7 9.9 Extremelypoor Headcount 4.7 10.3 Normalized gap 0.8 2.9 Hospitalization Headcount 25.4 27.5 Out-patient Headcount 25.4 31.9 j4The concentration curve is defined as two times the area between the concentration curve and the line of perfect equality. The concentration curve is constructed by plotting the cumulative share o f the sample along the x-axis (ranked by income) against the cumulative share o f persons exceeding the threshold along the y-axis. 86 Health care payments, especially outpatient health expenditure, induce some people to fall into poverty or deeper into it. While a quarter o f Albanians have per capita incomes below the full poverty line (44,891 Leks per person per month), income per capita o f almost 34% o f the population falls below the poverty line after health expenditure is subtracted from incomes. Similarly, although only 5% o f the population is considered extremely poor before health expenditure is taken into account, almost 10% o f the population falls below the food poverty line after such expenditure i s subtracted. Surprisingly, expenditures on outpatient care are more likely to render an individual poor compared with expenditures on hospitalization. 7. Determinantsof healthcare demand A standard probit analysis was used to study the determinants of the probability that an individual will seek care when ill.This analysis was done separately for children (0-14 years) and for adults that reported visiting a primary health provider when illor injured inthe last 4 weeks before the LSMS interview took place. The results o f this analysis indicate that income and ownership o f a health license significantly increases the probability o f seeking care in both samples. In the adult's sample the education o f the individual was also found to increase the probability o f seeking care. Inthe children's sample the distance to a pharmacy was found to decrease the demand for primary health services, and the distance to a public hospital to increase it. The last two results might be a consequence o f the low quality o f primary health services, as many people chose to bypass them or only used them for obtaining drug prescriptions. Therefore the farther away a pharmacy i s located, the less incentive there is to visit a GP in the primary health center to obtain drug prescriptions. Finally, although the LSMS did not capture outpatient care provided in public hospitals, there is anecdotal evidence suggesting that many people bypass primary health services and go directly to specialists in polyclinics and hospitals to obtain care. Price elasticities o f demand for health care were also computed. Price elasticities decrease with income, and people at the upper end o f the income distribution are less likely to change their health seeking behavior when prices change. The price elasticity o f demand for health care for children i s more elastic than that o f adults, but in general for all income levels and for both samples the price elasticities o f demand is very low, insome cases very close to zero. This contrast to the results found by Gertler and van der Gaag (1990) using data from Peru and Cote d'Ivoire and by Chawla (2001), using data from Nicaragua. However, other studies have found similar results (see Heller, 1982, Akin et al. 1984). 8. ConcludingRemarksand PolicyImplications Analysis o f the LSMS 2002 data shows that only 58% o f people who report an illness or injury seek outpatient care and there are large income and regional inequalities inhealth status and in access to health care services. There are large differences between rural and urban areas as regards access to water and sanitation, and while all urban dwellers have access to a sanitation facility inside their house, only half o f households living in rural areas have access to sanitation facilities. There are also large income differences inthe percentage o f people visiting an outpatient care provider, and while the poor and people livingoutside the Tirana region are more likely to report an illness than the non-poor, they are less likely to seek care when ill.Onthe other hand, residents o fthe Tirana region are less likely to report and illness, but more to seek care when ill.There are large regional differences in inpatient care utilization as well. Better off urban dwellers are more likely to have health insurance license than people in rural areas. Also individuals for whom the State covers insurance contributions are more likely to have insurance than those that contribute themselves. 87 Almost all women receive pre-natal care, and trained personnel attend almost all at birth. However, there are large income and regional differences in the number o f pre-natal consultations and in the type of provider attending the women at birth. Comparatively speaking, however, women in the Mountain areas are the least likely to receive pre-natal care, and receive the least number o f clinical consultations. In contrast, women in Tirana receive two times as many pre-natal consultations than women in other parts o f the country. There are large out-of-pocket expenditures in health care. Expenditure in medicines represents the largest share of these expenditures. Informal payments are widespread, especially in hospital care where 60% o f people reported making under-the-table payments. These payments also represent an important share of total expenditure, accounting for about a quarter o f total expenditure on hospitalization. Health care expenditures represent a significant share o f total household monthly expenditure, and vary with the income level o f the household. Though the poor pay less in absolute terms, they spend a larger share o f their income on health care than the rich. Poor households are not only more likely to experience catastrophic health expenditure, they are also more likely to have larger excesses in health payments over the catastrophic threshold. The results o f the analysis point to three priority areas for policy makers interested in improving access and effectiveness o f health care. First, coverage under health insurance needs to be increased, both in terms o f number o f persons covered as well as number o f services provided under the coverage. Health insurance i s not only important as a source o f funding for the sector; it also provides financial protection for the insured against catastrophic health expenditure. In particular, the poor and vulnerable sections o f the population, who are at risk o f falling into poverty as a result o f a single incidence of illness, need to be covered by health insurance. Indeed, even though the poor and vulnerable are the ones most likely to be covered by health insurance at present, there remain large groups o f the poor and vulnerable that are actually not covered. Further, health insurance only covers the services of the GP or the family doctor, and provides no protection against catastrophic health expenditure. In Tirana, HI1covers all outpatient care and not surprisingly, people inthis region spend the least on outpatient care, on average, compared to people in the rest o f the country. There is clearly a strong case for increase in the number o f services offered. Another area that needs immediate attention i s informal payments, the eradication o f which will have positive effects on access to health services. Informal payments in outpatient care represent about 11% o f the total expenditure, in contrast to hospital care, where they represent close to 25% o f the total expenditure. On average, Albanians spend almost 700 Leks per capita per month on health care, which i s about 18% o f the average per capita consumption o f a poor family. However, there are large differences inthe amount households spend across consumption quintiles. Households belonging to the poorest 20% of the consumption distribution on average spend about half what households in the richest quintile spend on health. As percentage o f total consumption, households in the poorest quintile o f the income distribution spend almost 10% o f their per capita monthly consumption on health. Besides increasing the burden o f out-of-pocket expenditure on health, particularly for the poor, informal payments also undermine all intent to improve quality and efficiency inthe sector. Anecdotal evidence also shows that quality o f care, particularly in outpatient care and particularly outside Tirana, where a large number o f people choose to bypass these services and go directly to hospitals, is sub-standard. Improving the quality o f the outpatient care services i s necessary for enhancing the scope and utilization o f health services, so that they can be used for more than to just obtain prescriptions for drugs. 88 Annex to Chapter V Not Poor Poor Total Infectious diseases 1.9 2.4 2.0 Diseases o fthe blood or bloodproducing organs 38.8 28.7 36.8 Diseases o fthe respiratory organs 8.9 9.9 9.1 Tiimnrs 1.o 1.2 1.o Source: Albania LSMS 2002; Standard errors inparenthesis 89 Table 5.A. 3: Health care utilization across gender, region of residence,poverty status, and whether Source: Albania LSMS 2002 I Total I 69.8 I 58.3 I35.8 I 74.4 I 57.9 Source: Albania LSMS2002 Table 5.A. 5: Pre-natal consultation by type of provider across regions andpoverty status Source: Albania LSMS2002 90 CHAPTERVI: EDUCATION 1. Overview At the start ofthe transition, Albania inheritedan education system that allowed equal access to education services for almost all. However, in the years after the end of the communist regime the country as a whole and the education system in particular have experienced many challenges that have made it difficult for the education system to contribute to the reduction o f poverty and inequality in the country. When the Hoxha regime broke down, Albania suffered a deep depression that affected all sectors, and education was no exception. Though the country has experienced periods of rapid economic growth since then, income per capita today is similar to that of 1989 and government resources have not reached the levels that existed before the start of the transition. This reduction of public resources has directly affected resources earmarkedfor education, which saw its share o f GDP decreasefrom 5% in 1991 to 3% today. Lack o f resources is not the only challenge confronting education in Albania; two periods of generalized destruction o f educational facilities, large internal movements of the population, and the new demands o f the transition economy have also put pressure on the system. The upheavals that followed the end of the communist regime and the collapse of the pyramid schemes in 1997 were combined with a generalized vandalism of public property that did not spare educational facilities; many of these facilities were destroyed and substantial resources are needed to completely rehabilitate them. Further,when restrictions to population movement were lifted, large migration of people from rural to urban areas and from Albania to other countries followed. Internal migration has created strong pressures on urban and peri-urban schools that were not ready to receive the large influx of new students coming from rural areas. The challenge is not only that there are a large number o f children that need to be schooled, but also, as the new migrants come from areas where school quality is lower, that their needs are different from those of city children. Migration has also created challenges in rural areas as well, as many o f the most qualified teachers have left intheir quest to find better payingjobs in the cities, an exodus that has exacerbated the low quality o f rural schools. Finally, the country i s going through a period of large socio-economic transformations as it transitions into a market economy. The market economy demands different skills in the labor force than those needed by the command economy. The education system therefore needs to adapt to these new skills, which also create a new and difficult challenge that needs to be overcome to help in the country's development process, and to help people obtain jobs that will keep them out of poverty. It i s inthis difficult environment that the education system inAlbania has to assure equal access to educational services, especially to basic and upper secondary education, and contribute to breaking the vicious cycle of lack of education and poverty. 2. EducationalAttainment Illiteracy One of the major accomplishments o f the education system in Albania duringthe communist era was the almost complete eradication of illiteracy. Data from the Albanian LSMS 2002 shows that today more than 90% of the population older than 14 can read and write without any difficulty. The great majority of adults who cannot read and write (about 5% of them), or can but with some difficulty, belong to older generations. While only 1% of adults younger than 50 are completely illiterate, about 15% of those older than 50 are illiterate. Among Albanian youth, 98% can read and write without any difficulty," compared to 95% ten years ago.j6 55 Ifthe literates includethose that canreador write with somedifficulty, this figure goesupto 99%. 56 World Development Indicators,World Bank. 91 The illiteracy rate varies with income, gender and place of residence. There are two times as many illiterates among the poor than among the non-poor, and also two times as many illiterate women as there are illiterate men. Similarly, while only 4% of the people living in urban areas are illiterate, about 6% of those in rural areas are illiterate. It is important to note that gender differences have disappeared in younger cohorts, as a result of the efforts of the education system in eliminating the gender gap in educational attainment. Years of schooling On average, Albanians 21 years and older have 8.5 years of schooling (Table 6.1). This average conceals variations across income and place o f residence. Adults living in a poor household on average have two years less o f schooling than those living in a better off household. Similarly, adults living in Tirana have about 3.5 years more of formal education than those living in rural areas. The average years of education among the adult population is relatively low when compared with other countries in the region. For instance, the average years o f schooling among young adults in the same age cohort i s one year less in Albania than in Bosnia and Her~egovina.~~ Source: Albanian LSMS 2002. a 60 basic years of schooling. In the case of -2 Albania, basic education finishes in grade 8, cc 40 s0 20 which explains the sharp drop in attainment 0 after that year in the figure. What is very different in Albania, as compared to other 57 Among adults 15-35years old the averageyears of schoolingamongAlbanians is about 9.5 years while inBiHis about 10.6(World Bank, 2003). 58 See for examplethe case of Kosovo (World Bank, 2001) and Bosnia and Herzegovina(World Bank, 2003) 92 completed. The poor at any level o f schooling complete fewer years, and these differences are small in the first eight years o f education but increase significantly afterwards. There are also large differences between rural and urban areas, and while these differences are small in the basic years o f schooling, they increase at upper secondary school level. Among those who complete primary education, more city dwellers than those in rural areas, both poor and non-poor, continue to secondary school. For instance, while close to 72% o f the urban non-poor start secondary education, only 37% o f the urban poor and less than 20% o f the rural dwellers start secondary education. This is not surprising, as the number o f secondary schools inrural areas has been decreasing over time, indicative o f a supply problem inthese areas. Enrolment rates There are large variations in enrolment rates at all levels o f education across regions and income levels. Enrolment rates for the poor, and especially the extremely poor, are much lower than those o f the non- poor; similarly, enrolment rates in rural areas are much lower than urban areas. These differences in enrolment rates might preserve and even intensify income inequality in the future, as poorer households tend to be headed by individuals with low levels o f education. There is no gender difference in enrolment, with the exception o f the university level where the enrolment rate for women is somewhat higher than for men. There are no major absenteeism problems at any level o f schooling among those enrolled. The percentage o f students enrolled that have not attended classes for more than a month i s less than 2%. However, qualitative data suggest that in the peri-urban areas, children might be either missing classes or are just getting only a few hours (some times even less than 4) o f instruction a day. Unfortunately, the LSMS is not representative at this level, and cannot be used to obtain reliable statistics for peri-urban areas. Table 6. 2: Net enrolment rates acrossgender, regions, and income levels Secondary (14- 17) 38.8% 38.7% 70.8% 59.5% 62.2% 24.9% 46.7% 19.4% 19.8% 38.7% Tertiary (18-22) 7.6% 9.6% 22.7% 16.6% 18.3% 2.6% 11.2% 0.4% 0.0% 8.7% computing these results, non-reporting observations (27%) were omitted from the calculation. Tertiary includes students in university andpostgraduate stu& Pre-school: Enrolments in pre-school have greatly declined during the transition period. While in 1990 more than one in two children between 3 and 5 years o f age were enrolled in pre-school (INSTAT, 2001),59less than one inthree i s enrolled today. To put these numbers in perspective, among all countries inthe Balkans only Bosnia and Herzegovina has a lower enrolment rate at this level6' But not only have the number o f children going to pre-school drastically declined, there are also large differences in enrolment across regions and income levels. The net enrolment rate o f children in poor families is about half that o f children in better o f f households. This i s a worrying trend as pre-school programs help 59 The 1990 data is based on gross enrollment rates. 6o Data from other Balkan countries comes from UNICEF TransMONEE database for the year 1999. 93 diminishpossible disadvantagesthat children from low socio-economic status might have when entering primary education. In the case of Albania, as children living in non-poor households are the ones taking advantage o f these programs, any initial difference intheir favor will probably be exacerbated by the time all children enter into their compulsory education. Fortunately, the compulsory starting age o f basic education i s low (6 years), and this might counterbalance the decline inpre-school enrolments. Box 6.1: Access to schoolin peri-urbanareas of Tirana and Durres Our oldest daughter is currently in third grade. She walks with two other neighborhood children to her school, which is an hour away. Class is@om 12.00 PMto 4.30 P M During the winter (rainy season) she misses school about two weeks per month due to mud andflooding. Overall we are not happy with the quality of schooling their daughter receives. We (Drita) provide extra instruction at home, in the morning, but it's not enough. The other children are still tooyoungfor schooling. - Ramiz and Drita's Household inDurres Swamp Area We are not registered in the Tiranaperi-urban area, so it is difJicult to enroll the children. Our eldest son, Peshopia, 9, is not currently enrolled in school, primarily due to the residency issue, however, we trying to resolve this issue and hoping the boy will be enrolled in the new semester that starts after the holidays. At present, however, none of the children in the householdare enrolled in school. - Household inBregu ILumiarea o fBabru, Tirana's peri-urban area. Education is the most important thing Uor the children], but none of my children are enrolled in school. I wanted to enroll my oldest son (7) but could not afsord the 7,000 Lek worth of fees for books and registration. Theseregistrationfees are necessary, because myfamily is notfiom the area. - Kujtim's Household inBahoris, Tirana's peri-urban area. My 16-year-old sonjinishedprimary school two years ago, but his grades were bad, so he was unable to attend the local secondary school. When myfamily lived in Kukas, my eldest son was a good student, but once he transferred to the school in Bathari, his grades started tofall. My second son (14) is in the 8th grade and receives instruction9om 8:OO A M to 11:OO A M During this time, there are 4 classes - each 30 minutes long. My youngest son (10) is in the 3rd grade and receives instruction9om 1:30 to 5:OOpm. - BersnikHousehold inBahoris 2, Tirana's peri-urban area. Source: Household Interviews inperi-urban areas of Tiranaand Durres (Bezhani and Tracy, 2003) These interview excerpts are not meant to be representative of the entire country or of the areas where they took place, but are indicative ofproblems faced by some of their inhabitants. Basic education: Compulsory basic education in Albania includes four years of primary and four years of lower secondary school. Enrolment rates at this level are almost universal for both boys and girls. The gross enrolment rate at the basic school level is about 100% while the net enrolment rate is about 94%. Although the enrolment rates at the basic education level are high, they hide some income variations that affect mainly the extremely poor. For instance, the enrolment rate of children living in households unable to meet basic food requirements is 5 percentage points lower than that of children living in non-poor households. It should be pointed out that the non-extreme poor have similar enrolment rates as the non- poor, suggesting that policies should target the poorest of the poor to assure that all Albanians receive basic education. Another indicator of the Millennium Development Goals of achieving universal access to primary education by 2015 is primary completion rate. Ifchildren start their school education at the age o f 6 years, they should complete their primary education by the time they are 14 years old. However, data from the Albania LSMS 2002 show that on average only 62% of children of that age have completed the eight years of basic education. This is probably mostly the result o f late enrolment, as many children 14 or even 94 15 years old are still enrolled in primary school. The percentage of children 16 years old that have finished 8 years o f basic education, for instance, is about 95%. Albania Croatia FRMacedonia 1990 2002 1999 1999 Total 79 38.7 Academic 33.1 19.9 22.2 VET 5.6 58.2 43.2 Upper secondary education: Enrolment rates in upper secondary schools are low, and as was the case in pre-schools, there has been a large drop in enrolment during the transition years. For instance, the enrolment rate in secondary school today is only half of the 1990 rate and is also lower than the rates of other countries in the region. Most o f the enrolment at this level i s in academic programs with an enrolment rate of about 33%; in contrast vocational and technical upper secondary schools only have an enrolment of about 6%. As Table 6.3 shows, the situation is very different in other countries in the regions, where most of the secondary enrolment occurs inVET programs. Box 6.2: Determinantsof secondary school enrolment A probit analysis was conducted to study the determinants of secondary school enrolment in both rural and urban areas of Albania. The data used was a sample of secondary school age children (14- 18 years) that had already finished primary school. Results show that in urban areas the only variables that significantly explain enrolment, across all the models estimated,6l were education o f the household head, the education of hisher spouse (variables that were used as proxy for parental education) and the age of the child. The price of notebooks and color pens has a significant negative effect on enrollment. Inrural areas, the main determinants of secondary school enrollment were found to be: (i)the presence o f a secondary school in the community, (ii) the education o fthe spouse of the household head, (iii) capita per household consumption, and (iv) ownership of cattle (probably a wealth effect). Variables that had a negative effect on enrollments rural areas were: (i) the ownership of horses, and in some of the models estimated the ownership of sheep, and (ii)the presence of other secondary school age children in the household. On average households with cattle own only one head, while households that own sheep own about 14 heads, which might require more work than inthe case of cattle. Household ownership o f horses and sheep thus indicates work opportunities for children inside the household, and their ownership therefore increasesthe opportunity cost of sending children to school. Enrolment rates in secondary school are not only low, they also vary largely across income and region of residence. For instance, the net enrolment in secondary education of poor children is less than halfthat of the non-poor. Across regions the inequalities in secondary enrolment are even more striking. While almost 70% of the children o f the relevant age group are enrolled in Tirana, and close to 60% in other urban areas, only a quarter of the children in secondary school age in the rural areas are enrolled. This is due, inpart, to a lower supply of educational facilities inrural areas. Tertiary education: Enrolment in universities in Albania is very low. With a gross enrolment rate o f 11% and a net enrolment rate of about 9%, Albania has the lowest participation in universities in all Europe. Among the transition economies in Europe and Central Asia, only Turkmenistan and Uzbekistan have lower enrolments rates (UNICEF, TransMONEE). At this level there are also large variations across 61 Five models were estimatedfor the urbansample, and 7 for the rural one, 95 income, regions and gender. Only at the university level the enrolments o f boys and girls differ, with girls participating slightly more than boys. Almost all enrolments concentrate among the non-poor students living inurban areas, especially inTirana. Cohort differences The transition period brought many socio-economic challenges that also affected the education system. Children who went to school during the transition saw their family income plummeting at the start of the period and also suffered the reduction in government resources earmarked for education. As shown in Table 6.4, the youngest cohorts, ones who were inthe compulsory school age duringthe transition, have a lower completion rate than the cohorts who finished primary school during the communist regime!* In the youngest cohort, people between 15-19 years, there are children that are still enrolled in school, but even if these individuals are considered as if they have completed 8 years o f basic schooling, the proportion o f people finishinggrade 8 in this cohort will still be less than that o f the cohort o f 25-29 years old. At the start o f the transition, individuals who are now 25 years old or more were 14 years old or older and thus had already finished or were about to finish elementary school. Younger cohorts went to primary school after the end o f the communist regime and experienced the deep depression o f the early 90s and the subsequent upheavals. Cohort Finished grade 8 Enrolled in Secondary 15-19 0.925 0.484 (0.011) (0.023) Source: Albania LSMS2002 Standard errors inparenthesis. There are also large cohort differences in the proportion o f students who continue to secondary school once they have finished 8 years o f compulsory schooling. There i s a large drop (of about 80%) in the proportion o f people continuing to secondary school in the cohort 20-24 years old. However, in the youngest cohort the proportion increases again. Interestingly enough, those 20 to 24 years old today were between 15 to 21 years, near secondary school age, in 1997 the year o f the large upheaval that followed the collapse o f the pyramid schemes, when many schools were vandalized. There is, however, no statistical difference in the proportion o f those that continue to secondary school between the youngest cohort and those between 25 and 45 years. 62There have been large movements of population in Albania during the transition years. The youngest cohort is also the one that has migrated the most. It is possiblethat these differences in school completion might be driven by selective migration where the most educated members o f this cohort leave the country. This will lower the proportion of peoplethat finishedprimary school amongthose staying. 96 3. Reasonsfor low enrolment The main reason for the low enrolment at the pre-school level appears to be the non-availability o f pre- schools. The total number of pre-school programs has decreased considerably in the last 10 years, from 3,426 in 1990 to 2,111 in 1999 (INSTAT, 2002). Many of the pre-school programs were connected to industries were the parents worked, and when those industries closed at the start o f the transition, the pre- schools also closed. Another reason cited mostly by parents inTirana is the preference to keep the child at home.63 Source: Albania LSMS 2002 Inthe case of secondary school, the mainreasons givenfor not enrolling are demand-side. More than one o f every two not enrolled young person did not have any interest in going to school, most because they considered their schooling completed after finishing 8 years o f basic education.64Another reason for not attending secondary school was that of having to work, suggestinga high opportunity cost of going to secondary school. In rural areas, 13% of the children who are out of school are involved in agricultural work. On the supply side, distance to school was mentioned as a reason for not going to school, particularly in rural areas. This is not surprising, since the number of secondary schools has fallen by more than 50% in the last decade. The bulk o f this reduction happened in vocational and technical secondary schools, that fell from 575 vocational and 177 mixed schools in 1990 to only 51 and 9 respectively in 2001. Incontrast, the number o f academic schools increased in the last decade and today there are 25 1 more schools than in 1990 (INSAT 2002). The reduction inthe number of schools affected mostly the rural areas, where a decade ago there were 550 vocational or mixedschools and now there are only 63Why this is the case is not clear. It might be that highunemploymenthas kept many women at home; it might also be a reflection ofthe low perceived quality of the existent programs. 64 More than 90% of those that said that they were not enrolled because they have completedtheir education only hadprimary school diploma. 65INSTAT, as quoted in Berryman, 2000. 97 Source: Albania LSMS 2002 4. Expenditure on Education Public Expenditure With the economic downturn that followed the collapse o f the communist regime, government resources diminished considerably. Today, many years after the depression and after several years o f highgrowth in the country, government expenditure (as percentage o f total GDP) is still only half o f what it used to be. Although the reduction in government expenditure in part reflects the transition to a market'economy, it also reflects the reduction in the government's resource collection capabilities. 1991 2000 2001 2002 Totalpublic spending/GDP 62.2 25.9 27.5 27.9 Spendingon Educationas % of GDP 5 3.2 3.3 3.4 Educationas % of total public spending 8.1 12.3 12 12.1 The reduction in government resources negatively affected government expenditure in education, which fell from 5% o f GDP in 1991 to only about 3% today. Public expenditure in education in Albania i s not only low when compared to what it used to be a decade earlier, but i s also low when comparedto several other countries in the region (Slovak Republic, 4.3% o f GDP in 1999, Czech Republic, 4.69'0, Poland, 4.1%, Hungary, 4.1%. OECD mean, 5.3%; UNICEF, OECD). Private Expenditure The reduction in public spending on education has been balanced by increase in private expenditures, which on average represents about 1.4% o f the total household monthly per capita expenditure (almost 2.5% for households with students). Households in the upper quintiles o f the income distribution pay on average about 2.5 times more than those in the poorest quintile. However, as a percentage o f total monthly expenditure, poorer households pay a similar or even a higher percentage (the difference, however, is not statistically significant) o ftheir total household expenditure. Higher household expenditures inthe basic years of schooling do not reflect a higher cost at this level, but are indicative o f lower enrolments in post-basic education. Likewise, at the basic years of schooling similar household expenditure across quintiles is the result o f a lower share o f school age children and higher expenditure in households at the higher end o f the income distribution. For instance, when looking at all households, regardless o f whether or not they have school age children, they all spend, on average, 98 around 300 Leks a month. However, when only households with school age children are considered, households inthe higher quintile o f the income distribution spend on average more than twice the amount what poor households spend on education. Among households with students, expenditure on education can be a considerable share o f their total expenditure, representing 2.2% o f the total monthly expenditure among the poorest households and 3.5% among the richest. Consumption Quintiles 1 2 3 4 5 Total Source: Albania LSMS 2002 Expenditure in secondary schools and in universities is in general higher than expenditure on the basic levels. Data from the Albania LSMS show that the total monthly individual expenditure for secondary school i s on average about 3.3 times the individual expenditure in basic education. Similarly, the total per student monthly expenditure on higher education is about 7.5 times the per student expenditure on the basic years o f schooling. In both secondary and tertiary educationthere are also large income variations in the total amounts spent. For instance, among those in secondary school and higher education, students belongingto households in the highest income quintile pay on average four times more than their poorest counterparts. Source: Albania LSMS 2002 The effects o f low public expenditure on education are also observed when looking at the percentage o f students that report making payments either in cash or in kind to the school and the teacher in the last academic year. In both basic and secondary schools, more than 40% o f the students report making such payments to schools and teachers. More students from richer quintiles report making payments to schools and teachers compared to the poorer quintiles. Qualitative evidence from peri-urban areas o f Tirana and Durres suggests that many households contribute with construction material and fuel for the repairs and maintenance o f the schools. 99 Box 6.3: Informal paymentsin education My second son(14) is inthe 8th grade ...the youngest son(10) is inthe 3rdgrade ...Imust pay for books, repairs, and a contribution to the salary ofthe guard at the school. - Bersnik's Household in Bahoris 2, Tirana'speri-urban area. Better education for my sons was one of the reasons Imoved my family to Tirana for Peshkopia. Our 12- year-old son attends class from 7AM to 2PM, and i s even instructed in English.Ipay about 300 Leks per month for books and his contribution to the school guard's salary. However, Ipulled my 16-year-old daughter out of school inthe 2nd grade. The reason?We didn't want the girl inschool. - Isuf'shouseholdin Bathori 1,Peri-urban area of Tirana. Household Interviews in Tirana and Durres peri-urban areas (Bezhani and Tracy, 2003). These interview excerpts are not meant to be representative of the entire country or of the areas where they tookplace, but an indication of problemsfaced by some of their inhabitants. Low public expenditure in education i s not only reflected in the large proportion o f households contributing to the schools and teachers, but also in the proportion of students receiving private tutoring, as the quality o f education has been negatively affected Figure6.2:Percentage of students over time. Proportionally more students in secondary recievingtutoring school receive tutoring than those in primary school, and there are large variations in the proportion of students `0.35 I I receiving tutoring across income quintiles. While a third o f the secondary school students living in households belonging to the richer 20% of the income distribution I receive tutoring, less than 1% of those living in the poorest 20% of households do. The large proportion of students receiving private tutoring 1 2 3 4 5 i s worrying as it can potentially affect equity outcomes. It ConsumptionQuintiles ~ suggests that there are quality problems in the instruction students receive in public schools and as higher income children are more likely to afford private tutoring, they are also more likely to obtain a better education and thus betterjob market opportunities. The majority of the students receiving private lessons have as tutors their own teachers or other teachers in their school. This might potentially compromise the instruction given in the classroom, as teachers might force the students to take private lessons, or the students mightfeel compelled to do so. Inprimary schools at least 50% of the students paying for private lessons received those lessons from their own teacher or teachers intheir school; inthe case of secondary school this percentage i s higherthan 60%. 5. Educationand Labor Markets Table 6.10 shows the relationship between education and labor force participation o f old and new generations. Albanians with low levels o f education, i.e., with primary school diploma or less, are proportionally more represented among the sub-employed agricultural workers, seasonal workers, laid off workers, and the population economically inactive, than those with higher levels of education. Incontrast, Albanians with a university degree are proportionally more representedamong the employed, regardless o f their age or gender. 100 Table 6. 10:Laborforce status of Albanian men across educationlevels,percentages IWorkinginagr. <15hrs/OtherWorkhot employedjLaid Off/SeasonalIDiscouragedJInactive/Total I ITotal 0.04 0.06 I 0.00 I 0.00 I 0.01 I 0.13 I 1.00 I Source: LSMS 2001 Note: Percentages add up to 100vertically. The labor force participation of Albanians with secondary school diplomas deserves a separate analysis. Secondary school enrollment decreased considerably in the years after the end of communism. Younger Albanians are abandoning school earlier than their older counterparts. As seen in Table 6.11, the share o f young Albanians menwith secondary general education among the unemployed, the laid off workers and the discouraged workers is larger than their share inthe total population, suggesting that secondary school training does not improve labor market opportunities for graduates. Similarly, among older generations, the shares of men with secondary vocational or technical school among the unemployed, the seasonal workers, the discouraged, and the inactive, are larger than their share inthe total population. Inthe case of those with secondary general education, their share among the laid off, seasonal and discouraged workers i s larger than their own share in the total population. This discouraging performance of secondary school graduates can explain in part the lower enrollments in these programs today, as they seem to offer very few benefits interms o f labor market performance. Source: Albania LSMS2002 101 Vocational training is a case in point. These schools used to train the labor force for the state owned industrial complex, the labor force for the collective farms, and it also used to train middle level technicians and to prepare them for technical University studies.66As many state owned enterprises have closed or downsized and new private enterprises are emerging and inneed of different skills, many of the old VET programs lost their appeal and many have closed. This in part explains why older generations, those who went to secondary VET programs duringthe previous regime have fared so poorly inthe labor market, and why the enrollment rates o f younger generations in these programs are considerably lower than before. For sure, vocational education programs are needed, but the instruction neededtoday i s very different than the instruction these programs traditionally offer. The labor market today requires skills that allow workers to adapt to changing environments andnewjobs or tasks. Source LSMS 200I Note: standard errors inparenthesis Education not only affects the probability o f having a job, it also affects earnings. Table 6.12 shows the average monthly real wages of Albanians 18 years and older. Among the young there i s no statistical difference between the earnings of those with primary school and those with secondary school, both general and vocational, which further points to the low value added of secondary school education, as it does not improve earnings of those secondary school graduates when compared to primary school graduates. This also explains the low levels of enrollment in secondary school, as the benefits of continuing to secondary school are not high when compared to the labor market benefits of being a primary school graduate. Among the older generations there are also no significant differences between the earnings ofthose with secondary school degrees and those with a primary school degree. Butthere is a significant difference between the earnings o f those with less than primary and those with primary education and among those with secondary education and those with auniversity degree. In a country like Albania, characterized by large migration, the effects of education on labor market opportunities is better understood if job opportunities outside the country are also considered. For instance, secondary education could bring benefits such as to better prepare young Albanians to work abroad or to look for job opportunities outside their place of residence. Although there is no reliable information on Albanians living outside the country, LSMS 2002 collected information on temporary and permanent migration. Among young men, there was no difference in the percentage o f people that migrated internally across education levels. Similarly, among young men, there was no statistical difference between people with primary and secondary education in the proportion that lived abroad for 66EuropeanTraining Foundation: Albania Labor Market-VocationalEducationand Training Assessment, Torino, April 2001. 102 more than 3 months. However, this evidence i s not conclusive, since information on those living abroad at the time o fthe interview is not available. 6. ConcludingRemarksand PolicyImplications Analysis o f the LSMS 2002 data shows large income and regional inequalities in access to education facilities and education outcomes. While Albanians 21 years and older have 8.5 years o f schooling on average, there are large variations across income and place o f residence, and an adult living in a poor household on average has two years less o f schooling compared to one living in a better o f f household. Similarly, adults living in Tirana have about 3.5 years more o f formal education than those living in rural areas. There are large variations as well inthe total years o f completed schooling across income and place o f residence, and even though the percentage o f those who finish basic education i s high, there i s a sharp decrease in the number o f students who continue schooling beyond basic education. There are also large differences between rural and urban areas, and while these differences are small in the basic years o f schooling, they increase at upper secondary school level. There are large variations in enrolment rates at all levels o f education across regions and income levels. Enrolment rates for the poor, and especially the extremely poor, are much lower than those o f the non- poor; similarly, enrolment rates in rural areas are much lower than urban areas. Enrolments in pre-school have greatly declined during the transition period, from more than one in two children between 3 and 5 years o f age in 1990 to less than one in three in 2002. The main reason for the low enrolment at the pre- school level appears to be the non-availability o f pre-schools, the number o f which has fallen by one-third inthe last decade. Enrolment rates inupper secondary schools are low, and have dropped duringthe transition years to about halfo fthe 1990 levels. Enrolment rates in secondary school also vary largely across income and region o f residence, and the net enrolment in secondary education o f poor children is less than half that o f the non- poor. Across regions, while almost 70% o f the children o f the relevant age group are enrolled in Tirana, and close to 60% in other urban areas, only a quarter o f the children in secondary school age in the rural areas are enrolled. The main reason for not enrolling in secondary school i s that most considered their schooling completed after finishing 8 years o f basic education. Another reason for not attending secondary school i s that students need to work after completing basic education. Public spending on education has fallen considerably in the past decade, to levels almost half o f what it was in 1990. The reduction in public spending on education has been balanced by increase in private expenditures, which on average represents about 1.4% o f the total household monthly per capita expenditure (almost 2.5% for households with students). Households in the upper quintiles o f the income distribution pay on average about 2.5 times more than those in the poorest quintile. However, as a percentage o f total monthly expenditure, poorer households pay a similar or even a higher percentage (the difference, however, i s not statistically significant) o f their total household expenditure. Informal payments are widespread, and in both basic and secondary schools, more than 40% o f the students report making such payments to schools and teachers. Interms o f the relationship between education and employment, Albanians with primary school diploma or less are proportionally more represented among the sub-employed agricultural workers, seasonal workers, laid o f f workers, and the population economically inactive, than those with higher levels o f education. Incontrast, Albanians with a university degree are proportionally more represented among the employed, regardless o f their age or gender. Education also affects earnings, and there i s a significant difference between the earnings o f those with less than primary and those with primary education and among those with secondary education and those with a university degree. However, there is no statistical 103 difference between the earnings of those with primary school and those with secondary school, both general and vocational, which further points to the low value-added of secondary school education. The results o f the analysis point to two priority areas for policy makers interested in improving access and effectiveness o f education. First and foremost, secondary school enrollments need to be increased, and quality o f education at this level improved. Secondary schools should impart market-relevant skills and training, emphasizing flexibility, adaptability and analytical skills over rigidly drawn curriculum. New vocational programs, in line with potential demand for skilled labor, need to be introduced to prepare students for job-market. New secondary school programs are especially needed in rural areas, where the lack of these programs i s one of the main causes of low enrollment. Poor quality o f education and low teacher salaries have caused a surge in private tutoring given by the same teachers. Private tutoring compromises the education received and equality in education outcomes as only those who can afford these classes will take them, and thus would receive better schooling. The fact that a large percentage of the students take these classes, and that most o f the tutors are teachers in the students' school, is indicative of serious incentive problems in the education sector, and this has to be corrected in order to improve equity and effectiveness inthe financing and delivery of education. 104 CHAPTER VII: COPINGWITH POVERTY-THE ROLE OF SOCIALPROTECTION PROGRAMS 1. Introduction Social protection in Albania today consists o f three major groups o f public programs, plus private transfers between and within families. The public programs consist o f a comprehensive social insurance system, social safety net programs, and labor market policies. These are discussed in detail below. Social Insurance System Albania enacted a new Law on Social Insurance in 1993, which created a mandatory, publicly managed, contributory system with universal coverage. All economically active individuals, including those self- employed in agriculture (farmers), are required to contribute. The Law provides for comprehensive protection against income loss due to old age, disability, loss o f a primary earner, death, unemployment, general sickness, occupational injury and work related illnesses, childbirth and maternity. There are approximately 20 different types o f benefits provided to the insured. Different groups o f economically active people are provided different types o f benefits. One pension program covers employees. Another covers the self-employed, but in practice this encompasses distinct schemes for urban and rural self- employed workers. Pensions are o f a defined-benefit type. The two-tier pension formula for employees includes a basic pension and an earnings-related part. The latter is equal to one percent o f the pension calculation base for each year o f insurance. The base is calculated as an average o f gross earnings on which social insurance contributions have been paid for the individual for three consecutive years out o f the last 10 prior to retirement. The base i s adjusted annually for the changes in the average contribution paid by all contributors. So pensions depend both on years of contributory employment, the income level on which contributions were based, and workers' average contributions in each year. The 1993 law tightened eligibility criteria for retirement, gradually raising the retirement age to 60 for men and 55 for women. In 2002, the Government initiated a further gradual increase o f five years in retirement ages for both men and women. It also raised the ceiling on contributory income, lowered payroll contribution rates by four percentage points, and strengthened mechanisms for enforcing collections compliance. Social Safety Net The social safety net (social assistance) in Albania consists o f three cash social assistance programs and a program o f social care services for orphaned, disabled, and elderly people. The social assistance programs are: a targeted poverty benefit in cash (called ndihme ekonomike or economic aid); a regular monthly allowance to those disabled since childhood; and price compensation paid to pensioners and their families. Ndihme Ekonomike This is the main cash social assistance program in Albania. It provides a means-tested cash benefit for eligible families with little or no earned income. The program i s structured through conditional grants from the national government that local governments are to administer using nationally specified eligibility and benefit rules. Larger local government areas are established as municipalities. Smaller local government areas are administered as communes. Local governments receive and process applications, including through a mandatory visit to each applicant family to check its eligibility. The final decision on whether to grant assistance i s then made by municipal and communal councils, which also set the benefit amount. The amount o f assistance awarded to a family i s to be based on a national Government decision dating from November 1996. This set the benefit ceiling for the recipient family at 2.5 times the unemployment benefit. The base benefit (awarded to each eligible family) cannot be higher than 95 percent o f the individual unemployment benefit. Subject to the overall ceiling, aged or disabled family members attract an additional amount o f 95 percent o f the individual unemployment benefit; people 105 below working age attract 25 percent o f the base amount; and people o f working age attract 20 percent. The program identifies categorically eligible families that satisfy various selection screens designed to exclude families on the basis o f family composition (e.g., no economically dependent family members) or ownership o f assets that imply wealth or income-generating capacity. For eligible families, the program specifies a maximum allowable benefit for families with no other income. For families with limited private income, a partial benefit may be payable. The partial benefit i s reduced from the maximum level to offset all private income (including imputed income) that the family generates through home production (e.g., from land or garden plots) or that it receives from outside employment. In practice, payment o f benefits is also limited by the program budget available. For some time, anecdotal reports have suggested that some local governments have disregarded the national eligibility criteria and substituted decision-making criteria o f their own to determine who will receive ndihme ekonomike. However, the Ministry o f Labor and Social Affairs recently undertook a review o f the eligibility o f all recipients o f ndihme ekonomike and, as a result, substantially reduced the current caseload. The Ministry now reports that evidence from the control department at State Social Services i s that payments to ineligible families constitute a low percentage o ftotal payments. Regular monthly allowance to the disabled The regular monthly allowance in cash for the disabled since childhood i s granted to a person certified as disabled prior to 21 years o f age (24, if attending university). It i s a flat rate benefit, amounting to 70 percent of the minimum wage. The amount o f allowance is doubled if the disabled person attends secondary school and tripled ifhe or she attends university. Since 1995, the benefit has been administered regional GASS offices performing general monitoring and supervision function^.^' inthe same way as the ndihme ekonomike program: at the local community level and with the central and Price compensation to pensioners and their families Compensation to pensioners and their dependents for bread, electricity, and food price increases was introduced in conjunction with the elimination o f price controls and subsidies, to cushion the impact and help maintain the consensus in favor o f economic reform.68Since 1994, the price compensation program has either been discontinued or consolidated into wages (for government sector employees) for all eligible categories, except for pensioners. In 2001, the Government reviewed the continued relevance o f price compensation for pensioners. As a result, it limited the category o f pensioners eligible for these compensation payments to those with dependents who are unemployed, children up to age 18, or students up to 25 who are not in employment-categories chosen because o f their consistency with the European Convention on Social Security. This measure, effective from January 1, 2002, was expected to reduce pension outlays in 2002 by an estimated 250 million Leks. Price compensation i s financed from the central budget. It is administered by the Social Insurance Institute, which identifies beneficiaries (pensioners and their dependents) and pays the benefit along with the pension payment. For most o f the pensioners the compensation i s part o ftheir regular pensionpayment. 67Prior to 1995, the Social InsuranceInstituteadministeredthe disability allowance program. Food price compensation was introduced in conjunction with the August 1992 liberalization of most consumer goods prices, including all food other than bread; bread price compensation was introduced in July 1993, when breadprices were raisedto cost-covering levels; electricity compensationwas introduced inApril 1994, with a sharp increase in electricity price. In 2000, the monthly amount of electricity compensation was Lek 350 per dwelling, while bread compensationwas Lek 150 per pensionerper month. The amount of compensationgiven to pensioners' dependents-that is, each family member supportedby the pensioner, irrespective of age - was Lek 414 per month. According to the Social Insurance Institutethere were 0.6 dependantsper pensioner.Rural pensioners were entitled only to electricity compensation. 106 Social Care Services Albania has a system o f residential social care services for orphaned, disabled, and aged people in need o f assistance. It i s the administered by the General Administration on Social Services (GASS). Albania's Social Services Delivery Project, which i s supported by an IDA credit, is seeking to strengthen, broaden, and restructure the current system o f services. Box 7.1: Social Protection Spending, 1998 The following table presents social protection spending data, as a percentage o f GDP, for OECD countries in 1998. It shows spending on social insurance, social assistance, and social protection (the sum o f the two). Social insurance and social assistance spending are shown both as a percentage o f GDP and o f overall social protection spending. Social Social Social Social Social Insurance Insurance Assistance Assistance Protection (% Labor Market The National Employment Service (NES) has responsibility for: (i)a national network o f labor market offices (currently limited to urban areas); (ii)the National Labor Inspectorate, which seeks primarily to 107 enforce work safety standards and to require employers o f informal labor to contribute to the Social Insurance Fund; (iii) passive labor market programs (administration o f unemployment insurance benefits, as well as income support for defined categories o f people made unemployed by mass privatization o f state-owned enterprises); (iv) active labor market programs including (a) the Employment Promotion Program, which offers either a year o f temporary employment or hands-on training to unemployedjob- seekers, as well as support for creation o f relevant institutions, and (b) the Public Works Program until its fundingwas terminated in2000; and (v) professional training. These programs were developed in part through the Albanian Labor Market Development Project, which received World Bank support. One legacy o f this project is a valuable capacity to assemble statistical data for the formal portion o f the labor market. A new NES strategy was approved thisyear through a tripartite agreement among the Government and employer and employee representatives. The strategy aims to reform the NES at all levels and it proposes creation o f an Employment Fundwithinthe framework of the Social Insurance Institute. Ina broad sense, the current safety net has replacedan inheritedsocialprotection strategy inwhich people were assured o f guaranteed employment at low wages on collective farms and in government enterprises, while a system o f administered pricing within a closed economy kept prices for consumer staples correspondingly low. The current social safety net system therefore represents a major change that was designed to complement movement to a market-oriented system with unemployment risks and prices for labor and consumer goods that increasingly reflect market forces. As in other E C A countries, the prerequisites for higher living standards and increased economic opportunity also brought with them new social and economic risks for individuals and families. Albania is still adjusting to these change through two major streams o f migration - one internal and the other external - as people seek to move closer to employment opportunities. One stream has involved a continuing flow o f families from former collective farms that are not economically viable to rapidly growing peri-urban areas around Tirana, Durres, and other cities in the country's central zone. These non-viable farms are located mainly in the most northern parts o f the country and, to a lesser extent, in the most southern parts as well. The second migration stream has been o f young men moving both to neighboring countries and others further afield in search o f jobs or higher incomes. To date, much o f their earnings have returned to Albania as remittances that support their families. Approximately one-third o f Albanian males aged 20 - 35 now live abroad in this way; and one-third o f all transfer income received by Albanian households comes as remittances from family members abroad. 2. Impact of Albania's Cash SocialProtectionPrograms Tables 7.1 and 7.2 highlight the importance o f public and private income transfers within the structure o f household income. In 2002, averaged across all households, public transfers reported in the survey represented 21.3 percent o f household income. Private transfers represented a further 14.4 percent. Together, therefore, these reported transfers provided more that one Lek out o f every three. That made them a more important source o f household income, on average, than either earnings from employment (3 1.5 percent) or agriculture (25.0 percent). For the population as a whole, transfer income sources that were quantitatively important as a share o f household income (Le., those contributing more than 2 percent) consisted o f all categories o f pensions [urban old-age pensions (10.5 percent), rural old-age pensions (3.8 percent), and other pensions (2.4 percent)], ndihme ekonomike (3.1 percent), and remittances from relatives abroad (12.8 percent). In short, public pensions and private transfers from Albanian relatives abroad dominated, with ndihme ekonomike contributing substantially less. 108 Sources of Income Total Urban Rural Urban Quintiles Rural Quintiles 1 1 2 1 3 1 4 1 5 1 1 2 1 3 1 4 1 5 ~~ Employment 11740 17904 6574 13153 17580 14884 16564 24849 4232 7260 7866 6935 7242 Wage 11392 17289 6451 12983 17101 14343 15947 23808 4165 7160 7732 6798 7033 Inkindwage 45 81 15 57 79 80 75 104 7 27 10 29 3 Bonus 302 534 108 114 400 461 542 936 61 73 124 108 207 Other Income 297 526 105 164 67 1186 345 663 22 5 241 73 243 Total Income 28833 41298 28833 22636 29644 38175 33732 70480 16485 23914 30964 33897 46276 109 Table 7. 2: Structureof Albanian HouseholdIncome, by Source,Locationof Residence, andIncome The most distinctive feature o f Albanian transfer patterns, relative to those o f other transition countries in the region, i s the key role o f private remittances from abroad, mainly in the form o f cash. On average, one Lek o f household income out o f every eight came from this source. However, among the 22 percent o f households receiving such transfers, their average size was 13,600 Leks per month. That is equivalent to 47 percent of average monthly household income for the country. Moreover, private transfers from abroad 110 were much more likely to be received by non-poor households (23.5 percent) than by poor households (16.8 percent) or extremely poor households (10.1 percent). Similarly, when these transfers were received, they were substantially larger on average for non-poor households (14,400 Leks per month) than for poor households (8,900 Leks per month) or extremely poor households (6,100 Leks per month). Exploration is warranted o f the extent to which receipt o f private transfers from abroad was itself sufficient to move families out o f poverty andor into the higher income quintiles. Pensions, ndihme ekonomike, and transfers from Albanian relatives both at home and abroad were somewhat more important sources for urban households than rural ones, while only rural pensions were more important in rural thanurban areas. The most direct evidence on the distributional impact o f these transfers comes from the Living Standards Measurement Survey for 2002. Of all households, 25.4 percent were below the poverty line, Le., the poverty line fell within the second quintile o f households. However, these poor households were significantly more likely than non-poor households to receive transfer income. For example, poor families comprised 53 percent o f the households receiving public transfers and 46 percent o f households receiving private transfers (Table 7.3). Table 7. 3: Distribution of transfer income across householdsreceiving transfer incomefrom that Of households receiving Ndihme Ekonomike, 76 percent were poor and only 24 percent were non-poor. Moreover, 43 percent o f all households receiving Ndihme Ekonomike payments were in the lowest quintile, which contains four out o f every five poor households. Another 24 percent were in the second quintile, which contains the remaining poor households. Only 12 percent o f Ndihme Ekonomike recipients were in the two top quintiles. However, poverty did not assure receipt o f ndihme Ekonomike. Only one- quarter o f all poor families received ndihme Ekonomike; and those families represented three-quarters o f all families receiving the benefit. Fully 69 percent o f households in severe poverty and 75 percent o f all poor households did not receive Ndihme Ekonomike. Many o f these, o f course, might have been ineligible. 111 Pensions and remittances from Albanian relatives are the largest sources o f transfer income. Urban old- age pensions are disproportionately received by the two upper quintiles; but the 25 percent o f households that are poor represent 37 percent o f the recipients o f urban pensions. Rural old-age pensions are disproportionately received by the two lower quintiles and the poor. It i s notable that poor households receive a more than proportional share o f each different form o f transfer income examined. Important Transfer Income Sources by Povertv Status. Transfer income can also be analyzed by source to show which sources made major and minor contributions to household income for the various income groups. Not surprisingly, ndihme ekonomike was the single most important source of transfer income for poor and extremely poor households, for whom it provided 25.8 percent and 31.7 percent respectively o f all transfer income. (However, ndihme ekonomike also represented 8.8 percent o f transfer income received by non-poor households.) For poor and extremely poor households, rural and urban old-age pensions were the two next most important sources o f transfer income. Urban pensions were proportionally much more important for non-poor households than for poor and extremely poor households, whereas rural pensions were considerably more important for poor and extremely poor households than for non-poor households. A similar divergence was apparent in relation to remittances from family members. Remittances from abroad were considerably more important for non-poor households than for poor and extremely poor households, whereas the poorer household groups benefited more from remittances from family members inAlbania. Most Common Sources o f Transfers by Poverty Status. Two households out of every three received assistance from at least one source o f transfers (Table 7.4). Just over one-quarter (28 percent) o f all households received private transfers. Twice that number received public transfers from some source. Poor households were slightly more likely that non-poor households to receive a transfer from any source or from any public source, but they were less likely to receive a private transfer. The most commonly received public transfers were urban pensions (25 percent o f households), private remittances from abroad (22 percent), rural pensions (19 percent), ndihme Ekonomike (11percent), and other pensions (9 percent). All sources o ftransfer income were received disproportionatelyby poor andvery poor households. Table 7. 4: IncidenceAmongHouseholds of Receipt of Transfer Income, by Quintile, Poverty Status, households within cat 112 All I Any Private Transfers 28.1 24.8 22.0 29.0 30.9 33.9 28.9 24.1 21.5 28.6 All /Any Transfers 66.9 71.6 66.9 66.0 67.1 62.9 65.9 71.3 65.9 67.8 Figure7.1: Distribution of Households Transfer Income (from all sources, by source of income) I Unemployment Family Abroad- 14% Ndihme Ekonomike 23 3% 12 3% Private, Non-Family Family inAlbania Rural Old-Age \ Pension Social Care 17.5% 0.3% Figure 7.2: Distribution of Non-Poor Household's Transfer Income (from all sources, by source of income) Unemployment Ndihme 1.4% Family Abroad- ,'Ekonomike 8.8% 250% ' / Private, Non- Family - 0.7% \ UrbanOld-Age Family in A l b a n i a 1 4.3% Other Pensions-f, 7.8% Rural Old-Age Pension Social Care 16.3% 0.4% 113 Figure 7.3: Distribution of Poor Household's Transfer Income(from all sources, by source of income) Unemployment FamilyAbroad Benefit 16.6% - '\ 1.4% Private, Non- , Ndihme Rural Old-Age Pension 1 '\ ,_Other Public J 22.0% Social Care 0.9% 0.2% Figure 7.4: Distribution of Extremely Poor Households' Transfer Income from All Sources, by Source of Income Family Abroad Private, Non- 9.4% Unemployment Family 1.1% 2.5% ?, \ Family in Albania Maternity Benefit Rural Old-Aae Urban Old-Age Pension - 1 \ 1 Pension 25.5% 11.7% Social Care- \ 0.1% Lother Public 0.7% Source. Tabulationsby WorldBank sta#j?om Albanian Living Standards MeasurementSurvey,2002 114 Sources o f Largest Transfer Amounts. Among those households receiving transfers from particular sources, foreign remittances provided the largest average annual benefits (13, 600 Leks) followed by urban pensions (8,900 Leks), domestic remittances (5,100 Leks) and other pensions (4,900 Leks). Ndihme ekonomike provided an average o f only 2,400 Leks. Social care, maternity benefit, other public transfers, and unemployment benefit all provided benefits o f the order o f 5,000 Leks per month, but assisted very few households. For Ndihme Ekonomike, average benefits were higher for extremely poor households (2,700 Leks) than for all poor households (2,500 Leks), while non-poor households received less still (2,300 Leks). This progressive pattern was not repeated for any other transfer source. Exactly the reverse pattern emerged, however, for urban pensions, "other social transfers," and private remittances from abroad. Estimating the Impact o f Public Transfers on the Incidence o f Poverty. Insight into how Albania's social safety net functions can be amplified by analyzing the impact o f the various public transfer programs, singly andjointly, on the incidence o f both poverty and extreme poverty inAlbania. However, it is important to understand how that impact is estimated. Albania's poverty line is defined in terms o f the value o f a household's consumption, rather than its income. This is primarily because the value o f goods and services that a family consumes i s thought to be conceptually superior, as a measure o f its living standard, to the amount the family receives in income. It i s also partly because measurement o f income is subject to significant underreporting in transition countries. What i s observed and measured in surveys like the LSMS are families' "post-transfer" consumption levels, which reflect their receipt o f income from all sources including transfers. To assess the impact o f transfer income on families' living standards and poverty rates, policy analysts must therefore estimate how different consumption levels would have been ifthere had been no (or fewer) sources o f transfer income available. Inother words, analysts must estimate what families' "pre-transfer" consumption levels would have been. The simplest way to do so i s to assume that a family, which received 100 Leks intransfer income, spent it all to consume extra goods and services. Inthat case, if the family had not received the 100 Leks, its pre- transfer consumption level would have been 100 Leks lower than its "post-transfer" consumption level. So, if this family's observed consumption level were below the poverty line, then the entire 100 Leks would have contributed to a reduction in its degree o f poverty. However, it i s possible - indeed, it is likely - that, inthe absence o f the 100 Leks, the family might have changed its behavior in ways that permitted some o f the last 100 Leks o f its actual consumption level still to be undertaken. For example, a family member might have worked longer in an informal job, or spent more time growing food in a family vegetable plot, or borrowed food or clothing or received additional cash from relatives, or even withdrawn a child from school to help expand home production activities or to work as a street vendor. To the extent that such adjustments would have occurred, the loss o f 100 Leks intransfer income would have led to a reduction of less than 100 Leks in the family's consumption. The extent to which a typical family could and would make such adjustments in practice i s unclear. It i s also unclear to what extent families in different sets o f circumstances would vary in their capacity to do so. Nevertheless, it should be recognized that the impact o f social safety nets on poverty i s likely to be overestimated if it is assumed that the receipt o f an additional 100 Leks o f transfer income will add 100 Leks to the family's consumption level. Accordingly, alternative calculations have been undertaken to show how the impact o f the safety net on the incidence o f poverty and extreme poverty would vary according to the proportion o f additional transfer income that results in additional consumption. For all Albanian households as a group, the overall poverty rate o f 25.4 percent in 2002 would have been 11.8 percentage points higher - that is, 37.2 percent - if there had been no public transfer programs (Table 7.5). Relative to this higher, implied level o f poverty before public transfers, the public safety net reduced the number o f poor households by almost one-third. The major contributions came from urban, rural, and other pensions and ndihme ekonomike (in that order). The other public transfers programs made modest additional contributions. Similarly, the extreme poverty rate would have been 6.46 percentage 115 points higher than its actual rate o f 4.74 percent higher -that is, 11.2percent. Thus, relative to this higher, implied level o f extreme poverty before public transfers, the public safety net reduced the number o f extremely poor households by 58 percent (Table 7.6). The individual programs had the same relative importance in bringing about this reduction as they did for the overall poverty rate. These estimates assume that all transfers contributed 100 percent to additional con~umption.~~ Table 7. 5: Estimated impact of income transfers onpost-transfer rates of poverty, if specified sources of transfer income had not existed, by source of transfer and assumedshare of transfers used to increase consumption,for all households in 2002 u % of impliedpre-transferpoverty rate 31.68 1.07 18.59 7.65 6.11 1.05 0.73 0.50 1.23 %points 8.03 0.73 3.61 1.38 1.29 0.23 0.05 0.13 0.18 % of impliedpre-transferpoverty rate 24.02 2.80 12.46 5.16 4.84 0.90 0.20 0.50 0.70 ImpactAssuming Consumption Increasedby 50 percent of Transfers 1% of implied pre-transferpoverty ratel 7.97 I I 0.57 I 4.23 I 0.98 I 1.02 I 0.55 I 0.00 I 0.06 I 0.20 I Source. Analysis of Albania 2002 Living Standards Measurement Survey 69Additional analyses are planned to examine the corresponding impacts of foreign and domestic remittances, all private transfers, and all public and private transfers considered jointly. Given the large known size of private remittances, it is reasonable to expect that they will add significantly to the measuredeffectiveness of the Albanian social safety net. Accordingly, when all public and private transfers are consideredtogether, the total gap between pre-transfer and post-transfer poverty rates is likely to emerge as much larger; and the proportion of that implied overall pre-transfer poverty level that is offset by the total system o f public and private transfers i s likely to be higher still. Further analyses will repeat this kind of analysis for various subsets of Albanian families to see for which groups of families the current safety net has providedmore or less assistancethan for the average family. I16 Table 7. 6:Estimated impact of income transfers onpost-transfer rates of extremepoverty, if specified sources of transfer income had not existed, by source of transfer and assumedshare of transfers used to increase consumption,for all households in 2002 Baseline Rate o f I n I Extreme Poverty Source of Income Transfers and EstimatedImpact All -Ndihme Unemp-Mater- BaselinePost-Transfer Extreme Poverty Rate Source.Analysis of Albania 2002 Living StandardsMeasurementSurvey 3. NdihmeEkonomike: Key Issues Social assistance programs have multiple, but competing, goals: to combat poverty through adequate benefits for people whose resources fall short o f socially recognizedneeds standards; to support economic efficiency, including favorable work incentives for participants; to maintain fairness between recipients and non-recipients o f assistance; and to keep program costs affordable." Balancing benefit adequacy, work incentives, equity, and program affordability unavoidably involves tradeoffs among these goals. The trade-offs are often seen as implying "target-efficiency" as an intermediate goal. However, target efficiency does not promote economically efficient resource use; usually, it makes this worse. So target efficiency i s only o f several criteria o f good performance. Public social assistance programs supplement private transfers from relatives, friends and charities. Public transfer programs can potentially displace private transfers to some degree. ' OWork incentives depend on the proportion of any additional earnings that a worker can keep without an offsetting reductioninthe benefit (or, equivalently, on the rateat which benefits are withdrawn when private earnings increase). 117 Box 7.2: Cashtransfer programs A policy dilemma arises from the inherently conflicting objectives o f cash transfer programs: high levels o f assistance to the poorest groups, strong incentives for them to work, and low overall expenditures on social assistance. They conflict because they involve three policy variables: the maximum benefit, the effective marginal tax rate, and the ceiling income level for benefit eligibility; but only two o f these can be set independently. Unfortunately, raising the basic benefit amount for someone without other income will also raisetotal program spending, unless higher spendingis offset by faster withdrawal o f benefits as private income increases. But that would worsen work incentives! More generally, policy changes to achieve any o f the three goals more completely would worsen attainment o f at least one o f the other two. So policy makers face unavoidable tradeoffs among the competing objectives, requiring them to strike an ootimal balance among benefit adeauacv, work incentives, and urogram affordability. Four main questions about the ndihme ekonomike program concern social assistance policy makers. First, how reliably is it reaching its intended recipients? Second, do families receiving ndihme ekonomike face appropriate incentives to work? Third, to what extent do families seeking or receiving assistance from the program under-report their earnings or other income? Fourth, are changes needed in the current structure o f the program? Evidence concerning how well cash assistance i s targeted is mixed. It shows that three-quarters o f the families that this program reaches are poor, but the program nevertheless misses three-quarters o f all poor families, including two-thirds o f all extremely poor families. Administrative accuracy within the program probably also varies geographically by local government area. Evaluatingthe ndihme ekonomike program, however, also requires knowing what proportion o f poor families i s ineligible and, more particularly, whether those poor families that do not receive benefits are indeed ineligible. Ifso, does the exclusion o f many poor families result legitimately from the national program criteria, or illegitimately from differences between national legislative criteria and local-governments' de facto administrative criteria? Alternatively, i s the problem underlying nonreceipt o f ndihme ekonomike by poor families simply that the program's scale is inadequate? Inthat case, would raising the program's budget substantially increase its coverage o f poor households - especially those that are extremely poor? In2002, the Ministry o f Labor and Social Protection began an ongoing administrative review o f the eligibility o f existing recipients o f ndihme ekonomike, which by February 2003 had cut the total caseload by 8,000 families (5.6 percent). Overall, the primary challenge concerning this program i s to identify and then overcome obstacles to more effective targeting and more comprehensive coverage o fpoor households. Box 7.3: Target Efficiency Target efficiency refers to the proportion o f a program's expenditures that the intended target group receives. For programs to alleviate poverty, it refers to the share o f benefits that go to people below the poverty line, rather than "leaking" to individuals who were not poor to begin with or "spilling over" by raising some poor people's post-transfer incomes to levels above the poverty line. The concept also reveals the tradeoffs among competing objectives o f benefit adequacy, work incentives, and program affordability. For example, a social assistance program that i s perfectly target efficient, with no leakage and no spillover, would also imply no incentive to work (or to work more) over the range o f income below the poverty line. Fiscal analysts often presume that, in a world o f limited budgets, a more target- efficient program is a better program. However, making a social assistance program more target efficient will usually make the allocation o f resources less economically efficient by eliminating incentives for DarticiDants to increase their work effort. Balancing the desire to provide adequate income support with a concern to maintain adequate work incentives i s an additional challenge for all social assistance programs. Providing public benefits reduces 118 the direct pressure on families to pursue all available sources o f private income or assistance. Moreover, denying benefits if any earnings are received, or reducing benefits to fully offset increases in earnings, can also reduce correspondingly the incentive to work. Albania has defined a partial benefit ceiling under the ndihme ekonomike program for low-income families that have land or other potentially productive resources. This i s equivalent to imputing to such families income from those resources. Families migrating to peri-urban areas near the cities are affected by this provision. A further possible response to this inherent dilemmawould be to reduce benefits by only a proportiono f any increase in earnings. Such a reduction would impose an implicit tax rate on earnings up to the ceiling level o f income for receipt o f any benefits. (Lithuania used this approach inthe early 1990s; and the U S food-stamps program also does so.) Other countries have responded by imposing work requirements on some or all recipients as a condition for receiving benefits. (Highly visible US welfare reforms in 1996 combinedthis approach with job training, subsidies for child and dependent care, retention o f health insurance assistance benefits, and a ceiling on duration o f benefits'l). Albanian policy makers have been actively considering pilots to impose increased work requirements on recipients o f ndihme ekonomike. They have delayed related decisions partly until the current review o f safety net programs i s complete, allowing a more thorough diagnosis and assessment o f program effectiveness, and partly to allow adequate time to define and gather baseline data that could be used to assess the impact o f any such pilot programs. Inall countries with large informal and illegal economies, non-reportingand under-reportingo f income is a major problem for administrators o f income-conditioned social assistance. This i s a serious challenge for Albanian policy makers as well. These challenges have reinforced the interest o f social assistance policy makers in work-testing pilots. It has also led some local governments to publishthe list o f families receiving ndihme ekonomike and the amounts they receive. The current study i s partly intended to improve the information base for deciding whether changes in program structure are warranted, and the appropriate nature o f any such changes. As well as possible changes to strengthen work incentives, policy options o f interest might also focus on processes: (a) to determine more systematically whether, in some localities, the allocation o f ndihme ekonomike significantly disregards intended targeting criteria; and (b) to intervene in those cases to achieve appropriate targeting practices. Current national authority to undertake or strengthen these processes would need to be retained as changes in national and local functions are introduced under the Government's new decentralization policy, under which it i s transferring additional responsibilities to local governments. Thus, these policy development processes would need to be reconciled. Discussions over the PRSC-2 adjustment credits envisaged that this will be accomplished through the preparation o f a policy paper through a transparent, participatory and technically supported dialogue among MoF, MoLSA, M o L G and associations o f local governments that defines respective authority o f central and local governments for social assistance. It shows that most, but not all, recipients are poor; most poor households are not receiving this benefit (possibly because they are ineligible); and anecdotal accounts emphasize the divergence between national eligibility criteria for benefits and the criteria that local governments adopt when administering the program. Concerning social assistance, a 1998 assessment, based on 1996 data, found that 36 percent o f all program expenditures went to families whose per capita incomes placed them in the lowest decile, and three-quarters went to families in the bottom four deciles. However, half o f all families in the lowest decile received no assistance; the allocation o f central government funds to local governments was less well targeted than the local governments' allocation o f funds to families; and local governments that received relatively higher levels o f funding used this primarily to extend coverage to less needy families rather than to raise benefits to the poorest families. 71 United States Congress: Personal Responsibility and Work Opportunity Reconciliation Act of 1996, which replacedthe oldwelfare system with a new program, Temporary Assistance to Needy Families (TANF), 119 The 2002 household survey found that 31 percent o f extremely poor households and 25 percent o f all poor households received this benefit, compared with only 8 percent of non-poor households. It also found that ndihme ekonomike was receivedby 24 percent o f households inthe poorest quintile, 13 percent o f those inthe second quintile, and 11 percent o f those in the third quintile.Thus, it remains the case that the benefit goes disproportionately to families who are poor or extremely poor, it goes as well to a significant number o f non-poor households, and it i s not received by many families who are poor (but who might be ineligible for other reasons). Administrative accuracy within the program probably also varies geographically by local government area. The key policy requirements now are processes (a) to determine when the allocation o f ndihme ekonomike systematically disregards intended targeting criteria and (b) to intervene to achieve appropriate targeting practices. Those requirements sit uneasily with the Government's new decentralization policy, under which it i s transferring additional responsibilities to local governments. In2002, the Ministry of Labor and Social Protection began an ongoing administrative review o f the eligibility o f existing recipients o f ndihme ekonomike, which by February 2003 had cut the total caseload by 8,000 families (5.6 percent). 120 CHAPTERVIII: COPINGWITH POVERTY-ROLE OF INTERNAL AND EXTERNAL MIGRATION 1. Overview Migration i s perhaps the single most important political, social, and economic phenomena in post- communist Albania, and has been a dominating fact of everyday life in the last decade. Since 1990, approximately one-fifth of ths total population o f the country has left and is living abroad, and Albania has experienced large-scale movements of population from rural to urban areas. Migration, whether rural to urban or international to Italy or Greece, is often the most important livelihood coping strategy employed by rural households, and serves as an important escape valve for unemployment and other economic difficulties brought on by the transition to a market economy. Official estimates make remittances the largest source of foreign exchange, greater than the combined value of exports and foreign direct investmentand currently constituting 14 percent of GDP. Migration i s a double-edged sword, however, and while in the past migration out o f rural areas was perceived as a potential solution to the problem of rural poverty in Albania,72 migration is also increasingly seen as contributing to social dislocation, agricultural labor shortages and rapid deterioration inthe provision of social services in urban areas. Furthermore, the large-scale legal and illegal migration of Albanians to other parts of Europe constitutes an important political and social challenge to the governments and citizens of the European Union. Stricter migration laws, increasing costs and associated risks also make migration less of a viable solution, particularly for the less-endowed households. And finally, although migration and the resulting remittances are likely to continue to be an indispensable ingredient in the Albanian recipe for rural development, the need and necessity to devise more appropriate, sustainable strategies to lift rural households out of poverty and promote the country's growth cannot be over-emphasized. Using evidence on the incidence and patterns of migration derived from the 2002 Living Standards Measurement Survey (LSMS) and the 2001 Population and Housing Census (PHC), this chapter documents the demographic transition underway in Albania and analyzes decisions of households to migrate. Section 2 briefly reviews the history of migration in Albania, discusses the principal factors affecting international migration, and quantifies the importance of migration and remittances in the income and livelihood strategies of rural households. Section 3 provides a descriptive account of internal and international migration, and presents the incidence, duration, and destination of migration. Section 4 examines remittances, with focus on the country source of remittances and the characteristics of households receivingremittances, individual remitters and non-remitters. Section 5 concludes. 2. Background Albania has undergone two great migratory waves in the modern period. The earliest Albanian migrants to Italy in 1448 were soldiers provided to the King of Naples by Skanderbeg, the military commander o f the Albanian Alliance of nobles and the national hero of Albania, in order to put down an internal rebellion. Followingthe death of Skanderbeg in 1468, who had lead the resistance to the expansion of the Ottoman Empire into Albania, until the first years of the 16`h century, approximately one-fourth of the total population of Albania fled from the Ottoman expansion. Most migrated to Southern Italy, where they founded many present day towns inPuglia, Molise, and Calabria (Piperno, 2002). ''Albania: Growing out of Poverty, The World Bank, 1997 121 The second great migratory wave began in 1990 and is still underway. Triggered by the fall of the communist government o f Enver Hoxha, which had governed Albania since end of World War 11,the end of controls on internal and external migration and the unraveling of the centrally planned economy unleashed a demographic shift as individuals and households migrated to the cities and/or left the country. The initial political instability, social unrest, and economic transition associated with the change in government led to the largest surge o f migration in recent times, with an estimated 300,000 individuals leaving the country during 1991-92, primarily to Greece and Italy (Piperno, 2002; Pastore, 1998). Stabilization of the political and economic situation after 1992 reduced and stabilized migratory flows as well. Inflationdropped to less then 10 percent in 1995 from a highof 226 percent in 1992, unemployment fell from 28 percent to 12 percent, and real GDP growth rate rebounded from -7.2 percent to approximately 9 percent annually from 1993 to 1996. The chaos of the original migration shock gave way to decreased but sustained illegal migration as well-organized smugglers started to illegally transport potential migrants across the channel to Italy and over the mountains to Greece (Pastore, 1998). Remittances immediately became a crucial component of the Albanian economy, with private transfers quickly dwarfing export earnings and composing an increasing share of GDP. The collapse of a national pyramid scheme in late 1996 sparked another peak in international migration. The pyramid scheme had its origin in a weak formal credit system and a thriving informal market unregulated by the government and fuelled in large part by remittances. At its highpoint, over 2 million deposits were made in the pyramid scheme, representingover half of 1996 GDP, as people sold houses, livestock and other assets in order to invest on the promise of receiving a 40 percent monthly return on investment.The collapse began on November 19, 1996, and took four months to complete, bringingdown the government and triggeringriots in which over 2,000 people were killed. The country fell into anarchy as the Army and police lost control, armories were looted and foreign nationals were evacuated [Jarvis, 19991. Tens of thousands of Albanians fled the country, starting first from the Vlore region where the first riots erupted, and then spreading throughout the country until March of 1997. Many were repatriated, and a multinational force led by Italy helped restore order and prevent a larger exodus [Pastore, 19981. Beyondthe traumatic political and social impact of this crisis, the economic consequenceswere relatively short lived. While inflation rose and GDP fell 7 percent in 1997, the economy started recovering in 1998. The return of political stability and economic growth again servedto stabilize the migratory outflow, and remains so today. Four types of migration can be discerned since 1990. First, a comparison o f the 1989 and 2001 census shows large-scale migration from rural to urban areas and from villages and small cities to bigger cities, particularly to Tirana. Second, short term international migration (for periods o f days, weeks, or months) is very common, almost exclusively to Greece, particularly from those regions bordering Greece. Third, there is the most common type of long-term international migration, to Greece and Italy as well as other countries of the European Union. And finally, legal long-term international migration to the USA and Canadahas started inrecent times. 73 According to official numbers, by the year 2000 Albanians composed 10 percent of the 1.388 million immigrants with residence permits in Italy, making it the largest source of immigrants from a single country. Approximately 63 percent of these have work permits as well. Family reunification permits accounted for 26 percent of all residence permits, and Albanians were again the biggest group (OECD, 2002). Of the 271,517 permits granted in 2000, 31,185 (over 11 percent) went to Albanians. Based on 73Given the nature o f the LSMS and the relatively small incidence o f migration to North America, there is insufficient data to analyze this type o f migration. 122 these numbers, and including the adjustment suggested by CARITAS, there were probably 170,000 legal Albanian residents inItaly alone in 2000. The numbers of Albanians in Greece appear to be much higher. Two large regularization programs in 1998 and 2001 led to a total of 720,000 applications, of which Albanians represented approximately 60 percent, or 430,000.74The March, 2001 Census shows 655,000 foreign residents, although critics contend that the number is more likely between 800,000 and 1 million (OECD, 2002). Considering that 720,000 immigrants had applied for legal residence and that by the end o f 2001, 585,000 immigrants had work permits,the critics' estimates are more likely to becorrect. Thus, by 2001 the number of Albanians legally residing inGreece and Italy was approximately 600,000, or about one fifth o f the current Albanian p~pulation.~' This does not include the number of Albanians illegally residingin Greece and Italy. This also does not include the number of Albanians legally residing in other countries of Europe or North American, but that number is likely to be only in the tens of thousands. Beyond the two big "push" migration spikes in 1990 and 1997 caused by political and social instability and economic crisis, continued poverty and unemployment as Albania transitions to a market economy serve as constant push factors for migration. Approximately 25 percent o f Albanians, and 30 percent of rural Albanians, live in poverty. Interms ofjobs, the public sector has shrunk to less then one-fourth its size in 1990, while the private sector has only slowly compensated for the loss in state jobs. Growth in employment in agricultural seems to be more a reflection o f refuge rather than of productivity. Unemployment rates have remained in double digits since 1992, and only in 2001 did real wages recover to the pre-crisis level o f 1995. Poor access to basic services is another push factor, particularly in rural areas. Less then half of rural households have access to running water inside or outside their dwelling, only 40 percent have a toilet inside their dwelling, and only 14 percent o f all Albanians receive electricity continuously (WB and INSTAT, 2003). Pull factors have also been important. Exposure to Italian television during the communist period helped transform that country into the Shangri-La of potential Albanian migrants in the early 1990s. Beyond the allure o f wealth and the Italian lifestyle projected through television, significant wage and wealth differentials between Albanian and its European Union neighbors have been obvious attractions. In2001, for instance, Per capita GDP in Italy was 16 times higher and the Greek 9 times higher than per capita GDP inAlbania. At the same time, some factors have reduced the attraction o f migrating abroad, including attacks and stigmatism in the Italian and Greek media; the tendency to blame immigrants for crime and other social ills; European increasing unease with migration, which has led to increasingly restrictive immigration policies; and the physical danger inherent in illegal migration. This is in part balanced by the relative success of Albanian migrants inItaly and Greece in obtaining legal status inthose countries. Finally, other migration processes are at work in Albania. First, Albania was a major recipient of migration during the 1999 Kosovo crisis, eventually receiving up to 450,000 refugees, and to a much lesser extent during the Macedonia conflict in 2001 (Piperno, 2002). Second, Albania serves as an 74 OECD (2002) reports that 65 percent of the 370,000 applications in 1998were from Albanians.This decreased in absolutenumbersto 351,000 applications in2001, but the percentageis not known. Assumingthe share in2001 is 55 percent, the average of the two years would be around 60 percent.Approximately 300,000 illegalimmigrantsin 2001 were also reported,but no informationwas providedas to their origin. 75 Basedon the latest PopulationandHousingCensus of Albania, the populationat April 2001 wasjust in excess of 3.069 millions. 123 important transit country for illegal immigration from South East and East Asia (Piperno, 2002). Third, Albania also serves as an important link in the trafficking of women for prostitution. Large numbers of women are taken from Eastern European countries - as well as Albania - and sold or forced to work as sex workers in Italy and other European countries.76 3. Types of Migration (i)Internaldemographic mobility Table 8.1 presents information on households classified according to having all, some or no members born in the municipality where they currently live. In all, 12 percent of households are composed of individuals born in other municipalities. This index o f incoming migration is much higher in Tirana (28 percent o f households) and cities in the coastal areas (20 percent), but almost non-existent in rural mountain areas (1 percent). I I I I Householdsize I 3.38 I 4.64 I 4.12 I 4.28 Average adult education (years) I 8.46 1 8.78 I 8.51 I 8.64 Percent of householdswhich participatein - dependent work 0.49 0.49 0.39 0.45 - agriculturalbusiness 0.24 0.57 0.56 0.52 - non-agriculturalbusiness 0.10 0.15 0.16 0.15 Number of observations 429 1773 1397 3599 Per-capitaconsumption (lekdmonth) 8520 7662 7818 7801 Povertvheadcount 0.21 0.26 0.25 0.25 Overall, 40 percent o f households have all members born in the same municipality where they currently live. This figure is higher for mountain rural areas and lower for Tirana and in the cities in the coastal stratum. At the same time, only 17 percent of the families in urban mountain areas are composed of local people, suggesting a considerable rural-urban migration within the mountain region. Not surprisingly, 76 See the Human Rights Watch, www.hnv.org, and InternationalCatholic Migration Commission, www.icmc.net, websites for information. 124 households entirely composed by individuals born in other municipalities are less likely to be involved in agriculture and more likely to be involved inwage work. These households are also smaller in size. As far as household heads are concerned, two-thirds of all household heads are currently living in the same place where they were born, while 22 percent moved to the place where they currently live before 1990 and 12 percent after 1990. Overall, 34 percent of household heads live in a municipality different than the one where they were born. Among the regions, household heads have moved the most in Tirana (60 percent), inthe cities on the coastal stratum (53 percent) and inthe urban areas inthe mountain region (69 percent). Figure 8.1: Internal mobility of household heads 100% 80% 60% 40% 20% 0% Tirana Coast, urban Coast, rural Center, urban Center, rural Mountain, Mountain, Total 1 urban rural moved before 1990 0 moved in 1990or a f t 4 I The figures indicate a strong incoming migrationtowards Tirana after 1990 of 29 percent compared to the national average of 12 percent after 1990, and 31 percent compared to the national average of 22 percent before 1990, and peaking after 1995. Internal migration towards central cities (excluding Tirana), on the contrary, slowed down after 1990 to half the national average compared to levels before 1990 that were 50 percent higher than the average. Clearly, Tirana i s the urban destination of choice for most migrants. ehold heads in the 1990s Tirana is the mainpullarea, with a net balance of 133 (142-9) incoming household heads, followed by the Coast region (net balance of 66 incoming heads). The Mountain and the Center regions, on the other hand, are strong push areas, with net balances o f 115 and 84 outgoing household heads, respectively. 125 Quite interestingly, the most important flow o f internal migrants is from Center (rural) to Center (urban), followed by Center to Tirana. The data point towards a very mobile population, with an increasing concentration o f migrants towards Tirana. Other cities, which in the past have been the destination o f choice for many internal migrants, are instead experiencing reverse trends and inflows in the late 1990s significantly lower than in the previous years. Another interesting finding i s the apparent dynamism in terms o f rural-urban migration within the more remote regions in the NorthEast o f the country, which appears to support the impression that the poorer rural households from the Mountain region may often lack the economic means to attempt migration to more distant, opportunity-rich places. Comparison of the 2001 PHC data and the 1989 census data permits disaggregation o f the migratory flows to the level o f districts. For this purpose, the districts are classified as "expelling" and "absorbing" districts, and two indexes are created: (i) for Fig8.2 expelling districts: share o f population living in each district in 1989 who no longer live there in 200177; and (ii)for absorbing districts: share o f population living inthe district in 2001 who did not live there in 1989.78 The three main expelling districts are all located in the Northern mountain region (Tropoje, Puke and Kukes), with less than two third of residents in 1989 still living there in 2001 (darker shade in the first map). In absolute terms, three Northern districts (Diber, Kukes and Puke) are the three main expellers in the 12 years between censuses. Among the primary destinations are, o f course, Tirana and Durres, but also smaller urban hubs such as Lushnje, Fier and Vlore. Among the absorbing district, Tirana is the main magnet, with only 69 percent o f current residents reporting living there in 1989 (darkest shade in the second map). Inabsolute terms, the district o f Tirana in the last 12 years has been the internal Fig. 8.3 migrant destination o f choice in almost one half In-migration of the cases, and has received close to 160,000 bydistrict new residents. Further disaggregation to the level of mini-municipalities shows that the largest inflow o f migrants in the past decade has been towards mini-municipalities 11 and 6 in Tirana, representing close to one half o f the population currently residing in those locations. In other words, the resident population o f mini- municipalities 6 and 11 has almost doubled inthe 12 years between the two censuses. These two peri-urban areas o f Tirana are also the two poorest mini-municipalities, with poverty "A '*Peopleborn lower value of either index indicates higher levels of between 1989 and 2001 were accountedfc 126 incidence well above the city average. Other mini-municipalities with high in-migration are numbers 4 and 7. The district of Durresis the distant second largest destination for internal migrants.Almost 72 percent o f current residents report living there in 1989 (also, darkest shade in second map), and the total inflow of immigrants during the same period numbered over 50,000. Although not large in absolute numbers, during the same period, the district of Mallakaster has the third highest share of current residents who were not living there in 1989, and is one of the few Albanian districts, which in the past decade has experienced a net inflow of migrants. The other main net absorbers, in term of the absolute number of migrants, are the districts of Lushnje (over 14,000), Lezhe, Fier and Vlore (all with around 10,000 migrants each). Regarding net population changes, the top six districts in terms of net inflows are all in the Center and Coastal regions (darker shade in map). Besides Tirana and Durres, there are Lezhe, Lushnje and Vlore. Interms of net outflows (lighter shade in map below), the North Eastern Migrant net inflow region mountain region dominate, with Kukes, Puke, Diber, Mat and Tropoje alone experiencing a total net outflow of almost 100,000 individuals over the past decade. Not surprisingly, most of the flow out of the districts comes from rural areas. Whether measured in relative or absolute terms, the main expellingrural areas are all in districts in the Northern region. In the past decade, close to 75,000 people have moved out o f rural areas in Diber, Kukes and Mat alone. Migration from rural areas is, of course, to Tirana and -36468..,2Jo, Durres, but also to urban areas inthe same districts (in Mat, over 36 percent of people moving from rural areas stayed in 3287.155165 the district). Both urban and rural areas o f the districts o f Tirana and Durresremain inabsolute terms the main destinations among Albanians. Over the past decade, the municipality of Tirana alone has attracted more than 100,000 internal migrants. One inthree of these new residents come from 5 districts inthe North of the country (Diber, Tropoje, Kukes, Mat and Korce). Another 10 percentcome from urban areas inthe center o fthe country (Berat, Lushnje,Fier and Elbasan), while another 11 percent have moved from ruralareas inthe same district. Based on the index, the rural areas in the district o f Kurbin ranks second among the absorbing districts, between Tirana and Durres. Although small in magnitude, in relative terms, the urban areas of Mat, Mallakaster and Has have experienced the largest inflows inthe 1990s. Virtually all the migration to these provincial urban centers comes from nearby ruralareas. Three conclusions on internal mobility are particularly important. First, Albania is undergoing a fundamental demographic transition, from rural to urban, and from the North Eastern Mountains to the districts of the Coast and Tirana. Second, Tirana i s by far the principal destination of internal migrants, a process that appears to have accelerated inthe second half o f the 1990s. Third, there is evidence of a local rural-urban migration within the North Eastern region. Whether these urban centers in the north are only interludesto more distant and lucrative types o f migration for these households remains unclear, although the large numbersof migrants from these provincial urban centers inthe Northto Tirana seems to suggest 127 that many, indeed, move on. Giventhe internal migration to the urban areas of the Coast, as well as local rural-urban migration, creating economic opportunities in urban areas beyond Tirana could help in re- establishing a more sustainable rural-urban flow, as Tirana alone i s struggling to accommodate and service the increasing number of migrants flocking to its poorest neighbourhoods. (ii)Temporarymigration 2001-2002: all types of temporary migration The LSMS data also provides information on adults who spent at least one month outside the household during last 12 month^.'^ Out of a total of over 10,000 adults, more than 4 percent have left home temporarily in the past 12 months, and about 10 percent of households have at least one adult who has beenabsent for at least one month inthe year prior the survey.Most of these types o f short-term migrants are heads of the household (53 percent) or their children (36 percent). On average, they spent 3.5 months out of the household (3.75 ifmales, 2.44 iffemales). On average, they are younger than the average for all adults (35 years old against 41), most likely to be male (85 percent), and slightly more educated(9.3 years o f schooling as against the average of 8.5). Householdsize Percentofhouseholdswhich participatein --- non-agriculturalbusiness dependentwork 0.46 0.35 0.45 agriculturalbusiness 0.50 0.71 0.52 0.15 0.11 0.15 Per-capitaconsumption(lekshonth) 7890 7069 7801 Povertyheadcount 0.25 0.33 0.25 As Table 8.3 shows, the phenomenon oftemporary migration is more concentrated inruralareas, with the highest shares in the rural mountain regions (17 percent o f households) and rural central areas (14 percent), and lowest in Tirana (4 percent) and in urban coastal areas (6 percent). Households with members away for part of the year are slightly larger, are more likely to work in agriculture, and less likely to be involved in wage work and non-agricultural businesses. These households are also characterized by lower levels o f consumption and higher poverty incidence (33 percent). In sum, ''No informationis available on the destination, however 128 households resorting to short-term migration are poorer, live predominately in rural areas, and are involved primarily in agricultural production. This suggests that households utilize short-term migration to complement low incomes from agriculture. 1997-2001:temporary external migration The LSMS collected migration histories of individuals who had been abroad for at least 3 months at any point in time since 1997. The data show that among adults, 7 percent (corresponding to 18 percent of households) reported having spent at least three months abroad since 1997.*' The temporary external migrants are younger than other adults (36 years old compared to the average of 41) and much more likely to be male (83 percent versus 44 percent among other adults). O f the total number of migrants, 43 percent were abroad in only one year from 1997 to 2001, while 15 percent were abroad every year during the same period. Incidenceofmigration is highest among people currently aged25 to 34 (14 percent) and 35 to 44 (10 percent) and is lower than the average among young adults (aged 15 to 24) and persons over 45. Migrants are typically household heads (51 percent) or their children (36 percent). Figure 8.5: Temporary externalmigration I 100% 80% 60% 40% 20% - 0% Tirana Coast, Coast, Center, Center, urban rural urban rural Mountain, Mountain, urban rural To Greece 0 To Europe 0 Other ~ Temporary external migrants come from households characterized by slightly bigger size, lower participation in wage labor and higher participation in agricultural work. The share of households with wage labor decreases with an increasing number of migration episodes, from 0.44 (1 episode) to 0.33 (5 episodes). In contrast, the share o f households with agricultural activities increases with an increasing number of migration episodes - from 0.62 (1 episode) to 0.74 (5 episodes). No significant difference exists between households with a temporary external migrant and households with no temporary external migrant interms of average education, per-capita consumption and incidence of poverty (though poverty incidence is lower among households with higher numbers of migration episodes). Greece is by far the most important destination o f temporary migration (about 80 percent o f cases), followed by Italy, whose importance increased considerably between 1997 and 2001, from 12 percent to 22 percent of cases, (during which time the percentage of migrants to Greece decreasedfrom 84 percent to 74 percent). The reported frequencies underestimate the full extent o ftemporary external migration as they exclude individuals who have beenabroad for less than 3 months continuously at any given time. 129 Overall, migrants to closer destinations like Greece and Italy are younger than non-migrant adults. Women are again less likely to migrate than men, but they are relatively more numerous among those who migrate to further countries, as they are more likely to be part o f entire migrant households. On average, migrants to Europe are slightly more educated than their counterparts going to Greece or the non-migrants, and also include a higher share o f women. The incidence o f poverty among migrants to Europe is significantly lower (17 percent). Migrants to Greece more often come from rural areas in the North and Center o f the country, while migration to Italy and the rest o f Europe is more predominant from both urban and rural coastal areas. As expected, the largest overall outflows o f temporary external migrants are from the rural areas in the Mountain and Central regions. Tirana municipality has the lowest prevalence o f temporary migrants going abroad. (iii)Permanentmigration Further evidence on migration comes from the fertility section o f the LSMS, in which women were asked about their children, when they left home and where they live. The data show that of the 6,065 children of household members who left the family at some point and are still alive at the time o f the survey, an astonishing 35 percent o f them live abroad. The share o f children living abroad i s even higher for children who left the household after 1990 (48 percent vs. 15 percent o f children who left the household before 1990 - not reported). In other words, almost one-half o f all children who have left home since 1990 are living abroad. The main destinations for these permanent migrants are Greece (49 percent) and Italy (35 percent). At the individual level, these migrants tend to be younger than those who remained in Albania, and are predominantly male. The large majority (65 percent) of those remaining in Albania are women. Education levels are similar, though migrants living in Europe and other places (primarily North America) are more educated. Table 8. 4: Individual andHousehold characteristicsof children who left home I Albania I Greece I Europe I Other I Total Individualcharacteristics Numberof observations Percentage 0.65 0.17 0.15 0.03 Age (years) 38.1 31.6 30.12 32.83 35.62 Percentfemales 0.64 0.38 0.3 1 0.43 0.54 Education (years of schooling) 9.68 9.83 10.04 11.63 9.81 Number of observations 1343 Percentof households(furthestmigration) 0.35 0.27 0.31 0.07 1.oo Household size 4.53 3.57 3.43 3.47 3.86 Average adult education(years of schooling) 7.73 7.32 7.8 9.73 7.79 Per-carita consumDtion 7234 9155 9385 9867 8472 Percenthouseholds independent work 0.36 0.28 0.32 0.50 0.34 Percenthouseholdsinagriculture 0.62 0.66 0.53 0.32 0.58 Percenthouseholds innon-agricultural business 0.13 0.09 0.12 0.22 0.12 Povertv headcount 0.3 1 0.17 0.12 0.18 0.21 130 Looking at household level characteristics, we observe that in 35 percent o f cases, all children who left home currently live in Albania.*l These households are bigger in size (4.5 members comparedto 3.5) and more likely to be poor than the families with some children abroad (31 percent versus 12-17 percent, depending on the country o f migration). Average household education levels are similar, with the exception o f the 'Other' category, where households have a significantly higher level o f education. Overall, however, the education levels o f households with children who have left home are lower then those households where the children have not left. In terms o f labor activities, the share o f households with wage labor and non-agricultural business is higher the further is the migration point, while the share of households in agriculture decreases. The poverty headcount is lower for all categories o f migrating households (12-18 percent), as compared to households with children inAlbania (3 1percent). Based on the above, the number o f Albanian citizens currently living abroad is estimated to be about 458,000. This i s a lower-bound estimate, as it excludes entire households that moved abroad before 2001, the children o f the children (i.e. women's grandchildren) and children of women who are no longer alive. Furthermore, it does not consider people who live in Albania - whether still in the households or not - who move abroad only temporarily every year. 4. Remittances Remittances play a very important role in the income strategy o f Albanian households. On average, remittances represent 13 percent of total income among Albanian households (14 percent for non-poor, 8 percent for poor). The share is higher among urban dwellers (16 percent) compared with rural (11 percent), likely reflecting differences in patterns of migration F i g u r e 8.6: R e m i t t a n c e s a n d GDP 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Data from the LSMS show that over the 12 months prior the survey, 28 percent o f households received some form of private transfer from either individuals or institutions. The vast majority o f individual remitters are either children o f the household head (55 percent) and siblings (25 percent, including brothers and sisters in law). Approximately 80 percent o f remitters live abroad, mainly in Greece (42 *'At the household level, if one household had multiple destinations, we placed that household in the category correspondingto the furthest destination, inthis order: Greece, Europe and Other. 131 percent o f those abroad) and Italy (40 percent). Virtually all remitters (98 percent) have been living abroad since after 1990 (38 percent since before 1997, 60 percent since 1997). Of those who left after 1990, 90 percent have been living abroad continuously. More than half (55 percent) o f recipient households live in rural areas. Individuals live in Albania? Yes No Total Percent 0.21 0.79 1.oo N(unweighted) 283 1070 1353 Locationof remitters Percent Percent Percent Amount Per-capita Albania 1.00 0.21 38,151 11,700 Greece 0.42 0.33 70,095 24,036 Italy1 0.40 0'32 1121,211 1 37,782 I Other EuroDel II 0.10 II 0.08 Tirana1 I 0.07 I 92.039 I 33.426 1 Remittancesused for Percent Amount Per-capita Purchase o f food andbasic necessities 0.33 74.249 24.438 If migrated after 1990,lived abroad continuously Percent Yes 0.90 An examination of migration patterns suggests that the individuals who go to Greece are generally from the rural and poorer households, while those who go Italy and other European destinations are generally from better-off families. At the same time, an examination o f remittances patterns shows that remittances from Greece are less compared to remittances from Italy. This would seem to indicate a migration- poverty trap, in that the less-endowed people from rural areas go to Greece, remit less and their families 132 stay poor, while the marginally better off from Coast and urban areas go to Italy, remit more and get better. Remittances are usually sent for no specific use or destination (37 percent of cases). When households report a reason, the main one is to purchase food and basic necessities (33 percent), followed by investment and purchase o f durable goods (12 percent) and for medical expenses (9 percent). The average amount o f remittances is 88,600 Leksg2per year (about US$670). The amount is higher when destined to households in rural areas (104,400 Leks compared to 68,900 Leks), and increases with the distance of remitters (higher when coming from abroad, in particular if from Italy or further). Remittances are of significantly higher amount when specifically destined to fund investment (business or dwelling renovation) or purchaseo f durable goods (above 220,000 Leks per year). Households receiving remittances are smaller in size (3.8 members versus 4.5), characterized by a lower level of adults' education (7.9 years o f schooling versus 8.9), and are less likely to be involved in dependent labor and in non-agricultural businesses. They are characterized by higher consumption levels and lower incidence of poverty. In the 12 months prior to the survey, Albanian families received about US$227 million worth of remittances. O f these, US$203 million came from individuals living abroad, of which US$193 million were in cash. This is an underestimation of total remittances from abroad, as it does not take into account the amount brought back by household memberswho have migrated inthe course ofthe year. To estimate this residual part, we estimate the total amount of months adult household membershave spent abroad in 2001 and multiply that by an expected earning for each of the main destination^.'^ Assuming a savings rate of 50 percent, total remittances of temporary migrants i s estimated to be US$156 million. By adding the two figures (remittances from abroad and earnings brought back by household member from abroad), the total amount of annual remittances is about US$359 million. Even then this number is probably an underestimation (the official estimate is US$543 million), since transfer amounts are likely to be under- reported. 5. ConcludingRemarksand PolicyImplications There is little doubt that Albania is a country on the move, both internally and internationally. This astonishing mobility is often part of a household-level strategy to cope with the economic hardship of transition. As eloquently put by Adriana, from Gramshi: "A poor family i s considered to be one that does not have a member who can emigrate abroad.'ys4Access to a migration network and family exposure to migration is seen by many as the only viable way out of poverty. In relation to internal mobility, three conclusions stand out. First, in the past decade Albania has experienced a demographic transformation from rural to urban, and from the North Eastern Mountains to the districts of the Coast and Tirana. Second, Tirana is by far the principal destination of internal migrants, a process that appears to have accelerated and become increasingly concentrated in the second halfo f the 1990s. Migration to Tirana i s strongest towards poorest peri-urban areas, re-enforcing a vicious cycle of poverty and adding strain to already overstretched municipal services. Third, there i s evidence of a local rural-urban migration within the North Eastern region. Whether these urban centers in the north are only interludesto more distant and lucrative types o f migration for these households remains unclear, although the large numbers of migrants from these provincial urban centers inthe North to Tirana seem to 82One US$=132 Leks. 83On the assumptionof an average daily wage of $30 for Greeceand $40 for Italy andEurope. 84Albania QualitativePovertyAssessment, The World Bank, 2002. 133 suggest that many, indeed, eventually move on. Given the internal migration to the urban areas of the Coast, as well as local rural-urban migration, creating economic opportunities in urban areas beyond Tirana could help in re-establishing a more sustainable rural-urban flow, as Tirana alone is strugglingto accommodate and service the increasing number of migrantsflocking to its poorest neighbourhoods. Clear patterns o f temporary migration can be identified.The vast majority of temporary migrants, mainly from rural areas in the Center and the North-East of the country, travel to Greece to seek short-term employment opportunities to complement the meager earning from agricultural activities. Despite the higher transaction costs, in recent years the flow of temporary migrants to more distant destinations such as Italy and Germany has increased substantially. Temporary migration appears to leave a dent on the welfare level o f the household only if repeated year after year. Evidence indicates that among the non- migrants, a large share has considered the option of migrating, and many have tried and failed. An astonishing 35 percent of children of household members currently live abroad. For children who left home in the 1990s, the share reaches exodus proportions, with one child in two currently living abroad. The poverty headcount for households with children abroad is significantly lower (12-18 percent, depending on the destination), as compared to households with children in Albania (31 percent). The amount of remittances from children living abroad is estimated to be over US$200 million per year. Almost one-half of Albanian households have access to migration networks, either through direct migration of a household member or through their children living abroad. This i s most likely an underestimate, and it i s comparable to a country like Mexico, which has over 100 years of migration tradition and experience, compared withjust over a decade for Albania. The LSMS data conservatively estimate that 458,000 currently live abroad. Information from the two principal receiving countries puts the number of legal residents in these countries as approximately 600,000 in 2000-2001, or 1/6 of the Albanian population. This number is typically cited in other sources as well. Although migration, and the resulting remittances, is likely to provide an indispensable ingredient in Albanians recipe for (rural) development, there is increasing consensus on the necessity to devise more appropriate, sustainable strategies to lift rural households out of poverty and promote the country's growth, as the phenomenon is also increasingly contributing to social dislocation, agricultural labor shortages and rapid deterioration in the provision o f social services inurban areas. While remittances are currently holding up the economy, contributing about 14 percent of GDP, it is unlikely to be a permanent, long-lasting solution. Inevitably demographic trends will reduce remittances as migrants settle. In any case, it is important that Albania is able to convert an increasing share of remittances into investment. The role o f policy-makers is to ensure the proper conditions are in place for that to happen. First, and most important, it must improve the investment climate in the country. Second, it must contribute to legalizing and regulatingmigration flows to Europeancountries, while also ensuring that opportunities are created for the ones left behind. Finally, it must facilitate remittances and reduce transaction costs and spillages o f transfers. Albania cannot afford another decade o f brain drain. Besides, the impact migration i s having on the social fabric of the population and on relations with its neighbors is becoming too evident. Moreover, the depopulation of the countryside and the unsustainable demographic pressure on cities like Tirana are already creating additional impedimentsto a balanced and equitable growth process. 134 CHAPTERIX: LOOKING AHEAD 1. ReducingPoverty Albania has enjoyedstrong economic performanceinthe last five years, and growthhas averagedabove 7 percent annually. Inflation has been contained at low levels of 4 percent or less for most of the period, during which both exports and imports have grown steadily, as has the current account balance.Despite the impressive performance of the economy, however, poverty in Albania has remained high and per capita income, at around US$1,230 in 2002, has remained one o f the lowest amongtransition economies and one-quarter of the Albanian population, or close to 780,000 individuals, are currently below the poverty line. A large number of individuals are clustered around the poverty line, so that a small exogenous income shock can potentially send disproportionately large number of people into poverty. There is a marked spatialand regionaldimensionto poverty in Albania, and rural areas and the Mountain regionare considerablypoorer than the rest of the country.Modernizationof the countrytriggeredby the transition reforms and the recent period of steady economic growth is contributing to improve overall living standards in Albania, but the benefits appear to be accruing mostly to urban areas, particularly Tirana, while rural and more remote areas are being left behind.There is little doubt that without direct, focused and well-targetedpolicy actionand support from the Government,the possibilityfor many ofthe rural poor, and for the uneducated urban poor, to escape poverty by connecting to the economic growth processseemsremote. The analysis of the previouschapters pointsto several areas that require immediategovernmentattention and focus in order to reduce povertyandbringpeople out ofthe situationinwhich they find themselvesto be susceptible to even small exogenous shocks. The first area that warrants priority is employment generation, for there exists a strong linkage betweenbeingunemployed and being poor. Recallthat only 33 percent of the individuals 15 years of age andolder hold a full timejob, while another 21 percentwork on a part-time basis, bringing the total of working people to 53.9 percent. Irrespective of how employment is defined, the poor exhibit substantiallyhigher unemployment rates when compared to the non-poor, and are almost double that among the non-poor.Amongthe extreme poor the unemployment rate is virtually triple. A significantly higher proportion of marginal workers (those classified as discouraged, seasonal and laidoffworkers) are found amongthe poor. Albania enjoys a high labor force participation rate (66 percent), but the levels of skills are low, and almost two-thirds of the labor force havingjust basic or less than basic education. Almost two-thirds of the working population is either self-employedor work as unpaid workers in a household farm or non- farm enterprise. Most self-employmentis in agriculture, and only one-thirdof the employed earn wages. About 10 percent of the employed in Albania work in the informal sector, and this share is about 24 percent inurban areas. There are morementhan women inthe informalsector. There is evidence of long- term employment in the informal sector, with almost 75 percent of individualsbeingengaged in informal activities for morethan three years, and about 10 percentworkinginthe informalsector for morethan 10 years. Unemploymentremains widespread, and the unemployment rate is especially high in urbanareas. Long-termunemployment reigns - it takes, on average, more than one year to find ajob. Tirana has the lowest labor force participationand the highest unemploymentrate comparedto other geographical areas inthe country. The predominantlabor marketand poverty issue in Albania is how to deal with the big excess supply of labor, given the limited market mechanisms available. The government is no longer able to keep labor demandartificially inflated,andthe newly created marketmechanismsare too weak to offset employment losses.Inaddition,the wages are low and there is not muchroomfor price-adjustments.The excess labor supply has brought high rates of unemployment, hiddenunemployment, and a largely subsistence-based informal sector. As the result, large segments o f the labor force, such as younger people, new graduates, 135 women, and people o f pre-pension age do not participate in the labor market. The recently adopted NSSED sees the development and the promotion o f employment as one o f the main components o f the poverty reduction policies. The document foresees preparation o f relevant labor market legislature to insure provision o f the professional training and the development of the workforce, the formalization o f the labor market, and the promotion o f employment. These goals could, however, be difficult to reach if not accompanied by structural reforms aimed at improving job creation and the creation o f a favorable business environment. Box 9.1: InvestmentClimateinAlbania A recently concluded study o f the investment climate in Albania conducted by the Foreign Investment Advisory Service (FIAS) o f the IFC and World Bank highlighted bureaucratic bottlenecks, weak law enforcement and strong corruption as major impediments to the growth o f private investment, and emphasized the need for removal o f administrative barriers as the key to encouraging new investment. The study concluded that investment procedures in Albania are seriously flawed and characterized by over-complexity, lack o f clarity and transparency, and investor unfriendliness. Private business face bureaucratic delays, discretionary and arbitrary treatment, and high costs, including frequent "unofficial payments, all o f which increase the costs o f investment. Listing the overarching issues, the study notes the systemic lack o f implementing regulations and operational guidelines. Newly introduced laws are typically not supported by efforts setting out necessary procedures, criteria, standards and responsibilities. This leads to a total lack o f transparency in implementation, and allows for considerable operational discretion. In the absence o f effective institutional structures, there i s no clear designation o f individual agency responsibility & accountability, and the investors often do not know exactly where to go for what. Legal and regulatory enforcement is poor, and dispute settlement mechanisms, both within public administration and the court system, tend to be ineffective. Areas o f concern include customs procedures, tax administration, land and property rights, construction permits, and sector/activity licensing, all o f which are characterized by a high level o f bureaucracy, lack of transparency, poor law enforcement, and pervasive practices o f corruption. Administration i s further weakened by a lack o f development o f the appropriate mindset and skills o f officials at all levels. In general, regulatory officials do not trust businesses, and tend to see themselves as policemen, not service providers. Inthe context o f this pervasive dysfunction, it is hardly surprisingthat corruption stands out as the most serious factor impeding reform efforts. Experience in many countries suggests that governments can advance the battle against corruption by simplifying the investment process and eliminating scope for discretion. The Government o f Albania should make more vigorous efforts to clarify institutional responsibilities and enhance agency accountability, enforce laws and regulations, and educate the general public and civil servants at all operational levels on their rights and responsibilities. At the same time, the effectiveness and credibility o f laws and regulations designed to safeguard the legitimate public interest should be increased. Undoubtedly, these initiatives will require a strong political will and long-term commitment. Developing afavorable business environment Development o f the private sector is critical for sustaining economic growth in Albania and providing employment. Creating a business environment favorable to the growth o f the private sector should therefore be a priority for the Government, which needs to take measures to transform informal business into formal business, increase the level o f competitiveness o f the economy, strengthen the legal and institutional framework, and encourage the inflow o f capital to the economy. 136 Inpractical terms, along with supporting new business initiatives, the Governmenthas to take measures to remove existing constraints to firm creation and growth. Most common obstacles to private business development in Albania are corruption, poor law enforcement, insecure property rights, bureaucratic harassment o f small firms (frequent inspections, audits, etc.) and other lengthy administrative and regulatory procedures. Reforming labor market institutions Like other transition economies, Albania inherited labor market regulations developed under central planning, which have proven to be ill suited to the needs o f a market economy. This legacy o f rigid regulations has posed a considerable hindrance to labor market performance in the country. The labor market reforms in Albania have to (a) improve employment flexibility through lessening dismissal costs (e.g., lowering severance pay, introducing temporary employment contracts); (b) enhance working time flexibility (e.g. through providing for rescheduling o f working hours); (c) increase wage flexibility (e.g. through eliminating rigid wage grids); and (d) reduce some o f the non-cash benefits. Simultaneously with the deregulation o f labor relations the reforms have to support their devolution via increased participation o f trade unions and employers' organizations in collective bargaining. Box 9.2: AdministrativeCorruption inAlbania Albania suffers high levels o f administrative corruption, exemplified by the shop owner forced to pay bribes to an endless list o f officials, the blind eye o f some customs officials in the trafficking with goods or human beings, or fraudulent tender procedures in some public contracts. The 1999 Business Environment and Enterprise Performance Survey (BEEPS) finds that Albania, with a score o f 4.0 (bribes as share o f annual revenues, on an increasing scale o f 0 to 7) has the highest administrative corruption amongst the Central Eastern European countries (average score 2.2) and at par with administrative corruption levels o f CIS countries (average 3.7). The BEEPS survey lists the main obstacles to the business sector in Albania as: organized crime/mafia, policy and financial instability, street crime/theft/disorder and a malfunctioning judiciary. The cost o f doing business in Albania is high: obtaining business licenses, acquiring land titles or leases, hiring employees, complying with government's laws and regulations, hooking up and maintain electricity and telephone services - all these activities are expensive and time-consuming for the Albanian business community. Employment in RuralAreas Since a large number o f the poor live in rural areas, it is absolutely imperative that the Government develops a rural development strategy focused on poverty reduction in the short run. A large share o f the population i s engaged in agriculture, but the farm size is, in many cases, too small to be a viable commercial option or simply one that allows a family to escape poverty. Since the agricultural sector provides the most substantial share o f income for the rural poor, addressing the factors hindering the development o f this sector have to be an essential component o f any policy mix aimed at reducing rural poverty. Agriculture i s the key income generating activity in rural areas, and even richer households rely heavily on agriculture as their main source o f income. Sales o f livestock are more important for poorer households and households that live in mountain areas. Measures that provide access to better small scale processing could potentially benefit the poor and the more remote farmers and increase their income generation from livestock. 137 Quintile Non-farm Labor Agriculture Sale of animal products Sale of Livestock Sale of Crops 1 2.6 27.7 52.4 8.3 11.7 32.4 2 8.5 I 34.5 I 45.6 7.9 7.4 30.3 3 6.0 31.3 42.7 7.8 9.3 25.5 4 7.8 32.4 42.6 7.5 5.9 29.2 5 4.9 31.5 33.8 12.6 3.5 17.7 Land by itself i s not a determinant o f the wealth o f the farm household, and 15 percent o f the households with more than 2.5 hectares belong to the poorest q~intile.'~ Production costs differ significantly between farms, with smaller and poorer farms spending more per hectare on farm inputs. A large share o f these differences derives from higher fertilizer costs. Use o f farm machinery for cultivation i s very limited in farm households. Few families own even small tractors and animal drawn ploughs are common for smaller land sizes. Very little other machinery - like trailers, seed drill, pesticide sprayers, etc. - i s used, regardless o f the land size. Greenhouses exist incoastal and central region, but not inthe mountain areas. Agriculture is an important source o f cash income as well. While a large share o f farmers tends to produce for their own consumption, farmers who are able to market some o f their products tend to rely heavily on agriculture as a source o f cash income as well. For some of the poorer households, cash income from agriculture might be one o f the few cash income sources that they have. Considering that the poor also have overall smaller non-income cash transfers, agricultural income is extremely relevant. Without doubt, agriculture i s important for the overall economy o f Albania, and a large share o f the population relies on agriculture as one o f their main sources o f income. Till such time as income opportunities become available in other sectors o f the economy, the importance of the agricultural sector will remain high. Agriculture has a strong impact on food security that should not be underestimated. Poorer households rely on consumption o f their own products to a large extend and agriculture also serves as a shock absorber. As long as other income opportunities are limited, agriculture provides households a means to survive as well as to generate some cash income. Flexible and locally adapted rural extension service can do a lot to improve current production patterns. Poorer farmers have higher production costs and less income. They are not so diversified and tend to be less engaged in processing. These are all issues that can be addressed by extension services that take the local conditions into account and advise farmers based on market possibilities. Increased quality and quantities will potentially be used by small to medium enterprises that start processing activities in rural areas and can thus be a potential outlet for poor farmers. Land markets and structural change are important for the development of rural areas. Land sizes are on average so small that is difficult to access markets, unless farmers are organized. It appears important to assure that land markets are functioning properly so that when people decide to migrate and are able to assure income elsewhere, land can be easily put on the market. However, a prerequisite for structural change i s that employment opportunities increase so that families feel secure to not keep their livestock as a security mechanism and protection for bad times. Migration i s and will clearly change the rural areas; however, people will stay in rural areas and not move out as long as income opportunities exist in the rural areas. A focus o f improved income generation form agriculture will contribute to the possibility o f some rural areas to thrive. 85Certainly, land quality is also relevant, but information on quality o f land is not available inthe LSMS survey. 138 Employment in the Non-Farm Rural Economy With the collapse of Albania's state-financed and operated rural enterprises, the non-farm rural economy (NFRE) almost totally collapsed, notable casualties being agro-processing and mining enterprises. Over the last decade, employment fell dramatically from 38,000 to a mere 4,800 in the mining sector alone. These enterprises were not competitive in the new environment of a market economy, and their attractiveness to investors was reduced by the difficulties o f renewing the physical infrastructure and utilities necessaryfor their profitable regeneration. Currently the NFRE i s restricted to low-grade construction, small technical and personal services, and petty trade. Employers in the service sector are usually family-run businesses that are small and do not create many new jobs. Additionally, work tends to be casual, intermittent, and low-wage. The few jobs that are created tend to provide little or no opportunity for learning new skills, further disadvantaging the already weak rural labor force. Inaddition, the NFRE provides only the most basic subsistencegoods and services, and is primarily based on kinship and traditional codes of interpersonal reciprocity rather than modern contractual practice. The NFRE is also starved of the necessary investment and initiative, which are now directed to the lowland urban and peri-urban areas, where the highest growth rates occurred during the 1990s. This has resulted in large-scale out-migration from rural areas, primarily by the most active male members of the rural labor force who are attracted by the job prospects in urban areas and abroad. Consequently, the rural population i s becoming increasingly composed o f women, the elderly and the unskilled, substantially narrowing the human resourcebase for NFREdevelopment. Albania's NFRE has also suffered from a severe lack of the institutional arrangements needed for development. Key institutional gaps are in education and training in technical and business skills, access to financial services, articulation with local governance and civil society organizations; and the enactment, dissemination and application of a body of appropriate commercial law. There has been a steady reduction in skills of the rural labor force due to migration, aging and the absence of education and training of new entrants. TheManufacturing Sector In terms of future growth, attention must necessarily shift to the manufacturing sector, which currently contributes only 11.6 percent of the country's GDP, and employs only 4 percent of the workforce in the country, making it the fourth-largest employer, after agriculture, wholesale retail and repairs, and construction. The agriculture sector employs more than half of the labor force, but most are unskilled and low-skilled, and over 15 percent are part-time. The extent of disguised employment in agriculture is not well understood, but there is always the distinct possibility that land consolidation and technology could well push a large number of the relatively unskilledout of the agricultural sector. As it is, migration is effectively sending a large number o f persons from rural areas to urban and peri-urban centers of the country. Employment in construction is already quite high.At the same time, there i s a limit to the extent that construction activities can be expected to grow, and it does not seem likely that the construction sector will be a catalyst for the creation of new revenue generating activities. Stimulating the manufacturing sector could potentially create several employment opportunities. Within the manufacturing sector, there exist several industriesthat could potentially offer different employment opportunities, and it could be possible to further target the industries with the highest growth potential. Consider, for instance, textile manufacturing, which i s the largest employer within the manufacturing sector, which also has the lowest wages. Another sector is leather goods which, given Albania's proximity to Italy and wide familiarity with the language combined with recent trends in dispersedmanufacturing, could well afford Albania the opportunity to ease into production. This opens up intriguing options, and is an area that needs to be explored in depth. 139 2. Areas for FutureResearch There are at least five areas that needto be explored further and in greater detail: Generating dynamism in rural areas Faced with an economy experiencing the kind o f radical changes that Albania has in the last decade, one has to ask how dynamism can be created in rural areas such that it benefits the poor. As stated earlier, the benefits o f the growth process in Albania have tended to accrue mostly to urban areas, and rural areas are being left behind. Without direct, focused and well-targeted policy action and support from the Government, the possibility for many o f the rural poor, and for the uneducated urban poor, to escape poverty by connecting to the economic growth process seems remote. Poor people tend to be concentrated inmicro-firms, and tend to be under-representedinthe public sector and highlyrepresented inagriculture. However, land-holdings tend to be small, so that yields from agriculture for the poor tend to be low. At the same time, in rural areas, the non-poor tend to be well represented in the non-agricultural sector, and the for many reasons - related, inter alia, to education, gender, and family size - the participation o f the poor in the rural non-farm private sector activities tends to be low. As a result, the poor tend to be employed in less profitable, less attractive firms. In such a situation, therefore, it i s necessary to explore how the necessary dynamism can be created in rural areas that would provide the necessary impetus for the poor people to get out - and potentially stay out - o f poverty. Therole of entrepreneurship inpoverty reduction The role o f entrepreneurial activity and self-employment in upward income mobility and escape from poverty needs to further examined. Ingeneral, entrepreneurs are more likely to experience upward income mobility relative to the rest o f the population, and it i s useful to explore and identify the role o f entrepreneurial activity among those who did manage to escape out o f poverty. At the same time, the interaction between entrepreneurship and investment climate needs to be better understood, since entrepreneurship and employment in small enterprises i s more likely to be associated with upward income mobility if the investment climate improves. Exploring these issues may help get at the question o f how entrepreneurship might reduce poverty - either through an effect that is direct, Le., where poor people start small businesses and escape from poverty, or indirect, Le., the wealthy and educated are more likely to be entrepreneurs and own small firms, butthen employ poor people. Migration and return migration Of the many issues related to migration, the two that warrant urgent scrutiny relate to consolidation o f land and to return migration. First, migration to urban and peri-urban areas has not led to consolidation o f land in rural areas. The existing land holdings are clearly inadequate for efficient agriculture, and yet, even with the heavy migration that Albania has experienced, there has been almost no positive effect on land consolidation. This is an issue that needs to be better understood. Second, there i s no evidence that return migration has started to happen. Returnmigration to the home country has a variety o f long term benefits in terms o f FDI, local investment, entrepreneurship, investment capital, etc. The high out migration may also be tied to some o f the same problems o f public corruption, ineffective management and impediments to entrepreneurship. Countries like Ireland and Taiwan have experienced high rates o f return migration, and it could potentially be useful to examine the implicit and explicit policy measures adopted inthese countries that supported this trend. Growthandpoverty Given the large number o f people near the poverty line, economic growth is likely to have disproportionately larger effects on poverty. In the absence o f panel data, linkages between growth and poverty could not be analyzed in this assessment, but this is an area that could benefit from future research 140 Thepotential of tourism as an enginefor growth Another area that deserves greater attention is tourism. Albania is a country of great natural beauty, enhanced by the mountains in the north and the sea inwest. At the same time, Albania is close to Greece and Italy, countries that attract lots of tourists from within Europe and outside. Development o f tourism within Albania presents itself as an attractive option, one that needs to be explored and exploited more aggressively. 141