Report No. 35361-AM Armenia Labor Market Dynamics (In Two Volumes) Volume II: Main Report May 2007 Human Development Sector Unit Europe and Central Asia Region Document of the World Bank SSP Social Support Program TFR Total Fertility Rate U L C UnitLabor Costs UNDP UnitedNations Development Program UNICEF UnitedNation's ChildrenFund USAID United States Agency for International Development USSR Union o f Soviet Socialist Republics VET Vocational Educationand Training Vice President: Shigeo Katsu Country Director: Donna M.Dowsett-Coirolo Sector Director: Tamar Manuelyan Atinc Sector Manager Arup Banerji Task Team Leader: Arvo Kuddo ACKNOWLEDGMENTS This report was written by Arvo Kuddo (Task Team Leader; ECSHD). It relies on the contributions prepared by the team o f Armenian experts fiom AVAG Solutions Ltd. led by Vahram Avanesyan. Backgroundpapers on particular topics were prepared by the following authors: (a) macroeconomic developments and the labor market: Armen Yeghiazaryan and Vahram Avanesyan; (b) rural employment: Armen Yeghiazaryan, Vahram Avanesyan, and Nairuhi Jrbashyan; (c) youth employment and unemployment: Vahram Avanesyan, Melik Gasparyan and Nairuhi Jrbashyan; and (d) labor legislation and labor market institutions: Vahram Avanesyan and Melik Gasparyan. The fieldwork on conducting labor market surveys was supervised by Ruben Yeganyan and Norayr Gasarjyan. The report uses the contributions provided by Sona Harutyunyan (Director), Anahit Parsadanyan and Hasmik Tadevosyan from the State Employment Service, Louisine Qalantaryan, Head o f Labor Market Division o f the National Statistical Service o f the Republic o f Armenia (administrative and Labor Force Survey data), and Gorana Krstic (Integrated Living Conditions Survey data). Tatyana Bogomolova and Asta Zviniene (HDNSP) provided assistance inpreparingpopulationforecasts for Armenia. The team i s grateful for the support of Donna M. Dowsett-Coirolo, Country Director (ECCU3), Arup Banerji, Sector Manager (ECSHD), and Aleksandra Posarac, Country Sector Coordinator for South Caucasus (ECSHD). The team benefited from comments o f peer reviewers Verdon Staines (AFTH3) and Carlos Silva Jauregui (MNSED), as well as Christian Peterson (ECSPE), Saumya Mitra (ECSPE), and Anthony Cholst (ECCU3), Special thanks to Diane Stamm, who edited the document. a ARMENIA LABORMARKET DYNAMICS Contents Page Acknowledgments .................................................................................................................................. a Introduction............................................................................................................................................. Contents ................................................................................................................................................... i Chapter I: Macroeconomic Developments and the Labor Market......................................................... 1 5 1.1. Economic Growth and Employment ..................................................................................... 5 1.2. 11 Productivity Improvements ................................................................................................. Job Flows, by Industry........................................................................................................ 1.3. 18 Chapter 11: Labor Market Trends......................................................................................................... 23 2.1. 23 Employment Ratio............................................................................................................... Labor ForceParticipation.................................................................................................... 2.2. 28 2.3. Unemployment Trends ........................................................................................................ 34 2.4. Inactive Population.............................................................................................................. 38 2.5. Older and Disabled Workers ............................................................................................... 39 2.6. ChildLabor ......................................................................................................................... 41 2.7. Gender Segregation............................................................................................................. 42 2.8. Urban-rural Dimensionand Regional Disparities .............................................................. -44 Chapter 111: Wage Dynamics and Unit Labor Costs ........................................................................... 2.9. Education Level and Labor Market Outcomes.,.................................................................. 46 -50 3.1. Wage Dynamics .................................................................................................................. 50 Chapter IV: Rural Employment ........................................................................................................... 3.2. UnitLabor Costs ................................................................................................................. 56 58 4.1. -58 Landownership and LandUse............................................................................................. Outputand Employment inAgriculture ............................................................................. 4.2. 63 4.3. Chapter V: Youth Employment andUnemployment........................................................................... Income from Agriculture..................................................................................................... 64 66 5.1. 66 Households....................................................................................................................................... 5.2. The Level and Structure ofunemployed Youth ................................................................. Sociodemographic Characteristicso f RegisteredUnemployed Young People and their 68 5.3. 72 Establishments .................................................................................................................................. 75 5.4. Past Interactions o f Unemployed Youth with the Labor Market......................................... Students inPrimary Professional Schools and in Secondary Specialized Education Chapter VI: Labor Migration............................................................................................................... 78 6.1. Outflow o f Population......................................................................................................... 78 6.2. 80 Urban-rural Migration......................................................................................................... Current Labor Migration..................................................................................................... 6.3. 82 6.4. Remittances......................................................................................................................... 83 Chapter VII: Business Environment and Labor Markets..................................................................... 86 7.1, 86 88 The Role of Labor Regulations ........................................................................................... Challenges to Creating a Business ...................................................................................... Importance of the Business Environment............................................................................ 7.2. 7.3. 91 7.4. Chapter VIII: Labor Legislationand Labor Market Institutions.......................................................... The Role o f Skill Level and Education............................................................................... 91 94 8.1. State Intervention inthe Labor Market ............................................................................... 94 8.2. Employment Protection Legislation.................................................................................... 96 1 8.3. 101 103 Unemployment Benefits.................................................................................................... Active Labor Market Programs......................................................................................... The Capacity ofthe StateEmployment Service................................................................ 8.4. 8.5. 107 8.6. 109 Chapter IX: Policy Recommendations.............................................................................................. Social Dialogue ................................................................................................................. 111 References........................................................................................................................................... 120 Annexes .............................................................................................................................................. 128 Tables Table 1.1. Volume Indexes of Gross Domestic Product (constant prices). 1991 = 100........................ 6 Table 1.2: Employment Dynamics inHealth. Education. Science. AndPublic Administration. 1995- 2005 (1000') ......................................................................................................................................... 10 Table 1.3: Jobs, Value Added, andLabor Productivity inthe Nonagricultural Sector o fthe Economy, 1995-2003 ............................................................................................................................................ 10 Table 1.4: Parametersof Cobb-Douglas Production Functionfor Industry,by Sector,...................... 15 1995-2003 ............................................................................................................................................ 15 Table 1.5: Dynamics of Agricultural Employment inArmenia And other Transition Countries, 1990- Table 1.6: Output and Labor Productivity inIndustry,2001-2003, By Enterprise Size..................... 2003 ...................................................................................................................................................... 18 20 Table 1-7: Labor Productivity: International Comparisons ................................................................. 22 Table 2.1: Employment o fPopulationinthe USSRand inSome of the NationalRepublics (including insubsidiary householdplots; 1989populationcensus, in%)' ............................................................ 24 Table 2.2: EmploymentPopulationRatios and Activity andUnemployment Rates In2004 in Armenia and Georgia, and insome European Countries...................................................................... 27 Table 2.3: Unemployment Rate Estimatesfrom Different Sources, 1996-2005 ................................. 34 34 Table 2.4: Economically Inactive Population, 2001 Population CensusData; '000 ........................... (Population aged 15+) ........................................................................................................................... 39 Table 2.5: Employment, by Occupation and Gender, inPercent, 2004 LFS....................................... 43 Table 2.6: Population andLabor Force, by Marz................................................................................ 47 Table 2.7: Have You Worked since Graduation? (Respondents aged23-33), Percent....................... 48 Table 3.1: Armenia: Summary of Earnings Distribution, 1998/99 and 2004 ...................................... 52 Table 3.2: Average Annual Earnings and Returnper Unit o f Cost of Education, 16-33 Year-olds Table 3-3: The Dynamics of Average Monthly Wages and MinimumWages InArmenia,................54 with Jobs, by Gender and Level of Education, In'000 Drams; 2001................................................... 55 Table 4.1: Number of Farms and the Total Area O ftheir Privatized Agricultural Lands ...................59 1997-2006, Drams................................................................................................................................ 55 Table 4.2: Employment-to-Population Ratio and Employment Status by Stratum, Percent...............60 Table 4.3: Economic Activity and Unemployment by Age Group, June-July 2005........................... 62 Table 4.4: The Average Monthly Income from Agricultural Activity (Crop production, animal husbandry, processing) per Household, Drams .................................................................................... 64 Table 5.1: School Attendance of De Jure Population, by Age and Type O f Educational Institution, Table 5-2: Youth Unemployment Rates andthe Ratios o f Youth-to-Adult Unemployment Rate ......67 2001 Census.......................................................................................................................................... 68 Table 5.3: Unemployed Youth withProfessional Qualification, by Area of Specialization (percent) and Average Durationof BeingRegistered,At the SES (months) ....................................................... 71 Table 5.4: Distribution o fReasons for Not Obtaining Training Or Retraining by Unemployed Youth, Percent.................................................................................................................................................. 72 Table 5.5: Unemployed Youth, by Sourceo f Job Search, Percent ofTotal ........................................ 74 .. 11 Table 5.6: Circumstances with a Decisive Influence on the Selection Of Specializationby VET Students. Percent................................................................................................................................... 77 Table 6.1: The Number o f Population o fArmenianNationalityLiving inthe USSR and inSome o f the CIS States; inMillions (According to the 1989population census and the latest census data)......80 Table 6.2: Estimates o f the Average Remittance that Armenian Temporary Workers inMoscow and Rostov Send Home. 2003 ..................................................................................................................... 84 84 Table 6.4: Economic Benefits and Costs o f Remittances to a Receiving Country .............................. Table 6.3: The Impact o f Remittances on Household Incomes inArmenia (drams) ........................... 85 Table 7.1: Starting-a-Business Indicators-Measures to Register A Business Formally inSome Countries in2006.................................................................................................................................. 89 Table 8.1: Key Benchmarks on Labor Regulations from the National Labor Laws (Regular Table 8.2: Armenia: Public Expenditure on Labor Market Programs, 1998-2003 ........................... employees, that is, excluding minors, disabled, and so forth) ............................................................. -99 Table 8.3: Involvement o f Officially Registered Job Seekers inActive Policy Measures................104 106 Table 8.4: Reasons for Unwillingness o f Unemployed to Register at the Regional Employment 108 Table 9.1: Impacts o f ALMPs Based on Experience o f EUCountries.............................................. Centers inthe Opinion o f Directors .................................................................................................... 113 Figures Figure 1.1: Structure o f Value Added and Employment. 1990-2005 ................................................... Figure 1.2: Dynamics o fMainEconomic Indicators. (1991 = 100) ...................................................... 7 9 Figure 1.4: Job Creation and Destruction inIndustry. 1999-2003 (Inreporting enterprises) .............12 Figure 1.3: NonagriculturalJobs inthe State andPrivate Sectors. 1995-2003 ................................... 13 Figure 1.5: Job CreationandDestructioninConstruction. 1999-2003 ............................................... 14 14 Figure 1.6: Number and Size o f Industrial Enterprises inArmenia,.................................................... (Inreporting enterprises) ...................................................................................................................... 16 Figure 1.7: MainIndicators for Agriculture inArmenia, 1990-2004, Percent.................................. 1996-2003 (1995 = 100) ...................................................................................................................... 16 17 Figure 1.8: Structure o f Output and Employment inIndustry,by Sector, 1990-2003 ........................ 19 Figure 1.9: Ratio of Labor Productivity-to-GDP Per Capita inArmenia, ........................................... 21 Figure 2.1: Population Pyramid o f Armenia, 2001 Population Census............................................... 1994-2003 (1990 = 100) ...................................................................................................................... 21 25 Figure 2.2: Population Pyramido f Armenia in2025, Scenario I........................................................ 26 byMainAge Groups, '000s.................................................................................................................. 28 Figure 2.4: The Dynamics o f the Able-bodied Populationo f Armenia, 2001-30, .............................. Figure 2.3: Population Pyramid o f Armenia in2025, Scenario II....................................................... 26 28 Figure2.5: Employment-to-PopulationIndex inArmenia; 1990= 100.............................................. 29 Figure 2.6: Employed by Actually Worked Hours; 2004 LFS ............................................................ 29 Figure 2.7: Employed, by Gender and Type o f Work, inPercent, 2004 LFS...................................... 30 Figure 2.8: Employed by Employment Status, 2004 LFS, inPercent ................................................. 31 Figure 2.9: Tax Compliance inArmenia, CIS, EU8, and ECA Countries in2005, According to the EBRD-World Bank Business Environment and Enterprise Performance Survey (BEEPS) ................32 Figure 2.10: Forecast o f the Dynamics o f EmployedPopulationAged 15-64, to Maintain the Employment Rate o f Able-bodied Population at the Current Level o f 55.9 Percent, or to Raise to the Level o f 65 Percent and 70 Percent ...................................................................................................... 33 Figure 2.11: Unemployment Rates at Age 15 and Over, According to the Labor Force Survey Data Figure 2.12: Officially RegisteredYouth and Adult Unemployment, 1994-2003 (End-year data) ....35 for 2001-05, inPercent......................................................................................................................... Figure 2.13: Structure o f Registered and Self-reported ( L O definition) Unemployment, .................36 37 iii By Selected Age Groups. 2003............................................................................................................. 37 Figure2.14: Unemployed by the Durationo f Job Seeking, 2004 LFS, inPercent.............................. Figure2.15: PopulationProjection for Age Groups 7-17,65-79, Figure 2.16: Employment Rates o f Older Workers, 2004 LFS and 2004 ILCS Data .........................and 80 and Over, InArmenia ......40 38 41 Figure 2.17: Hours Worked by Employed Childrenina Week, 2004 LFS......................................... 42 Figure 2.18: Participation, Employment, and Unemployment Rates inUrban and RuralAreas, Figure 2.19: Unemployment Rate by the Level o f Education, inPercent, 2004 LFS ......................... According to the 2004 ILCS, PopulationAged 15 and Older ............................................................. -45 48 Figure 3.1: Dynamics o f Gross Domestic Product, Employment, Real Wages, and Labor Productivity .................................................................................... 51 Figure 3.2: Average Wages o f Employees with Different Levels o fEducation,................................. inArmenia, 1990-2004 (Indexes, 1990= 100) 53 According to 2004 LFS Data, inDrams ............................................................................................... 53 Figure 3.3: Dynamics o f Unit Labor Cost inArmenia, 1990-2004 (1990 = 100)............................... 56 Figure 4.1: Dynamics o f Value Added, Employment, Labor Income, and ProductivityinAgriculture, 59 Figure 4.2: Employment by Age Group, Percent o f Population......................................................... 1990-2004 (1990 = 100) ...................................................................................................................... -61 Figure 4.3: Average Size o f Agricultural Landper Household, Hectares........................................... 63 Figure 5.1: Households o f Unemployed Youth, by Number o f Working-age And Employed Members, Percent ................................................................................................................................. 69 Figure 5.2: Unemployed Youth without Professional Qualification (specialty) By Intentionto Continue Educationinthe Future, Percent ........................................................................................... 70 Figure 5.3: Unemployed Youth, byReasons o f LosingLast Job, Percent .......................................... 73 Figure 5.4: Reasons for Failure to Take a Job Offered Bythe SES, Percent ...................................... 75 Figure 5.5: Households o f VET Students byNumber o f Working-age and EmployedMembers, Percent .................................................................................................................................................. Figure 7.1: Problems o f DoingBusiness (percent o f firms indicatinga problem) .............................. 76 88 Figure 7.2: Underemployment Dueto Labor Regulations inArmenia, and ECA, EU8, and CIS Figure 7.3: Skills and Education o f Available Workers As a Constraint to Business Expansion .......91 States, Percent, by the BEEPS 2002 and 2005 ..................................................................................... 93 Figure 8.1: Government Interventions and Regulation inthe Labor Market...................................... Figure 7.4: Current Level o f Employment in SurveyedFirms (inpercent); FIAS 2003 Study...........92 -95 Figure 8.2: Labor Regulations as a Problem o fDoingBusiness (percent o f firms indicating labor regulations as a problemdoing business, 2005, According to the BEEPS survey) .............................. 97 Figure 8.3: Labor Market Regulation inArmenia (before the approval o f the new Labor Code) and in Selected Transition Economies............................................................................................................. 98 Figure 8.4: Assessment o f Labor Protection Legislation inArmenia (for old and new Labor Codes) and Comparisons with OECD Countries ............................................................................................ 100 Figure8.5.: Annual Expenditure onPublic Employment Services and Administrationper PESTotal Staff: International Comparisons (2002, 2003; U.S. dollars) ............................................................. 102 Figure 8.6: Distributiono fRegional Employment Centers by Budgeted Annual Expenditure per Officially RegisteredJob Seeker (2004) ............................................................................................. 103 Figure 8.7: Public Expenditure on Labor Market Programs per Officially RegisteredJob Seeker (Drams) ............................................................................................................................................... 105 Boxes Box 2.1: Employment Targets inEUMember States.......................................................................... 33 Box 8.1: Benefits o f Social Dialogue ................................................................................................ 110 iv Annex Figures Figure 1: Share ofMain Sectors inTotal Employment. Percent ....................................................... Figure 2: Employment Dynamics by Sector in 1994-2005 (1990 = 100).......................................... 128 128 Figure 3: Jobs inthe State and Private Sectors o f the Non-agricultural Productive Sphere in 1998- Figure 4: Labor Indicators inIndustry(1990 = 100) ......................................................................... 2005 .................................................................................................................................................... 129 129 Figure 5: Labor Indicators inConstruction (1990 = 100) .................................................................. 130 Figure 6: Labor Indicators inTrade andPublic Catering (1990 = 100) ............................................ 130 Figure 7: Total Fertility Rate inArmenia, 1985-2004....................................................................... 131 Figure 8: Labor Force Participation, Employment and Unemployment Rates by Age, Percent, 2004 LFSData............................................................................................................................................. 131 132 Figure 10: Ratio o f Annual Wages-to-GDP Per Capita in 1990-2003 .............................................. Figure 9: Unit Labor Cost Dynamics by MainBranches in 1990-2003 ............................................ Figure 11: Public Employment Service Staff: International Comparisons (2002, 2003) ..................132 133 Figure 12: Annual Public Expenditure on Labor Market Programs Per Officially Registered Unemployed: International Comparisons (2002, 2003; U S dollars)................................................... 133 Annex Tables Table 2: Main Labor and Employment Indicators inIndustryin 1990-2003 (Thousand Persons) ...134 Table 1: Employment and Wages by Sectors o f Economy. 1995-2005 ............................................ 135 Table 3: Main Labor and Employment Indicators inConstruction in 1990-2003 (Thousand Persons) ............................................................................................................................................................ 136 Table 4: Number and Size o f Enterprises inIndustryin 1990-2003 ................................................. 137 Table 5: Armenian Agricultural Sector: a Comparative View .......................................................... 137 Table 6: Labor Productivity and FixedCapital per Personnel inIndustryby Sector in 1990-2003 (million 1995 constant Drams)*......................................................................................................... 138 Table 7: Unemployment Rate by Age, 1998/99 and 2004 (Percent), ILCS ...................................... 139 Table 8: Participation, Employment and Unemployment Rates by Regions, 1998/99-2004 (population 16 f), ILCS...................................................................................................................... 139 Table 9: PopulationForecast o f Armenia, 2001-2025....................................................................... Table 10: Distributiono f Households by Main Source o f Incomes, Percent, 2004 LFS...................139 140 Table 11: Distributiono f Employed Who Want to Change the Work Situation, ByReasons and Table 12: Average Wages, MinimumWages and Pensions inCIS States In2003 and 2004 ...........140 Status of Employment, Percent, 2004 LFS......................................................................................... 141 Table 13: Average Time Spent onActivities, Proportion Undertaking An Activity and Average Time respectively; Armenian population 15-80 years; all days) ................................................................. Spent by Those Undertaking Activities DuringJuly 2004 (hours, minutes and percentage 141 Table 14: Distributiono f Respondent Rural Households by Size ..................................................... 142 Of Agricultural Land, Percent ............................................................................................................ 142 Table 15: Investments inFixedAssets (inbillionDrams), 1999-2004 ............................................. 142 Table 16: Stratification and Distributiono f the Sample by Strata: Rural Employment Survey 2005 ............................................................................................................................................................ 142 Table 17: Economic Activity and Unemployment: Rural Employment Survey 2005....................... Table 18: Average Monthly Income from Crop Productionper Household, Drams......................... 143 143 Table 19: Average Monthly Income from Livestockper Household, Drams ................................... Table 20: Inflow and Outflow o f Populationby Means o f Transport, 1992-2005 ............................ 144 145 Table 21: Distributiono f Regional Employment Centers (REC) by Number o f Job-seekers and Table 22: Public Expenditure on Labor Market Programs: International Comparisons.,..................146 Unemployed per Staff Member, December 2004 ............................................................................... 147 V INTRODUCTION Labor markets in Armenia, like those in other former socialist countries in Central and Eastern Europe and Central Asia, have gone through a profound transition since the start o f political, economic, and social reforms inthe early 1990s. The labor force movements inthis region are the result o f systemic changes inthe economy and society, rather than being only cyclical. The dynamics and patterns o f employment and unemployment in all the transition countries are o funique origin and character compared to the advanced economies. The major aspects o f such systemic change to employment dynamics are as follows: 0 On one hand, employment trends from the early 1990s were strongly influenced by political transformation, such as the disintegration o f the Former Soviet Union, the rapid demise o f its political and economic structures, and the vertical and horizontal integration o f the member countries established over decades. On the other hand, the founding o f the independent states and the subsequent institutional reforms have created new jobs in such areas as defense, customs and tax services, frontier guards, and the diplomatic service, and created new patterns o f labor mobility, such as interstate labor migration. Transition reforms have led to rapid structural shifts in the economy through the ownership o f firms, in their structure by sectors, occupations, or skills in demand; accompanying adjustment o f the economy to market needs, which have ledto a heavy contraction inaggregate output; and to structural unemployment inparticular. 0 Inrestructuringtheir economies, the countries were not able to avoid a rapid decline inaggregate demand inbothgovernment andprivateconsumptionandinfixed capital investments, which has led to demand-deficient unemployment. Therefore, the transition has been accompanied by both a negative supply shock and a negative demand shock. 0 Many new alternative forms o f employment andjobs have been generated, including inthe informal sector and in domestic economics. Widespread nonparticipation has become common. 0 The labor relationship has been influenced by aspects o f the sociocultural and psychological transition o f society. The ideologies o f employees o f the state-owned firms have been replaced by those o f owners, employers, entrepreneurs, wage earners o f private firms, self-employed, or employees inprivate entities. These aspects have induced new work ethics and discipline. The previous universal and mandatory system o fjob security and employment stability has been replaced by a more liberal institutional fi-amework for hiring and firing, and more flexible labor movements. These and many other changes have made labor markets mobile but vulnerable to external and internal developments inthe economy. The social costs o f such transitional shocks have been enormous, leading to the rapid and deep stratification o f society. Social and labor market policies must partially compensate for the cutbacks caused by systemic change. Armenia i s no exception. 1 The labor market is a key factor market, which influences overall economic efficiency. In a period o f rapid structural change, labor market policies are required to maintain an adequate equilibrium between labor demand and supply, notably by enabling entry and reentry into employment. As a result o f the shocks associated with the transition process, in Armenia considerably fewer employment opportunities were created, and a significant share o f able- bodied population i s thereby "forced" to stay out o f labor markets. The country has had one o f the highest unemployment rates among the transition countries-over 30 percent, according to Labor Force Survey (LFS) data, and over 20 percent, according to Integrated LivingConditions Survey (ILCS) data, throughout the past decade. Box 1: Country Characteristics Armenia is a landlocked country o f 29,743 square hlometers and has a population o f 3.219 million people (2006). Turkey lies to the west, Georgia to the north, Iran to the south, and Azerbaijan to the east. The borders with Turkey and Azerbaijan are closed. The nationhas few naturalresources and generally mountainous land: it relies heavily on imports for its food, and most crops are grown on irrigated land. Agricultural land comprises 47 percent o f the country. Its GDP based on Purchasing Power Parity in2006 is estimated by IMFat US$4,516 (www.imf.org). In 1988, a devastating earthquake struck Armenia, killing about 25,000 people, leaving another 500,000 homeless, and destroying parts of its fuel pipeline and industrial plants. Armenia was also engaged in a conflict with Azerbaijan over Nagorno Karabakh (NK), an ethnic-Armenian enclave inside Azerbaijan, closing some o f Armenia's trade routes. As such, it must rely on the development o f its human resources to acheve economic growth. Like other transition countries, the surge in unemployment in Armenia was largely anticipated, given the need for massive reallocation o f labor across firms, sectors, and regions. However, the persistence o f high unemployment and growing underemployment, even after more than a decade o f rapid economic growth, i s worrisome. Likewise, the reality o f a rather stagnant unemployment pool, with a high incidence o f long-term unemployment (close to 80 percent o f the registered unemployed are out o f ajob for more than one year), i s at odds with the idea o f a massive but also rapid reallocation o f workers across firms and sectors. Moreover, many households have turned to subsistence agriculture and petty trade, staving off poverty, while jobs inthe formal sector have continued to disappear. InArmenia, there is clearly a lack of demand for labor, which requires government efforts aimed at expanding the economic base, promoting more labor-intensive investments and industries, and making labor market incentives more market friendly. The current weak regulation o f the labor market may have helped employment flexibility, but it also appears to have had tradeoffs in promoting informalization o f the economy (through its avoidance), lower worker productivity, and reduced worker welfare, such as low wages, growth o f in- kindsubstitutes for remuneration, andwage arrears. The objective of the study is to determine the main factors behind poor labor market outcomes in Armenia: high unemployment o f long duration despite rapid economic growth. To do so, it will assess, first, the key characteristics o f the demand for labor. These include (a) the impact o f macroeconomic policies on job growth; (b) wage flexibility and unit labor costs; (c) cost-of-doing-business factors, including costs, risks, and barriers to competition faced by firms; and (d) employment promotion legislation and labor market institutions. Recommendations are made on policies that can promote an effective and sustainable 2 increased demand for labor; second, the key characteristics o f the supply o f labor, including the impact o f long-term demographic developments and labor migration, and the impact o f social benefits on the reallocation o f labor. Policy options are analyzed that can support the development o f an efficient supply o f labor to a modern high-wage market economy; and, third, the mechanisms by which employers and workers are brought into contact with each other, and recommend policies that support efficient and effective labor market institutions. Box 2: MeasuringEmploymentandUnemployment Information on the social and economic situation in the labor market has been mainly collected from the following three broad groups o f sources: (a) population censuses, labor force surveys, household surveys, and other specially designed surveys; (b) establishment surveys; and (c) administrative recordsh-egister-based statistics. In particular, the official data on employment in Armenia is compiled by the National Statistical Service (NSS) using the following sources and methods: (a) the data are collected from all large and medium- size enterprises and state organizations by the NSS; (b) small enterprises (as a rule, less than 100 employees) are surveyed selectively, and the employment data are then aggregated to reflect the whole sample o f registered small firms; and (c) self-employment and informal employment are calculated using the data o f land registers and administrative registers. For example, a person with landownership is considered employed. The State Employment Service (SES) collects the data on registered unemployment. Inthis report, the data on overall employment and unemploymentlevels in2004105 are derived from the labor module o f the Integrated Living Conditions Survey (ILCS) that was conducted between April 2004 and March 2005, as a more reliable source. This survey has a longer period o f observation and a large sample o f 6,816 households. For a snapshot o f more detailed labor issues, and for the dynamics o f employment and unemployment during a longer period (1996-2005), the data from the Labor Force Surveys (LFS) are also used. The 2004 LFS was conducted inAugust 2004, with 2,539 households participating. During2001-03, the LSFs were conducted inMarch and October, and the annual data reflect the average for these two months. A valuable source of information on labor is the 2001 population census conducted between October 10 and October 19 o f that year. Although the recent household and LFSs and the 2001 census followed International Labor Organization (ILO) definitions o f employment and unemployment, the variance inthe data is significant. The statistical categories o f unemployment and employment, which are highly simplified, can give only broad hints as to what a situation o f employment in Armenia would resemble. Statistics on unemployment and employment are partly an attempt at the indirect measurement o f changing levels o f satisfaction or dissatisfaction within the labor market. Thus if more people are found to be employed, the employment situation is assumed to have improved. The ILO's concept o f "decent work," which takes into account both the qualitative and quantitative aspects o f employment, is a much more appropriate way o f measuring progress in the labor market. The concept o f decent work implies that work should not only provide adequate income, but should respect workers' rights and dignity (see Ghai 2003). The study findings are based on different types o f data and analytical tools. Sections on macroeconomic developments will mainly use available data on aggregate output, employment, productivity, and wages. The section on business environment and labor market outcomes will primarily draw on indicators o f the quality o f the business environment included in the World Bank's Doing Business, the European Bank for Reconstruction and DevelopmentNorld Bank Business Environment and Enterprise Performance Survey (BEEPS), and the International Finance CorporatiodWorld Bank Foreign Investment Advisory Service (FIAS) database. 3 The overview o f labor market developments i s based on the data from labor force surveys, household surveys (ILCSs), and administrative statistics (see Box i.2). These data are supplemented with information on policy settings and changes in policy, including information available on the activities o f the State Employment Service (SES) in the provision o f relevant services, and changes in labor legislation and in other labor market institutions. In addition, findings of the report are based on a few other surveys conducted in 2005 by AVAG Solutions with support from the World Bank, to derive additional data on the labor market, as follows: (a) a survey o f unemployed youth aged 16-24 registered at the SES; (b) a survey o f students studying at primary vocational and secondary specialized education establishments; (c) a survey o f rural and agricultural employment; and (d) a survey o f managers o f public employment services. Valuable insights on labor migration were obtained from the survey o f labor migration from Armenia during 2002-05, conducted in 2005 by Advanced Social Technologies with support from the Yerevan office o f the Organization for Security and CooperationinEurope. 4 CHAPTER I:MACROECONOMIC DEVELOPMENTSAND THE LABORMARKET Despite impressive economic growth rates, Armenia has yet to experience growth in employment. Typical of many other Europe and Central Asia countries, workers displaced in the "old" sectors have hadfew chances tofind jobs in the "new sectors. Thus, the scarcity ofproductive job opportunities and the underuse of labor are major socioeconomic problems in the country. From the point of view of labor, of the two main ways to achieve GDP growth-an increase in the number of employed or improvements in productivity-in Armenia, GDP increased mainly due to the latter. Labor shedding in the "old" sectors has also been impressive. There is a clear trend of decrease in the share of agriculture and manufacturing, and of increase in the share of the services sector, which, for the last 10 years, is characteristic of countries in all income groups in almost every part of the world. I n this regard, Armenia, as a country in transition in the group of lower-middle-income countries, has had, as of 2004, an economic structure corresponding to that group of countries. Nevertheless, the trend of the decline in the share of agriculture after its unprecedented increase during 1990-94 generally corresponds to global developments. Thus, in the near future, we can expect a further decline in the share of agriculture and manufacturing in GDP, and continuous increase in the share of services; stabilization and certain growth in nonagricultural employment, which will happen mainly at the expense of the absolute and relative decline in agricultural employment; and stabilization in overall employment and its future growth of 0.I to 0.5percent annually. 1.1. EconomicGrowthandEmployment In the last decade, Armenia has been one of the fastest-growing transition countries. Moderate but robust economic growth in the initial years o f the recovery (5 percent on average during 1994-2000) accelerated inthe late 199Os, reaching an average annual rate o f 12.3 percent between 2001-2005 (NSS 200%; www.armstat.am). Armenia is also one o f the first Commonwealth o f Independent States (CIS) countries in which the volume o f gross domestic product (GDP) recovered in 2003 to its pre-transition level (Table 1.1). In recent years, the most rapidly expanding sectors o f the economy have been construction, which almost tripled its volume o f output during 2000-03-an increase o f 2.8-fold in constant prices; followed by trade and catering-an increase o f 51 percent; and industry-an increase of 38 percent. Agricultural output increased for the same period by 20 percent. GDP growth has been driven by total factor productivity, reflecting efficiency gains from macroeconomic stabilization and structural changes, including reallocation o f workers into higher value- addedjobs. 5 Table 1.1: Volume Indexes of Gross DomesticProduct (constant prices), 1991= 100 1992 1995 2000 2002 2003 2004 2005 2006"" Armenia 58.2 59.8 76.8 95.3 108.5 120.0 136.8 155.1 Azerbaijan 77.4 42.2 59.3 72.0 80.1 88.3 111.6 150.1 Belarus 90.4 66.1 89.7 98.7 105.6 117.2 128.4 141.1 Georgia 55.1 35.8 47.5 52.5 58.4 61.8 67.6 ... Kazakhstan 94.7 69.0 78.0 97.2 106.3 116.5 127.5 141.0 Kyrgyz Rep. 86.1 55.0 72.2 76.1 81.2 87.1 96.6 99.2 Moldova* 71 48 42 48 51 55 59 61 Russia 85.5 65.4 70.8 77.9 83.6 89.6 95.4 101.8 Tajikistan* 70 41 41 50 55 61 65 70 Turkmenistan ... ... ... ... ... ... ... ... Ukraine 90.1 52.3 47.4 54.4 59.6 66.9 68.6 73.4 Uzbekistan 88.9 81.6 98.6 106.8 111.5 120.1 128.5 137.9 * -CIS average" 86.0 63.1 68.3 76.3 82.2 88.9 94.9 102.0 Estimate;** -calculated by the author. Source: CIS STAT 2007 (http://www. cisstat.com/eng/mac-01.htm). In the transition period, several distinctive periods of development can be identified. The early 1990s was a period o f economic collapse and hyperinflation, and a drop in real output accounted for 60 percent, following the regainingo f independence in 1991. Economic growth was resumed in 1994. The early stages o f the recovery, during 1994-98, were driven by one-time factors (land reform, energy) and growth in enclave industries (mining, metallurgy) that had a limited impact on the broader economy. At the same time, privatization o f state enterprises shed workers that the gradually expanding formal private market was unable to absorb. Since 2000, the government's reform policies have begun to make themselves felt, and the country's economic performance has further improved. A recent International Monetary Fund study uses an endogenous growth model and cross-country regressions (covering 50 countries) to explore the determinants o f Armenia's recent rapid growth and the prospects for sustaining this growth in the future (Iradian 2003). Three main factors explain the strong growth performance: first, the catching-up process following the sharp decline in output in the early 1990s (estimated to account for almost a third o f the average annual growth during 1994-2001 o f CIS countries inthis situation); second, sound macroeconomic policies(reflected inrelatively low inflation and small fiscal deficits); and third, the large private transfers inrecent years (reflected in a high investment rate-much higher than the national savings rate). The boom i s also characterized by a dramatic rise in exports, which have grown by 13 percent per year since 1995, including 26 percent on average during 2000-03. The export boom has been fairly broad, with most significant export branches experiencing rapid growth. Exports o f precious stones and metals, however, still account for more than 40 percent o f total exports (World Bank 2005e). Investment has also boomed, growing at 21 percent annually since 2000. The FDIincreased by 46 percent in2004 compared with 2003. (IMF 2005). Private consumption growth has been lower, but still impressive, at 8 percent per year. This positive macroeconomic performance has already fundamentally altered Armenia's economic situation. 6 However, there exist serious structural problems, such as narrowly based growth, poorly functioning credit market, biases against small and medium-size enterprises (SMEs), and inadequate publicinvestment ininfrastructure.' The new sector, comprising small firms, has yet to develop the critical mass needed to generate enough jobs to offset job losses in the old sector. Facilitating the entry o f new firms and the growth o f small enterprises i s hence a prerequisite for faster job creation, employment growth, and eventually the reduction o f unemployment. Given the country's factor endowments and geographical situation, its potential comparative advantage is in high value-added and skill-demanding, labor intensive, light-weight sectors (World Bank 2002b). The Armenian economy during the Soviet period, especially starting from the 1970s, had a rather significant share of the informal economy. In addition, a considerable volume o f non-reported economic activities in the form o f different services rendered to the population was performedby individuals. Both types o f activities were considered illegal. In 1987, legal private enterprises, mostly in the form o f cooperatives, started to emerge in different sectors. According to National Statistical Service (NSS) estimates, in 1990, private sector value added accounted for 11.7 percent o f GDP. Figure1.1: Structureof Value Added and Employment,1990-2005 1Public infrastructure has largely benefited from capital inflows from the Lincey Foundation to reconstruct roads and other public infrastructure, and investments associated with the celebration of 1,700 years o f Christianity in2001. 7 The NSS estimates that, currently, 30 to 35 percent of total GDP i s in the shadow economy.' The World Bank Doing Business Report 2006 estimates that inArmenia the share o f the informal economy in gross national product i s as high as 46.3 percent, the fourth- highest levels among the transition countries, following Georgia (67.3 percent), Azerbaijan (60.6 percent), and Belarus (48.1 percent) (World Bank 2005a). During 1990-2005, initially due to the systemic crisis of the economy during 1990-93, and the subsequent continuous economic growth during 1994-2005, Armenia's economy was impacted by major structural shifts. The structure o f GDP and employment changed significantly. The importance o f the primary and secondary sectors in the economy and inthe total employment rate is decreasing, and the role of the tertiary sector i s increasing. Figure 1.1 reveals that the share o f the services sector was 33.8 percent of GDP in 2005, compared to 25.8 percent in 1995 and 30.8 percent in 1990. This i s still much less than the over 60 percent in new European Union member countries. The share o f agriculture in GDP dropped to 20.9 percent in2005, compared to 42.3 percent in 1995 and 17.2 percent in 1990.3 At the same time, the share o f the primary sector (industry and construction) amounted to 45.3 percent o f GDP in 2005, compared to 32.0 percent in 1995 and 51.9 percent in 1990 (NSS 200%; the share o f employment by sectors see Annex Figure l).). Strong economic growth in Armenia in the last decade was not accompanied by an increase in employment. Like many other transition countries, Armenia has faced "jobless growth" or actually even "job loss growth." As recorded by the NSS, employment has declined over the last 15 years by 32 percent-from 1.630 million employed individuals in 1990 to 1.108 million in2005 (CIS STAT 1994; NSS 2006). However, there are indications that the labor market situation might have reached the turning point. The NSS reports an increase in the number o f employed population from 1.082 million individuals in 2004 to 1.112 million on average in January-December 2006. Estimates based on the Integrated Living Conditions Survey (ILCS) show among the population 16 and over an increase inthe employment rate: from 46.0 to 48.7 percent and a decrease in the unemployment rate: from 27.0 to 19.3 percent between 1998/99 and 2004 (World Bank and NSS 2006). The number o f registered unemployed job seekers declined from 147,000 at the end o f 2003 to 88,900 in December 2006. The NSS adjusts GDP estimates for unreported activity by using the so-called "Italian method." Firms' reported employment and output are used to calculate labor productivity by sector o f the economy. Results o f a labor survey that is done once a year are used to estimate levels o f employment by sector; the labor survey is believed to much more accurately capture true employment than employment reported by firms. The shadow economy i s then estimated in monetary terms by sector by multiplying estimated labor productivity by unreported employment (the difference between survey and firms' reported numbers.). Calculated inbasic prices, that is, without net indirect taxes (indirect taxes less subsidies). 8 Figure 1.2: Dynamicsof Main Economic Indicators, (1991 = 100) 160 140 120 100 80 +Agricuturalproduction 60 40 I R 20 m 0 4 I 1991 1992 1995 2000 2002 2003 2004 2005 2006 Source: NSS 2007 (www.armstat,am); CIS STAT 2007 (www.cisstat.com). Duringthe transition years, the dynamics and structure of employment by main sectors of the economy has been highly uneven (Annex Table 1; Annex Figure 2). The number o f employed inindustryand construction declined from 679,000 in 1990, or 41.6 percent o f the total employment, to 169,200 in 2005, or 15.4 percent o f total employment. At the same time, the number o f employedinagriculture and forestry increased from 289,000 to 507,600, or from 17.7 percent to 46.2 percent o f the total employed (CIS STAT 1994; NSS 2006). The ratio o f employed inthe service sector has been relatively stable, but the sector experienced a rapid decline inemployment as well. Recent years have seen a noticeable shift of employment from the state sector into the private sector. Total employment in the private (non-state) sector was estimated at 80 percent in 2005-an increase from 21 percent in 1990 (CIS STAT 1994; NSS 2006). Employment i s shrinking inthe public services sector and inthe civil service, and their share intotal employment has declined from 27 percent in2000 to 21 percent in2004. Science has experienced the most notable reduction in employment, from 53,000 employees in 1990 to 14,000 in 2005 (Table 1.2). During 2003-05, as part o f the reforms inthe education system, 37 schools were merged or closed and more than 6,900 teachers in general public schools were made redundant (almost 16 percent o f the total), following the optimization o f the student-teacher ratio (from a low 11.7 in 2003 to over 14 in 2005), class size, teacher loads, and other indicators. The objectives o f the reform are achieving an effective use o f public financing, increasing the salaries and living standards o f teachers, and improving school performance. Teacher salaries increased by 65 percent in2005 (IMF2005). 9 Table 1.2: Employment Dynamics inHealth, Education, Science, And Public Administration, 1995-2005 (1000') 1990 1995 1997 1998 1999 2000 2001 2002 2003 2004 2005 Health Care, 94.0 85.5 80.7 77.7 76.6 80.7 79.3 66.9 60.3 49.8 50.6 Physical Culture, and Social Security Education, Culture, 195.7 180.2 161.3 154.8 153.4 153.9 155.2 138.3 132.2 120.6 119.0 and Art Science and 53.1 22.8 16.0 7.3 15.8 15.1 14.6 13.2 12.3 12.2 14.0 Scientific Services Government Staff 27.8 29.3 29.5 29.1 28.2 27.4 25.2 23.8 25.0 26.6 28.2 Source: NSS: Statistical Yearbooks (relevant years); www.armstat.am. Economic growth in Armenia so far has taken place almost exclusively through labor productivity increases, paired with job redundancies. In the last decade, two distinctive periods-1 995-99 and 2000-03-can be identified, where the latter period witnessed a marked increase in economic growth and labor productivity growth, with the relative stabilization in the number o f job cuts (see Table 1.3). To this end, 2002 may be characterized as the year when stabilization in the number of employed in the main productive sectors started. Most o f the labor shedding during 2000-03 took place in the overstaffed social service sectors, where, on average, 9,100 jobs were lost annually, compared with the annual average o f 6,400 job cuts during 1995-99. Table 1.3: Jobs, Value Added, and Labor Productivity inthe Nonagricultural Sector of the Economy, 1995-2003 1995- 1995- 2000- 2003 2003 1999 2003 Annual Average Increase o f Value Added in 34.38 15.21 58.35 96.79 Nonagricultural Sector, Billion 1995 Constant Drams Annual Average Change inEmployment in -42.7 -49.5 -34.3 -7.0 Nonagricultural Sector, Thousand Annual Average Labor Productivity inNonagricultural 0.538 0.393 0.718 0.966 Sector, Million 1995 Constant Drams Annual Average Increase o f Value Added inIndustry, 6.92 2.05 13.01 17.08 Billion 1995 Constant Drams Annual Average Change inEmployment inIndustry, -24 -32 -14.1 -4.3 Thousand Annual Average Labor Productivity inIndustry, Million 0.773 0.555 1.046 1.339 1995 Constant Drams Annual Average Increase o f Value Added inConstruction, 12.195 3.331 23.27 44.89 Billion 1995 Constant Drams Annual Average Change inEmployment inConstruction, -6.62 -8.64 -4.1 1.1 Thousand Annual Average Labor Productivity inConstruction, 1.511 0.734 2.48 1 3.924 Million 1995 Constant Drams Annual Average Increase o f Value Added inTrade, 10.336 8.087 13.146 16.163 Billion 1995 Constant Drams Annual Average Change inEmployment inTrade, 4.59 9.1 -1.05 5.4 Thousand Annual Average Labor Productivity inTrade, Million 0.712 0.538 0.93 1.16 1995 Constant Drams Source: National Statistical Service and authors' calculations. 10 1.2. Job Flows, by Industry The transition from the policies in the early 1990s of artificial hoarding of excess labor by low wages, extensive administrative leave, and part-time work, in order to avoid widespread unemployment under the conditions of severe economic crisis, to the free- labor-market policies was made possible by the following factors: Wide-scale privatization of medium- and large-size SOEs that started at the beginning o f 1995 was largely completed in 2003. As o f July 2004, 1,894 medium- and large-size SOEs, or 82 percent of all medium- and large-size enterprises, and 7,226 small enterprises, or 97 percent o f all small enterprises, were privatized. Only a small number o f firms in infiastructure are still controlled by the State or municipal governments. The lack o f investor interest inthe privatization o f the remaining SOEs i s mostly due to their non-competitiveness and obsolete technologies. H Increase of budgetary discipline and gradual phasing out of barter trade and payments via compensation of debts, due to which barter trade was practically eliminated by the end o f 1996. The payment mechanisms via compensation o f debts were rather common practice in the early stages o f economic reform, mostly due to the nonpayment or late payment on the obligations to suppliers o f goods, works, and services, when accumulated tax arrears o f these economic agents were cleared against these payments. In 2001, this practice was substantially restricted, and starting in 2002, the government has not been usingthe clearing mechanisms at all. Starting in 1998, tax arrears o f economic agents inArmenia can be written off only by adoption o f a corresponding law. 9 Gradual phasing out of direct and hidden subsidies to enterprises, followed by privatization o f energy distribution networks in 2002, which brought down quasi- fiscal subsidies to the industry and other productive sectors from 1.2 percent o f GDP in 1997 to 0.3 percent of GDP in 2001 (Freinkman, Gyulumyan, and Kyurumyan 2003). Improvements in the business and investment climate, namely, market entry, contract enforcement, and market exit. (See Chapter VII.) During 1995-99, privatizationwas the mainfactor responsible for the emergence of the private sector, which currently accounts for about 75 percent of GDP, and which provided, in 2005, 63.2 percent of all jobs in the nonagriculturalsector of the economy4 (Figure 1.3). Starting in 1999, when the first and biggest wave o f voucher privatization was mostly completed, the new private enterprises gradually replaced privatization as the main factor responsible for the growth o f a private sector. 4In 2005, 94.3 percent of jobs in industry, 95.1 percent of construction jobs, and 99.8 percent of trade and cateringjobs, were inthe private sector. 11 Figure 1.3: NonagriculturalJobs inthe State and Private Sectors, 1995-2003 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Jobs in state sector Jobs in private sector Source: National Statistical Service and authors' calculations. Meanwhile, privatization was also the factor responsible for the main bulk of job redundancies in production sectors of the Armenian nonagriculturaleconomy. Transfer o f the SOEs to private ownership, in parallel with elimination or substantial restriction o f hidden subsidies for the most privatized enterprises, and an increase in financial discipline, created an issue o f downsizing the excess labor. A considerable part o fprivatized enterprises, mostly large and medium-size manufacturing enterprises, turned out to be noncompetitive, lost their markets in the former USSR, were subsequently closed down, or were forced to operate only with the small fraction o f their capacity. This increased the magnitude o f job cuts even more. However, only a small fraction o f these enterprises underwent significant restructuring through downsizing, division, and profile change. Others served as a base for the emerging private enterprises via assets stripping and liquidation. A considerable percentage o fthese firms remainnon-operational. Labor shedding in the state nonagricultural sector has not been compensated by the corresponding increase in employment in the private sector. On the contrary, during 1995-99, about 375,000 jobs inthe state sector, or 52 percent, were lost, while the number o f private sector jobs increased during the same period by 191,500, or only about half o f the jobs cut in the state sector. Starting in 2000, when the entry o f new private companies into markets somewhat intensified, more jobs were created in the private sector to replace labor shedding in the state sector, and in 2003 an increase in private sector nonagricultural jobs almost compensatedjob redundancies inthe state sector (the difference was only 7,000 jobs). 12 Figure 1.4: Job Creation and DestructioninIndustry, 1999-2003 (Inreporting enterprises) 5 0 11Jobs created in industry -5 Jobs elilninated m v) of which a)"formal"jobs E0 3 -10 b)actualjobs Jobs elilninatedin public sector -15 -20 -25 1999 2000 2001 2002 2003 I Source: National Statistical Service and authors' calculations. In the production sphere, for first time in 13 years in the history of an independent Armenia, in 2003 the number of jobs created in the private sector (15,100) surpassed the number of job cuts in the state sector (8,900), and a total increase of 6,200 jobs was achieved. (See also Annex Figure 3). Inpublic services and insome other sectors where state sector employment exceeds 50 percent, labor shedding will continue in the medium term. This is due to the currently ongoing reforms in health, education, and public governance aimed at optimizing the number of state-owned entities (such as schools, hospitals, and government agencies) and employment in these sectors in order to increase quality o f provided services andproductivity, and to provide accelerated salary increases. Job creation in Armenian industry in recent years was mostly concentrated in new private enterprises (Figure 1.4; Annex Table 2; Annex Figure 3). Out of the 26,500 jobs created in industry during 1999-2003, 22,000, or 83.4 percent, were established in newly created private enterprises. Job elimination was mostly concentrated in"formal" jobs, that is, inoperating enterprises, and the people on administrative leave were dismissed first. Total job elimination during 1999-2003 was 98,600, o f which 76,600, or 77 percent, were due to staff reduction. Out o f the 76,600 personnel made redundant, 34,800, or 45.4 percent, were in the state sector, which resulted inthe decline o f the share of personnel employed in SOEs in industryfrom 34.9 percent in 1999 to 16.5percent in2003. Job creation in construction during recent years was also concentrated innew private enterprises (Figure 1.5; Annex Table 3; Annex Figure 5). All o f the 26,500 jobs created during 1999-2003 were in the private sector. Job elimination, due to the relatively small 13 number o f personnel on administrative leave in the sector, was mostly the result of accelerated productivity growth. Out o f 31,400 personnel cuts for the same period, 23,500 employees, or 74.8 percent, were inthe public sector. Figure 1.5: Job Creation and Destruction in Construction, 1999-2003 (Inreporting enterprises) 10 5 Jobs created -0 0 c Jobs eliminated m w e=I 0 of which a) "formal" -5 b) actual ~t Jobs eliminatedin public -10 sector -12.3 -15 1999 2000 2001 2002 2003 mrcc: National Statistical Service and authors' calculations, F o r short-term estimates of employment dynamics in different sectors, Cobb-Douglas production functions for 1995-2003 can b e used, with ordinary elasticity o f return, calculated for industry (without energy sector) and its main sectors (see Table 1.4). According to the production function parameters, only in the food-processing industry will an increase o f employment result in output increase, whereas in other sectors, the increase in the employed tends to negatively affect output. However, as labor elasticity coefficients show, the reserves for staff reduction are negligible inthe timber industry, and rather small in the building materials industry. According to the production function for metallurgy, labor elasticity i s negative. In reality, taking into account narrow profit margins in nonferrous metallurgy, due to high transportation charges, the level of output in this sector, and consequently the level o f employment, depends on the level o f internationalprices on copper, molybdenum, and gold concentrate^.^ The quality o fproduction function estimates for machinery, lightindustry,and other industrial sectors is rather low due to the absence o f dynamics inthe output o f the first two sectors, and low reliability o f the data on other sectors. In our opinion, substantial parts o f light industry and the machinery sectors can be described as depressed sectors, where the dynamics o f output depends o n nonpermanent factors, such as one-time orders, which usually are not renewed insubsequent years. 14 Table 1.4: Parameters of Cobb-Douglas ProductionFunction for Industry,by Sector, 1995-2003 Intercept Output- Output- R(2) F-Statistics Fixed Labor Capital Elasticity Elasticity Industry (excluding energy) 1.163 0.707 0.293 0.971 265.202 Standard error of estimate 0.035 0.043 Metallurgy 0.237 1.571 -0.571 0.719 21.510 Standard error of estimate 0.596 0.339 Manufacturing 0.730 1.583 -0.583 0.944 135.694 Standard error of estimate 0.074 0.136 Machinery 2.443 0.880 0.120 0.357 5.446 Standard error of estimate 0.526 0.377 Chemical Industry 1.359 3.737 -2.737 0.883 61.269 Standard error of estimate 0.143 0.477 Timber Industry 1.365 1.065 -0.065 0.403 6.391 Standard error of estimate 0.370 0.421 Building Materials 0.842 1.161 ' -0.161 0.704 20.023 Standard error of estimate 0.134 0.259 Light Industry 2.376 1.157 -0.157 0.449 7.524 Standard error of estimate 0.722 0.422 FoodProcessing 2.326 0.584 0.416 0.987 591.131 Standard error of estimate 0.029 0.024 Other Sectors 0.067 2.046 -1.046 0.392 6.158 Standard error of estimate 1.264 0.824 Source: Authors' calculations. Based on these data, in the near future, employment in the food-processing, nonferrous metallurgy (provided that the favorable dynamics of relevant international prices continue), timber, and building materials sectors, while taking into account the continuation of strong growth inconstruction, i s expected to increase as well, whereas inmost o f the sub-sectors of the machinery, light, and chemical industries, growth will probably not create additional employment, due to the still significant reserves o f excess labor and lower productivity levels compared with the other industrial sectors. Taking this into account, and provided that strong economic growth in industry continues, employment stabilization or growth o f employment inindustryinthe range of0.1to 0.5 percent annuallymaybeanticipated inthe near future. The other characteristic of industrial development i s the ongoing process of an increase in the number of enterprises in industry in line with the downsizing of enterprises. During 1990-2003, the number of registered enterprises operatinginindustry increased 2.7- fold. Data on registered enterprises are available from 1995, and the number o f registered industrial enterprises increased more than 3-fold during 1995-2003 (Figure 1.6). The biggest increase took place during 1995-97, that is, starting from the beginning o f a large-scale 15 privatization o f SOEs. The annual average increase in the number o f registered industrial enterprises was 1,550 during 1996-98, whereas during 1999-2003 it was only 391 enterprises. Figure 1.6: Number andSize of IndustrialEnterprises inArmenia, 1996-2003 (1995 = 100) 400 ~~ 200 300 - 200 100 0 1996 1997 1998 1999 2000 2001 2002 2003 ~ZZEEZ!Registeredenterprises mReportingenterprises 0Non-reportingenterprises +Average size of registeredenterprises (right scale) Average size of reportingenterprises (right scale) =Averagesize of non-reporting enterpises (right scale) )urce: National Statistical Service and authors' calculations. The average size of industrial enterprises declined from 110 employees in 1995 to 15 in 2003, a more than 7-fold reduction. (Annex Table 4). For reporting enterprises, there was a 12.7-fold decline inthe average size compared with 1990, whereas output declined only 3.6- fold. During the privatization in 1995-98, the average enterprise size declined from 298 employees to 152 employees, or 48.3 percent, whereas in the post-privatization period o f 1999-2003, the average enterprise size declined from 122 employees to 41 employees, or about 3-fold. This was mostly due to the intensification of post-privatization restructuring and entry o f new private, but much smaller, enterprises. The other employment characteristic i s the dramatic decline in the number o f personnel on administrative leave due to labor shedding. Out of total labor shedding inindustry, in 1996, 128,600 of those dismissed were on administrative leave, but by 2003, only 2,500 o f those dismissed were on administrative leave. 16 Figure 1.7: Main Indicators for Agriculture inArmenia, 1990 -2004, Percent 160 I t 140 120 100 80 60 40 20 0 1990 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 0Laborproductivityinagriculture-to-laborproductivityinnon-agriculture - II Share of agriculture inO P Share of agriculture intotal employment Labor income inagriculture-to-wages innon-agriculture Source: National Statistical Service and authors' calculations. Employment in agriculture in Armenia i s currently one of the highest among the transition countries. (Figure 1.7; Annex Table 5). The trend o f increase in agricultural employment was characteristic of a number o f countries inthe region, due to the depth and severity o f the transition crisis. It served as one of the mainmeans for cushioning the blow o f hardships.The main causes of highlevels o f agricultural employment inthe country are low productivity and the limitednumber o f newly created nonagriculturaljobs. (See Chapter IV.) Table 1.5 shows that inall the listed countries, with the exception o f Estonia and Tajikistan, a relative decline in agricultural productivity occurred, due to an increase in the share o f the employed, or the faster growth in other sectors.6The concentration o f surplus labor force in agriculture, on the other hand, i s one o f the main causes o f low productivity inthat ~ e c t o r . ~ The reasons were different in the two countries: in Estonia, a relative decline in agricultural productivity was due to a sharp increase in productivity and a 3-fold decrease in the share o f employment, and in Tajikistan it was due to a reductioninproductivity and the unchanged high share o f employment inagriculture. Our comparison of 75 countries during 1990, 1995,and 2001 reveals that there is a direct negative relationship between the share o f those employed ina riculture and productivit .The relationship is expressed as follows: Ln(LPA1990) = 7.1136- 0.5652 Ln(ESA1991) R2= 0.359, F= 42.41 SE= z031 1990 LnhPA2001]= 7.639 -0.6166 Ln(ESA 20 1) R2 - 0.545, F= 89.8 SE=0.825 LnPA1995 =7.321 -0.5834Ln(ESAl9954 R2=0.436, F=58.19 SE=0.952 1995 2001 where LPA is labor productivity inagriculture and ESA is the share o f employment inagriculture. 17 Table 1.5: Dynamics of Agricultural Employment inArmenia And other Transition Countries, 1990-2003 1990 1995 2003 The GDPper %of YOof Produc- Produc- Highest Capita GDP GDP tivity tivity Decline (US (1990) (2003) (%of (%of (YOto Dollars total total 1990 2003) produc- produc- level) tivity, tivity 1990) 2003) Armenia 17.74 37.38 45.96 63 910 17.2 23.50 90.19 51.14 Azerbaijan 30.91 30.64 39.95 60 810 26.5 14.30 85.7 35.79 Bulgaria 18.49 23.85 25.82 16 2130 17.0 10.00 91.9 38.73 Estonia 17.97 9.27 5.79 35 4960 15.6 5.40 86.8 93.29 Georgia 25.59 48.37 54.85 78 830 29.9 20.60 116.8 37.55 Kazakhstan 22.82 22.00 35.03 41 1780 34.9 7.60 152.9 21.70 Kyrgyz Rep. 32.72 47.19 52.61 50 330 32.9 39.20 100.5 74.51 Latvia 16.54 16.73 13.41 51 4070 20.3 4.50 122.7 33.56 Lithuania 18.94 23.72 17.77 44 4490 27.7 7.30 146.2 41.09 Moldova 32.74 48.83 42.88 63 590 24.0 22.50 73.3 52.48 Romania 29.10 40.22 35.63 21 2310 16.45 11.30 56.5 31.72 Russia 13.23 15.70 11.80 40 2610 15.5 5.20 116.9 44.07 Uzbekistan 39.29 40.93 38.50 18 420 33.2 35.20 84.49 91.42 Tajlkistan 43.07 60.07 46.19 50 190 21.7 23.40 50.3 50.66 Ukraine 19.76 22.46 19.73 59 970 24.6 14.10 124.4 71.46 Source: World Bank (1995b); World Bank (2002e); ILO (2004b). 1.3. ProductivityImprovements In Armenia, labor productivity growth outperformed GDP growth, and capital productivity growth was also high. (Annex Table 6). As a result, the productivity gap between Armenia and industrialized countries has narrowed but remains substantial. The import-substitution and export-oriented clusters, in which excess labor was cut the most, achieved higher absolute productivity levels (Figure 1.8). So far the highest labor productivity i s in the food-processing sector, which i s the fastest-developing sector in Armenia, and where the number o f individuals on administrative leave i s limited. In 2003, the food processing sector accounted for 36.5 percent o f total industrial output, whereas the share o f employed in the sector accounted for only 16.1 percent o f those employed. Nonferrous metallurgy accounted for 21.0 percent o f industrial output and 10.3 percent o f those employed. Machinery, chemical, and light industry are, in general, depressed sectors, with a total o f 8.0 percent o f industrial output and 39.6 percent o f industrial personnel in 2003. Together these three sectors, in2003, had a total o f 8,700 personnel on administrative leave, or 56.5 percent o f all industrial personnel on administrativeleave. 18 Figure 1.8: Structure of Output and Employment inIndustry, by Sector, 1990-2003 I 1990 I 1995 1 1996 I 1997 I 1998 1 1999 I 2000 I 2001 12002 1 2003 I 0 Foodprocessing 0 Lightindustry 0 Chemical,timber & buildingmaterials lilMachinery EI Metallurgy Energy Source: National Statistical Service and authors' calculations. Analysis shows that there i s no direct link between productivity increase and labor shedding. On the contrary, the industrial sectors with relatively small numbers o f layoffs have the highest productivity. For example, in food processing, 17,000 personnel were cut during 1990-2003, or 59.4 percent, whereas labor productivity was the highest in the industry. Inmachinery, where there were 152,600 personnel cuts during 1990-2003, or 88.9 percent, labor productivity was about 10.7 percent o f the labor productivity in food processing. In light industry, there were 92,500 personnel cuts during 1990-2003, or 90.2 percent, but still labor productivity in light industry remained the lowest o f all the industrial sectors. During 1995-2003, in manufacturing (that is, industry excluding energy and metallurgy), the main tendency was that in industries with relatively small enterprises, the output per enterprise tended to be higher that in the industries with relatively larger enterprises. Panel regressions o f the output o f the enterprises in the sectors o f industry revealed the dependency of the output on the size of enterprise-the smaller the average size of the enterprise, the higher the output8 (Table 1.6). Downsizing via restructuringand creation o f new enterprises may be considered as the main tendency for the development o f the industrial sector and labor productivity increase during 1999-2003, * Panel regression for 7 manufacturing sectors, including 63 observations for 1995-2003, showed the negative relation between the average output and the size o f enterprise: OUTPUT = 151.65 0.527 SIZE; R(2) = 0.137; - F = 10.88; SEfor intercept - 18.87; SE for dependent variable - 0.159. 19 Table 1.6: Output and Labor ProductivityinIndustry, 2001-2003, By EnterpriseSize 2003 2002 2001 Output inReportingEnterprises,Billion2003 ConstantDrams 372.87 319.72 275.68 of which: Microenterprises (5 and less employees) 7.38 10.19 13.21 Small enterprises (6-50 employees) 41.16 32.88 30.90 Medium-size enterprises (51-100 employees) 15.75 12.00 10.10 Large enterprises (101 and more employees) 308.58 264.65 221.47 including: large enterprises in manufacturing 138.57 103.67 61.59 AnnualAverage Personnel,Thousand Microenterprises (5 andless employees) 2.63 2.20 1.84 Small enterprises (6-50 employees) 11.60 11.26 10.07 Medium-size enterprises (5 1-1 00 employees) 7.60 7.25 8.33 Large enterprises (101 and more employees) 56.72 51.10 51.34 including: large enterprises in manufacturing 31.03 24.27 27.02 Number of Enterprises Microenterprises (5 and less employees) 1045 884 811 Small enterprises (6-50 employees) 718 649 605 Medium-size enterprises (5 1-1 00 employees) 101 100 115 Large enterprises (101 and more employees) 189 182 190 including: large enterprises in manufacturing 72 69 83 Average Size of Enterprises Microenterprises (5 and less employees) 2.52 2.49 2.27 Small enterprises (6-50 employees) 16.16 17.36 16.65 Medium-size enterprises (5 1-100 employees) 75.24 72.46 72.44 Large enterprises (101 and more employees) 300.11 280.75 270.19 including: large enterprises in manufacturing 430.99 351.71 325.49 Labor Productivity inReportingEnterprises,Million 2003 Constant 4.75 4.45 3.85 Drams of which: Microenterprises (5 and less employees) 2.81 4.63 7.19 Small enterprises (6-50 employees) 3.55 2.92 3.07 Medium-size enterprises (51-1 00 employees) 2.07 1.66 1.21 Large enterprises (101 and more employees) 5.44 5.18 4.3 1 including: large enterprises in manufacturing 4.47 4.27 2.28 Output in Non-reportingEnterprises,Billion2003 ConstantDrams 50.85 47.45 44.72 EmploymentinNon-reportingEnterprises,Thousand 55.10 48.40 60.50 Labor Productivity in Non-reporting Enterprises, Million2003 0.92 0.98 0.74 ConstantDrams Share of Non-reporting Enterprisesin Employment, Yo 41.23 40.26 45.82 Share ofNon-reporting Enterprisesin Output, YO 12.00 12.92 13.96 Source: National Statistical Service and authors' calculations. 20 The ratio of labor productivity-to-GDP per capita has increased in the entire economy (Figure 1.9).The only exception i s agriculture, which in 2003 constituted only 68.7 percent of the value of the corresponding indicator for 1990. The largest increase in the ratio was recorded in sectors in which labor force reductions were the most intensive. In -particular, productivity in construction increased four-fold and in industry two-fold, more than the increase inGDP per ~ a p i t a . ~ Figure 1.9: Ratio of Labor Productivity-to-GDP Per Capita inArmenia, 1994-2003 (1990 = 100) 450 200 400 180 350 160 300 140 250 120 200 100 80 150 60 100 40 50 20 0 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 hdustry (left scale) Construction(left scale) Trade (left scale) Transport and comnications (left scale) =Total(rightscale) --@-Agriculture (right scale) kn-agriculture (right scale) `ouvce National Statistical Service and authors' calculations. Moreover, during 1990-2003, an increase in ratio of productivity-to-GDP per capita was recorded in Armenia, contrary to all other countries presented in the table, where the ratio declined" (Table 1.7). Analysis o f economic development for 1990-2003 in Armenia allows us to distinguishtwo factors causing the above-mentioned phenomenon. The first one i s linked to the deep economic crisis o f 1990-93 and the subsequent structural reforms in the economy, as a result of which, due to the widespread use o f administrative leaves and the sharp decline in wages, the phenomenon o f hidden employment was widespread in the nonagricultural sector o f the economy. After economic growth was reestablished, especially after the implementation o f wide-scale privatization, these "fictitious" jobs were eliminated. 9 The ratio o f labor productivity-to-GDP per capita was used to ensure the comparability o f data for the subsequent years with data from 1990. lo unchangedratio ofproductivity to GDP per capita, all other factors being equal, can be interpreted as The economic growth with no change in employment. Lowering o f the value o f that ratio shows that economic growth occurs inparallel to increase inemployment, and higher values show that economic growth i s occurring inparallelto the reductioninemployment. 21 Table 1.7: Labor Productivity: InternationalComparisons Total Industry Agriculture GDP Shareof Ratio of Per Capita Agricultural Labor Employment, Productivity YO to GDP Per Capita Armenia 1990 3579.73 3532.88 3196.12 1632.49 17.74 2.19 Armenia 1995 871.60 1030.82 948.68 341.66 37.38 2.55 Armenia 2003 3610.98 5731.25 1675.32 1245.10 45.96 2.90 Brazil 1990 5035.50 9658.00 3077.00 2680.00 22.80 1.87 Brazil 1995 5303.31 11855.00 3532.00 3640.00 26.10 1.46 Brazil 2001 5430.94 13483.00 5176.00 3070.00 20.60 1.76 China 1990 795.19 2696.00 516.00 370.00 53.50 2.15 China 1995 1090.83 5280.00 665.00 620.00 48.60 1.75 China 2001 1526.65 7050.00 806.00 890.00 46.9 (2000) 1.71 Czech Republic 1990 6586.65 11526.00 3250.00 3140.00 12.3 2.09 Czech Republic 1995 6387.21 12246.00 8032.00 3870.00 6.6 1.65 Czech Republic 2001 7357.23 15419.00 14377.00 5310.00 4.80 1.38 India 1990 617.92 2589.00 678.00 350.00 69.10 1.76 India 1995 736.57 3373.00 671.00 340.00 66.70 2.16 India2001 902.05 4003.00 755.00 460.00 1.96 Netherlands 1990 33961.29 39013.00 15048.00 17320.00 4.60 1.96 Netherlands 1995 38828.27 44889.00 19897.00 24000.00 3.60 1.61 Netherlands2001 35301.99 50552.00 23137.00 24330.00 2.90 1.45 Poland 1990 5061.99 2167.00 1690.00 25.20 2.99 Poland 1995 6377.88 2575.00 2790.00 22.60 2.28 Poland 2001 9041.60 2923.00 4230.00 19.00 2.13 Russia 1990 3648.42 2669.00 4160.00 13.90 Russia 1995 2567.07 1933.00 2240.00 15.70 1.14 Russia 2001 2931.39 2170.00 1750.00 8.10 1.67 Source: World Bank (2004b); ILO (2004b). 22 CHAPTER 11: LABOR MARKET TRENDS During the transition years, labor market participation has passed through a profound transition. Low levels of economic activity and high unemployment rates have become a reality. Since social guarantees to the unemployed are quite limited, many individuals are in a transient status of employment, unemployment, and inactivity, and it may be even dgficult to clearly determine the actual status of individuals. The data on labor force participation are controversial. While the ofJicial administrative data on employment, and the LFS data on unemployment, show some stability in labor market dynamics, the ILCS data confirm some improvement in the last year or so. The overall unemployment rate of 20.7percent in 2004, according to the ILCS, and over 30 percent, by the LFS, is still very high by transition economies standards. 2.1. Labor Force Participation Inthe pre-transition (socialist) period, every person of working age was required to be employed. Since vacancies were abundant in most Former Soviet Union (FSU) republics, participation rates were relatively high; on average, 85 percent of the working-age population in the USSR was employed." In the 1980s, at least 1.5 million vacancies were registered by the authorities in the USSR annually; the number o f surplus work places was even estimated to be as highas 15 million to 17 million (Oxenstierna 1990; Kuddo 1995).Although these figures reflect deficiencies inthe planning system o f the Soviet economy and thus contain caveats, the socialist economy based on extensive investment policies was able to absorb a significant number of laborers. At that time, a job was guaranteed to all graduates. Moreover, graduates had compulsory assignments for their first jobs and, as a rule, young specialists had to serve at least three years at their first assignment. However, even at the end of 1980s, in republics such as Armenia, Azerbaijan, Georgia, Kazakhstan, Tajikistan, and Uzbekistan, more than a fifth of the working-age population was not employed in state-owned enterprises (SOEs) or other state institutions. Some were disabled, employed in home economics (especially mothers with many children), or involved in informal activities. Despite huge labor shortages, and in addition to frictional unemployment, disguised underemployment, hidden unemployment, and regional and seasonal unemployment were apparent in the USSR. This was determined by the rapid growth of the able-bodied population in many regions, monostructural economies in certain regions and cities, and deficiencies in investment policies to meet the requirements for newjobs. The employment rates of the able-bodied population in Armenia were particularly low. Only 67 percent of the population aged 16-54/59 reported being employed during the 1989 population census, and 13 percent o f working-age population indicated other sources o f income (other than stipends, pensions, or being dependants) or did not indicate their source l1Working age was defined as 16-54 years for females and 16-59 for males. 23 o f income. Low participationrates inthe 1980s can be partially attributedto the aftermath o f the 1988 earthquake, in which thousands lost not only their jobs but their lives (Table 2.1). Thus, even at the onset o f transition reforms, the labor market situation in Armenia was precarious, and the reforms hit the labor market especially hard. Like in other transition countries, labor supply in Armenia in the last decade has been affected by many demographic factors. In particular, matrimonial and reproductive behavior and migration patterns underwent radical change. Explanations for the worsening demographic situation inthe whole region have been proposed, including economic hardship and the drop in real income reflecting rapid inflation and the elimination o f consumer subsidies; job insecurity; psychosocial stress; weakened social protections, including in health care and child care; and worsening nutrition patterns (see, for example, UNICEF 1995). Lately, Armenia's average life expectancy has tended to increase, but two main changes in demographic behavior have affected or will affect the labor supply in coming years.12 One o f them i s the rapid decline infertility rates, and another i s a significant outflow o f population-permanent or temporary-from the country. (Migration issues will be discussed inChapter VI). Table 2.1: Employment of Population inthe USSR and inSome of the National Republics (including in subsidiary household plots; 1989 population census, in %)' USSR Including: Russia Ukraine Armenia Georgia Azerbaijan A t working age 84.6 86.2 85.5 66.8* 79.2 74.2 Over working age 17.8 18.3 14.4 24.3 32.0 22.6 Males at working age 86.6 87.6 86.4 70.9 84.6 82.0 Males over working age 23.2 23.8 19.7 32.2 39.5 33.5 Females at working age 82.5 84.7 84.6 62.5 73.6 66.2 Females over working age 15.6 16.4 12.4 20.6 28.7 18.0 1. Inthe 1989population census, employment was defined as an economic activity that brought remuneration or income to the respondent (except income from subsidiary plot). Employees who during the census were in training courses, or mothers who were on maternity leave, seasonal workers, or people who worked less than three weeks due to transfer to another job, were also counted as employed. 2. Working-age population = males aged 16-59 and females aged 16-54. * - Thirteenpercent o f the Armenian able-bodied population indicated that they had other sources o f income (other than stipends, pensions, or being dependants) or didnot indicate their source o f income. This may reflect the aftermath of the disastrous Spitak earthquake in 1988 Inthe USSR, the share o f the working-age population ~ ~.~ that did not indicate their source-of income was 0.3 percent. Source: GOSKOMSTATSSSR 1991a. Fertility rates are currently well below the level needed for a full replacement of the current population. The total fertility rates in Armenia have dropped from 2.6 in the mid- 1980sto 1.2 in2001, and 1.4in2004 and 2005 (Annex Figure 7). A rise could occur if some o f the existing barriers and disincentives to family formation and childbearing (related notably to labor market, housing, and child care) were removed. As a result of these changes, the population age structure is significantly distorted. l2Such trends could prove a fact well known to demographers-that inperiods of hardship, some demographic indicators, such as infant or maternal mortality, and life expectancy, can actually improve. 24 Three main cohorts can be identified that are less numerous: first, age group 0-10, reflecting decreasing fertility rates inthe 1990s; second, age groups between 20 and 40, reflectingrapid outmigration; and third, age groups around age 60, reflecting the decline inbirthrates during World War 11. Consequently, while the overall number o f working-age population will increase in the near future, the shnnking number of new labor market entrants will have a significant impact on labor supply and the structure o f the working-age population (Figures 2.1-2.3).13 Limited job opportunities in Armenia have led to discouragement and massive labor force withdrawal, especially among younger and older cohorts and women. The labor force participation rate provides an indication o f the relative size o f the supply of labor available for the production o f goods and services. (Summary data on the labor markets in Armenia and in some other countries are presented in Table 2.2).14 According to the 2004 ILCS data, the economic activity rate o f the population aged 15-64 was 66 percent. This i s higher than the level in Georgia, Hungary, and Poland but less than the level in the EU-15 c~untries.'~However, in Armenia the labor force participation rate i s very high due to extremely highunemployment rates, which are among the highest intransition countries. Figure 2.1: Population Pyramid of Armenia, 2001 Population Census I 40 20 20 MALES FEMALES I l3The first scenario o f the Armenian population forecast is based on the assumption that current fertility(TFR o f 1.176 on average during 1999-2003) and mortality rates will remain stable inthe foreseeable future, while in the second scenario, the TFR will gradually increase to 1.475 during 2011-20, and 1.65 during 2021-30, and age-specific mortality rates will slightly decline. Net migration is assumed to be nilunder both scenarios. l4 Although all the surveys use ILO definitions on employment and unemployment, there is a significant difference among the data from different sources. As noted inBox 2.1, for estimates o f the general level o f LFP rates, the 2001 census and 2004 ILCS seem to be a more reliable source, while the 2004 LFS provides more detailed data on particular characteristics o f the labor force. 15Hungary and the Czech Republic have among the lowest unemployment rates, and Poland and Slovalua have among the hghest inCentral and Eastern Europe. 25 Figure 2.2: PopulationPyramid of Armenia in 2025, Scenario I 40 20 20 40 MALES FEMALES Figure 2.3: PopulationPyramid of Armenia in 2025, Scenario I1 I 40 20 20 40 MALES FEMALES 26 Table 2.2: EmploymentL'opulation Ratios and Activity and Unemployment Rates In2004 in Armenia and Georgia, and in some European Countries Age Armenia Armenia Armenia Georgia Hungary Czech Poland SlovakiaEU-15 2001 LFS ILCS LFS Republic Population 2003 Labor Force ParticipationRates 15-64 72.0 58.2 65.9 60.5 60.5 70.1 64.2 69.7 70.8 15-24 55.3 34.4 39.6 30.3 27.9 35.8 33.9 39.4 48.2 25-54 83.2 71.1 76.8 69.9 77.9 87.8 82.2 88.9 83.9 55-64 51.2 42.8 60.9 69.6 32.0 45.1 31.7 31.7 44.6 Employment/Population Ratios 15-64 45.6 40.5 52.2 52.5 56.8 64.2 51.9 57.0 65.0 15-24 28.7 14.6 22.6 22.7 23.6 28.5 20.0 26.5 40.7 25-54 54.5 51.0 63.6 61.1 73.6 81.4 68.3 74.7 77.8 55-64 37.2 44.3 52.5 64.7 31.1 42.6 28.0 26.8 41.8 UnemploymentRates 15-64 36.8 33.2 20.7 13.2 6.1 8.4 19.3 18.2 8.2 15-24 48.1 57.6 43.0 24.9 15.5 20.4 40.8 32.7 15.6 25-54 34.5 28.3 17.2 12.6 5.5 7.3 16.9 16.0 7.3 55-64 27.2 31.0 13.7 7.0 3.1 5.4 11.6 15.4 6.3 *De jure population. **Employed: including people temporarily absent (up to six months) from the republic for work. Note: In Armenia, the 2001 population census was conducted in October; the 2004 LFS was conducted in August. Source: OECD(2005); Armenia: LFS 2004; ILCS 2004; Georgia:2003 LFS. According to the 2004 ILCS and the 2001 census data, youth (especially young females) aged 15-24 have the lowest participation rates. (See also the 2004 LFS data in Annex Figure 8). Nevertheless, youth participation rates tend to be higher than for the same groups inthe EU-15 and successful transition countries. Typical of other Trans-Caucasus countries, in Armenia participation rates of older workers are higher than in most other transition countries. This i s attributed to high involvement o f older members o f the population in subsistence agriculture. Especially for youth, high labor force participation rates are determinedby significant unemployment rates. More specifically, 43 percent of youth aged 15-24 in the labor force are unemployed (World Bank and NSS 2006; Annex Tables 7 and 8). Duringthe next decade, the labor supplywill significantly increase while the generation of youth born at the peak of high birthrates in the 1980s will enter the labor force (Figure 2.4). The population aged 15-64 is expected to increase from 2.121 million in 2001 to 2.451 in 2015 (according to the first scenario o f population forecast; Annex Table 9). The economy must absorb a significant number o f new entrants in order to maintain the current employment level. 27 Figure 2.4: The Dynamics of the Able-bodied Population of Armenia, 2001-30, by MainAge Groups, '000s 3,000.0 T- -----"~-----~-~~-"--~ ~ ~ ~ ~ - - . ~ ~ - ~ ~-~~~ 2,500.0 - 2,000.0 - +15-64 25-54 1,500.0 *55-64 1,000.0 3C15-24 500.0 Source: Authors' calculation based on NSS data. 2.2. EmploymentRatio The combination of unemployment and labor force withdrawal led to a substantial fall in the employment-to-population ratio. Out of the able-bodied population aged 15-64, only around 52 percent according to the 2004 ILCS data, and 46 percent according to the 2001 population census data, are employed. This indicates the low ability o f the Armenian economy to create jobs. According to the 2004 ILCS and the 2001 population census data, however, the employment rates o f older workers aged 55-64 are higher than in other comparator countries, as i s the casewith labor force participation (Table 2.2). W o r k arrangements in Armenia are among the most diversified and flexible in the transition countries.I6 The number o f workers who are employed in atypical work-hour arrangements i s significant. Based on the 2004 LFS data, 23.3 percent o f the employed worked part time. This i s a much higher share than on average in EU-15 countries, and sevenfold to tenfold the ratio inmore developed Central Europeantransition countries (Table ii). wasmainlyinvoluntarypart-timeemployment: only 12percentofthoseworking This part time were unwilling to take full-time work, 33 percent were unable to take full-time work, 49 percent were unable to find full-time work, and 7 percent worked part time on orders o f their administratiodemployer. Overtime work is very common: 26.0 percent o f the workforce i s working 51 or more hours per week (Figure 2.6). Adjustments in the labor l6 The forms of labor flexibility include external numerical flexibility, or adjustments in the number o f employees; internal numerical flexibility, or adjustments inthe number o f working hours; functional flexibility, when workers' job assignments are modified; wage flexibility, when labor costs and thus wages are adjusted; and externalization, when part of a firm's work i s contracted to other companies (Dijk 1995). Standing (1988) distinguishes four forms o f labor mobility: geographical, employment (or interfii), job (intrafirm), and skill mobility. 28 market and the flexibility o f labor have also taken the form o f flexibility in amount o f time worked (internal numerical flexibility o f labor). Intransition countries, typically adjustments inworking hours are less common, largely explained by the low level o f real wages, which means that workers cannot "afford" to work part time. Figure 2.5: Employment-to-PopulationIndex inArmenia; 1990 = 100 Source: National Statistical Service. The ratio of self-employed and unpaid family workers-29 percent and 13 percent of the total employment, respectively-is also very high by the standards of transition countries. Moreover, according to the 2004 LFS data, for 23 percent o f the employees (hired labor), their work arrangements were based on an oral agreement with their employer- typically without social contribution payments and without any insurance against sickness, work injury,unemployment, or retirement (NSS 2005b). Figure 2.6: Employedby Actually Worked Hours; 2004 LFS upto 9 10-20 21-30 31-40 41-50 51+ hours hours hours hours hours Tource: NSS (2005b). 29 The greatest change since transition has been the shift from stable wages and salaried jobs to casual and less-formal jobs and self-employment. Only around two-thirds o f the employed population have permanent jobs, while one-tenth have temporary jobs, and more than one-fifth participate in seasonal (mostly agricultural) jobs (Figure 2.7). Wages are still the main source o f household income, on average 47 percent o f the total but income fiom self-employment i s also significant, on average 19 percent of the total household income. (Annex Table 10). Figure2.7: Employed, by Gender and Type of Work, inPercent, 2004 LFS 70 60 50 40 30 20 10 0 Permanent Temporary Seasonal One-time, occasional Soz~rce:NSS (2005b). To a large extent these flexibility arrangements are of a forced nature (involuntary) due to very high and fluctuating unemployment inthe labor market, and many workers are not satisfied with their current employment status. According to the 2004 LFS, 48.6 percent o f those employed wanted to change their work situation, for example, to work full time or to have a permanent position. The dominant motivation of 81 percent of those who wanted to change their employment situation was to earn a higher income, followed by the feeling that their work did not correspond to their specialty. Among the self-employed, 63.1 percent were not satisfied with their employment situation. (Annex Table 11). This indicates that a significant portion of employment inthe Armenian labor market cannot be classified as decent work. (See Box I. 1). The labor market is dual in several dimensions, the most striking being between the formal and informal sectors.I7 The estimates by NSS indicate that a significant proportion 17The NSS defines the employed in the informal sector as people who work in an organization or firm without registration as a legal person. Informal employment is typically estimated as a residual: employment in establishments o f more than 10 workers (obtained from establishment surveys) is subtracted from total nonagricultural employment (obtained from labor force or household surveys). Ths method misses workers in large firms who are hired informally without the benefit o f employment contracts or social insurance. 30 o fjobs (40 to 45 percent) are informal sector jobs. These data indirectly reflect the extent o f collapse o f the formal sector o f the economy. The majority o f workers inthe informal sector are in marginalized economic activities and in labor-intensive sectors such as agriculture, construction, retail trade, catering, and domestic services. Self-employment accounts for the bulkofemploymentinthe informal sector, especially subsistence agriculture. Figure 2.8: Employed by EmploymentStatus, 2004 LFS,inPercent 70 60 50 40 30 20 10 0 Employee Employer Self-employed Unpaid familyworker Source; NSS (2005b). The European Union (EU) i s using the concept of "undeclared work," described as "any paid activities that are lawful as regards their nature but not declared to the public authorities" (EU 2004b). This definition excludes criminal activities and work that does not have to be declared. Inparticular, as far as the Law on Mandatory Social Insurance Contributions i s concerned, social contributions are paid by employers, hired employees, and entrepreneurs. Since 2002, people who are engaged in agriculture and in scientific and creative work can be excluded from paying social contributions. According to SSI data, in 2004, 420,100 individuals paid their social security contribution, or 39 percent o f the total employment, by the official statistics. Rural households are obliged to pay only land tax. On the other hand, as for the workforce in the formal sector o f employment is reported for tax purposes, the rate o f tax compliance in Armenia i s the highest among ECA countries: 95.9 percent o f the actual employment that has to be taxed by law i s reported by employers to tax authorities. With respect to the wage bill reported for tax purposes, Armenia has the second- highest ratio-95.6 percent-after Uzbekistan (Figure 2.9). Tax discipline in Armenia i s relatively high, but many workers avoid, legally or illegally, beingtaxed. 31 Figure 2.9: Tax Compliance inArmenia, CIS, EU8, and ECA Countries in2005, Accordingto the EBRD-WorldBank BusinessEnvironment and EnterprisePerformance Survey (BEEPS) I ECA I I' Percent of wage bill reported i I I I I I I for tax purposes Percentof w oruorce II EU8 reported for tax purposes CIS I I I I I I I1i Armenia I I I I I I I I I 80 82 84 86 88 90 92 94 96 98 I Source: EBRD-WorldBank (2005). Nevertheless, the share of labor undeclared for tax purposes or not obliged to report their economic activities i s significant. The main attraction o f the undeclared economy i s financial. This type o f activity allows employers, paid employees, and the self-employed to increase their earnings or reduce their costs by evading taxation and social contributions. The combination o f the minimum wage and the tax and benefit systems should be made sufficiently attractive for people to take ajob inthe formal economy. Furthermore, the issue of undeclared work should be addressed. The size o f undeclared work reduces social security contributions and leads to ahightax burden on registered labor. Underemployment and the increased number of subsistence jobs inthe informal sector are of great concern. This situation means that large parts o f the population are essentially unprotected against economic risk and receive no help from the State. On the other hand, self-employment i s a very important category of employment, both in the sense that displaced workers from old enterprises might find livelihoods working independently, and becauseit may represent the beginnings o f entrepreneurship. Given the current age structure of the population, the Armenian economy should grow particularly fast, with the focus on labor-intensive sectors and activities, to absorb an increasing supply of the able-bodied population, The population forecast for Armenia indicates that in order to maintain the current employment-to-population level at age 15-64 o f 55.9 percent, by 2015, around 120,000 more people should be placed injobs, compared to the current number o f employed. To raise the employment rate to the current EU-15 level of 65 percent, 340,000 new jobs should be created, and 470,000 new jobs are needed to reach the EU Lisbon target o f 70 percent o f employment o f the able-bodied population." Only by 2030 will the number o f able-bodied population decline to the level in2005 (Box 2.1). l8Itis assumedthatthe net migration ofpopulation is nil. 32 Figure 2.10: Forecast of the Dynamics of Employed PopulationAged 15-64, to Maintain the Employment Rate of Able-bodied Population at the Current Levelof 55.9 Percent, or to Raise to the Level of 65 Percent and 70 Percent 1700 1600 Employmentrate 0.65 1500 1400 1300 1200 Source: Authors' calculations. Box 2.1: Employment Targets inEUMember States The EULisbon Summit in March 2000 adopted the employment scenario, "From combating unemployment to activating economically inactive human resources," for the period until 2015, with the aim o f making the EU the most competitive and dynamic knowledge-based economy in the world, capable o f sustainable economic growth with more and better jobs and greater social cohesion (EC 2003a). The strategy that followed had three main dimensions-employment targets, competitiveness, and social inclusion-and was designed to enable the EUto regainthe condition of full employment and to strengthencohesionby2010. EUmember states aim to achieve full employment by 2010 by implementing a comprehensive policy approach incorporating demand- and supply-side measures, and thus raise employment rates toward the Lisbon and Stockholm targets as follows (on average for the EU): - An overall employment rate o f 67 percent in2005 and 70percent in2010 -Anemploymentrateforwomenof57percentin2005and60percentin2010 -Anemploymentrateof50percentforolderworkers(55-64) in2010. National targets in member countries should be consistent with the outcome expected at the EU level and should take into account particular national circumstances. 33 2.3. UnemploymentTrends Armenia's labor market, despite certain notable improvements in recent times, still has a supply and demand gap, which is fairly large by any international comparison. Inthe last decade, the unemployment rate (ILO definition) in Armenia exceeded 30 percent o f the labor force, according to the 2001 population census data and LFS data, which are significantly higher than average indicators for developed and transition countries. The 2004 ILCS survey indicates that the unemployment rate among the able-bodied population i s 20.7 percent, which i s still highby transition economies standards. Only some transition countries in the Balkan region, such as FYR Macedonia, and Bosnia and Herzegovina, have higher unemployment rates (Table 2.2). The unemployment rate has declined mostly among the prime age and older workers, while it remained high among the youth (Table 2.3 and Annex Table 7). Under these conditions, finding employment in Armenia i s much more complicated, especially for young people.l9 Based on the available information, two main periods and trends of unemployment can be identified. Both the overall and youth unemployment rates in 1996-2001 had an increasing trend, followed by a continuous decline during2002-04. Table 2.3: UnemploymentRate Estimatesfrom Different Sources, 1996-2005 (Populationaged 15+) 1996 19981 2004 2004 2005 1999 2001 2001 2002 2003 HHS ILCS LFS Census LFS LFS LFS ILCS LFS Total 29.2 27.0 38.4 36.8 35.2 31.2 31.6 20.7 31.3 Youth (15-24) 46.4 45.9 48.1 60.0 57.6 43.0 HHS =Household Surveys. LFS = Labor Force Surveys, 2001 census data. ILCS = Integrated Living Conditions Survey. Source: NSS; www.armstat.am. By the mid-l990s, the unemployment rate in Armenia was already close to 30 percent. According to the results o f the 1996 household survey o f 10,702 working-age respondents, self-reported unemployment rates (the first data based on the ILO definition) among various age groups inArmenia for 1996 were as follows: 29.2 percent for the active population aged 17-64, 46.4 percent for the population aged 17-24, and 18.8 percent for the population aged 50-64. Also, the first labor force survey was conducted in 1996, according to which the unemployment rate equaled 29.1 percent (Ministryof Statistics 1997; World Bank 1999). The 2001 population is particularly informative with regard to the assessment of unemployment rates and labor supply. This i s especially true for the mainindicators o f the *'The first attempt to quantify unemployment inArmenia was made only in 1996, within the framework o f the household survey. Preceding attempts were more o f a qualitative nature and could not be viewed as reliable assessments o f the unemployment rate. Assessments o f unemployment rates are not available for the years following 1996, and some years have a number o f significantly varying assessments. 34 supply side inthe labor market-for the assessmento f absolute numbers of the economically active population, employed individuals, and unemployed individuals. The census indicated an overall unemployment rate at a high36.8 percent. Most probably the 2001 census data are the most reliable source on the scope o f labor force participation, employment, and unemployment. Figure 2.11: Unemployment Rates at Age 15 and Over, According to the Labor Force Survey Data for 2001-05, inPercent 45 40 35 30 25 20 15 10 5 0 2001 2002 2003 2004 2005 Source: NSS: LFSs of relevant years. The most common reason for becoming unemployed (40 percent) i s liquidation or bankruptcy of enterprises, followed by staff reduction (NSS 2005b). Ninetypercent o fjob seekers are looking for paid work; only 4.0 percent tried to start their own business, o f which 6.0 percent were male and 2.5 percent were female. Most job seekers are desperate in their job search and ready to accept any job with decent wages. According to the 2004 LFS, 57 percent o f the unemployed are seeking any kindo f work with a highsalary; 20 percent o f the job seekers want the work to correspond to their qualifications, and the remaining 23 percent are looking for any kindo f work, irrespective o f salary level. Like in other transition countries, the negative effect of high unemployment rates i s especially damaging for youth. Relative to prime-age workers, youth are especially likely to be more affected by negative output shocks because their productivity i s generally lower due to differences in skills and experience. According to the 2001 population census data, the youth unemployment rate was 48.1 percent, and according to the ILCS 2004 it was 43.0 percent. This i s much above the average for the country (36.8 percent and 20.7 percent, according to the two sources, respectively). Young females are especially vulnerable. Their unemployment rates are almost 50 percent higher than the rates for young males (World Bank and NSS 2006). Also, young, uneducated people are inthe worst situation inthe labor market. This group, with no vocational education and experience, i s so uncompetitive that even considerable improvement in labor market conditions and employment growth may not 35 improve their situation. Unemployment early in a person's working life has been shown to increase the probability o f future joblessness and lower future wages and can lead to alienation, social unrest, and conflict. (Youth unemployment i s discussed indetail in Chapter VI. Another dimension of unemployment i s the trend/dynamics of registered job seekers. The data on officially registered unemployment became available in 1992, after passage o f the Employment Act o f Armenia in December 1991. Two main conclusions can be drawn from these data. First,the economic growth recorded during 1994-99 was accompanied by a rapid increase inthe registered unemployment as a result o f a wide-scale restructuringo f the economy (predominantly in urban areas), reaching its peak in 1999, followed by a clear declining trend during 2000-06. Second, the number o f unemployed adults o f prime age (30 and over) in 2006 was 1.5-fold the figure for 1994 (69,200 and 44,800 individuals respectively in December 2006), whereas the number o f young unemployed, aged 16-29, declined by more than three times (47,000 and 15,400 persons respectively). As a result, during 1994-2006, the share of unemployed youth aged 16-29 in the total number o f registered unemployed declined sharply to around 18 percent in2006, compared to around 49 percent in 1994 (Figure 2.12). Figure 2.12: Officially RegisteredYouth and Adult Unemployment, 1994-2003 (End-year data) 200 180 __ 160 140 c m 2 120 0 L :- E 100 c -0 80 5 60 13 40 C 20 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Unemployed aged 16-29 Unerrployedaged 30 and above burce: National Statistical Service. The pool of unemployed i s relatively highly educated. According to the 2004 LFS, one in eight registered unemployed have higher education (13.3 percent in 2004), and 81.5 percent have secondary specialized or secondary general education (NSS 2005~).The share of young people looking for jobs i s relatively high, but a large percentage are recent graduates o f different types o f schools. Both cases reflect the mismatchbetween the skills and knowledge o fjob seekers, and labor demand, which has altered inthe transition years with developments 36 inthe structure ofthe economy. Surveysshow that the education systemgenerates significant skills inadequacies, which contribute to a very large unemployment rate among recent graduates. (See Chapter V). As result o f this skills mismatch, firms must incur the high cost o f training new hires, which may take several months, without any productive input from these workers. Hidden (concealed) unemployment i s very high, and self-reported unemployment rates are significantly higher than unemployment rates recorded at employment services. According to the 1996 household survey, only 25 percent o f the unemployed were registered at employment services. According to the 2001 population census data, 570,500 unemployed were self-reported, including 562,500 aged 15-64, while the State Employment Service (SES) had 146,500 officially registered unemployed on the roster.20Based on those data, it can be estimated that only around 26 percent o f thejob seekers are registered by the SES, and around three-fourths o f unemployment i s hidden from authorities. Discrepancies between the two indicators are also impacted by the fact that, according to the Employment Act o f Armenia, the status o f unemployed i s not granted to individuals who had not been employed for at least one year. Understandably, this provision refers, first o f all, to young people, and together with other factors results in significant differences inage compositions o f registered and self-reported unemployment. Figure 2.13: Structure of Registered and Self-reported (ILO definition) Unemployment, By Selected Age Groups, 2003 100% 80% 60% 40% 20% 0% Self-reported Officially Self-reported Officially Self-reported Officially (ILO definition) Registered (ILOdefinition) Registered (ILO definition) Registered Total Female I\rale IAged16 30 - ti3 Aged 30 and above Source: National Statistical Service. 2oAccording to the 2002 Law on Employment Promotion, only the able-bodied population at working age can beregistered bythe N E S as officially unemployed. 37 The duration of unemployment remains extremely hi h, with nearly two-thirds of the unemployed looking for a job for one year or longer?'ln 2006, the average uncompleted unemployment spell was 14.6 months both for men and for women. (NSS 2006). The long average duration o f unemployment for most workers may indicate a stagnant unemployment pool in Armenia. Long-term unemployment i s a serious problem, since many o f those who wait several years for ajob eventually become inactive, which represents a waste o f human capital and an additionalburden on social funds and on the family. A marked reduction in long-term unemployment must be a priority for Armenia. The extent to which long-term unemployment i s a reflection o f low aggregate demand, a mismatchbetween the skills o f the unemployed and the skills demanded by enterprises, or a consequence o f incentives and social benefits that hinder job search, are matters for closer scrutiny. In general, the longer a person o f working age and in the labor force stays out o f active employment, the likelier that the person's skills will become obsolete. Figure 2.14: Unemployed by the Duration of Job Seeking, 2004 LFS,inPercent Less than I 1-3months 4-6 months 7-12 months More than 12 month months Source: NSS2005b. 2.4. Inactive Population Inanalyzing labor force dynamics, we cannot ignore the labor potential of the currently inactive population. Although a vast majority o f the economically inactive population aged 15 and over, or around 75 percent according to the 2004 LFS, do not want to work, cannot work, or are unable to work (students, retirees, disabled, household members taking care o f children or other dependants), around one-quarter o f the inactive population at that age can be considered discouraged workers who want to work and would be ready to start working, 21 The major disadvantage o f the LFS data on unemployment lies in the fact that the data do not represent a panel, so it is not possible to monitor the same individuals over time and investigate changes in their labor market status. 38 butfor various reasonsthey arenot actively lookingfor ajob. Inmost cases such respondents report that they have lost any hope o f finding work, or do not know where or how to look for work. While the general labor market situation might improve, a large portion o f the discouraged workers might enter the labor market, thus putting additional pressure on the workforce and wages. The reasons for inactivity vary according to gender and age. According to the 2004 LFS, four major groups o f almost equal size (aged 15 and over)-housekeepers, pensioners, discouraged workers, and students-formed the group o f inactive people, and the category o f discouraged people equaled 24.4 percent o fthe inactivepopulation (or 7.4 percent o fthe total survey sample aged 15 and over). According to the 2001 population census, the number of discouraged workers aged 15 and older (those with no hope o f finding ajob) equaled 63,300, and constituted 7.5 percent o f the inactive population, or 2.6 percent o f the total population at that age. The number o f discouraged males and females was approximately equal, but in relative terms, there were more discouraged males (10.3 percent o f all inactive males compared to 6.0 percent o f discouraged females out o f all inactive males and females, respectively). Although currently not in the labor force, they might potentially start actively searching for ajob when the economic situation improves, andjoin the rank o f employed or unemployed. Table 2.4: Economically Inactive Population, 2001 Population Census Data; ,000 Total Reasonfor Economic Inactivity Home- NoHope N o Student Age, Other Not maker of Finding Need to Health Reason Stated Job Work Total 841.7 99.4 63.3 16.5 164.6 407.0 37.1 53.8 Males 279.4 2.6 30.8 5.4 70.5 144.7 16.7 26.7 Females 544.3 96.8 32.5 11.1 94.1 262.3 20.4 27.1 Urban 611.2 77.8 45.5 12.2 123.1 283.2 29.5 39.8 Rural 230.5 21.6 17.8 4.3 41.5 123.8 7.6 14.0 Source: NSS (2003). 2.5. Older andDisabledWorkers F o r the economy as a whole, the increase in the employment rate of older workers i s crucial for sustaining economic growth, tax revenues, and social protection systems, including adequate pensions. This reflects the fact that the structure o f society i s changing radically: there are more older workers (aged 55-64), elderly people (aged 65-79), and very elderly people (aged 80 and over), and fewer children, young people, and working-age adults. According to the population forecast for Armenia, under one scenario, in which the fertility and mortality rates do not change, the share o f able-bodied population aged 15-64 will drop from 69.0 percent in2005 to 66.5 percent in2030, and the share ofpopulationaged 65 and older will increase from 11.1percent to 18.4 percent22(Annex Table 9). The aging o f 22Under another scenario, inwhch fertility and mortality rates improve, in2030 the working-age population is expectedto equal 66.7 percent, and the population over age 65, 18.1percent. 39 the population will have a major impact on public finances, especially for health services, long-term care, and social transfers in general. At the same time, a continuous drop in the number of school-age youth, aged 7-17, would require further adjustments inthe education system. The share of older workers in the total working population i s also increasing due to a move away from the culture of early retirement so as to encourage older workers to remain in the labor force. According to the 2002 Law on State Pensions, the retirement age for females will gradually increase from 60 in 2005 to 63 in 2011 (the retirement age for males has already reached this benchmark). Also, the retirement age for service under especially harmful working conditions will be increased from 55 in 2005 to 58 in 2011 for males, and from 49 to 55 (and to 59 for less-harmful working conditions) for females. International practice confirms that the two main arguments used to justify the use o f early retirement are invalid: (a) there i s no proof that older workers cannot participate fully in the employment growth o f expanding sectors, and (b) there i s no proof that young workers and older workers are interchangeable (EC 2005a). Figure 2.15: Population Projection for Age Groups 7-17,6579, and 80 and Over, InArmenia 800.0 700.0 600.0 500.0 -ePopulationaged7-17,Iscenario Populationaged 7-17, I1scenario 400.0 -A- Populationaged 65-79,l scenaric 300.0 +Population aged80+ 200.0 100.0 Source: Authors' calculations. Employment rates of older workers in Armenia (aged 55-64) are relatively high. Infact, according to the ILCS 2004 data, the employment rate at that age, 53 percent, exceeds not only the rate for most other transition countries, but also the employment-to-population rate for older workers in the EU-15 o f 42 percent (Table 2.2). The employment rate o f people aged 55-64 in Armenia even exceeds the target established by the EU at Stockholm o f employment o f 50 percent o f older workers by 2010. Also, the relatively high employment rates o f population aged 65 and older o f 29 percent, according to ILCS data, are a peculiarity o f the Armenian labor market, but these highrates largely reflect high employment o f older workers and the elderly insubsistence agriculture. One of the main "incentives" for older age groups to continue working i s very low pensions received from the state social security system. Compared to other 40 Commonwealth o f Independent States (CIS) countries, pensions are low inboth absolute and relative terms. InArmenia, the ratio o f average pension to average wage i s only around 20 percent, while in Belarus, the ratio exceeds 40 percent, and in Ukraine, it i s more than one- thirdofthe averagewage (Annex Table 12). Disabled persons belong to groups whose job opportunities on the open labor market are considerably restricted. The labor market situation o f these people i s determined by two main factors: their relatively low level o f education and vocational preparation, and their low level o f economic activity. During the socialist period, the system o f protected employment (the so-called protected work establishments) for the disabled was an alternative to the employment o f these people on the open labor market. However, this system encouraged the undesirable isolation o f the disabled from mainstream vocational and social life. Currently, disabled individuals have to compete for jobs on a par with other employees andjob seekers. While, according to 2004 LFS data, the employment rate o f the population aged 15 and older was 45.0 percent for menand 28.1 percent for women, the employment rate for disabled men was 15.6 percent, and for disabled women, 7.7 percent, for example, employment rates among the disabled are threefold and fourfold lower than for the general population. (NSS 2005b). Figure 2.16: Employment Rates of Older Workers, 2004 LFS and 2004 ILCS Data 70 60 50 40 30 20 I o 55-64, LFS 55-64, ILCS 65+, LFS 65+, ILCS I Source: NSS 2005b; World Bank andNSS 2006. 2.6. ChildLabor Compared to other countries at the same economic level, child labor i s not very common in Armenia; still, the number of working children i s significant. According to the 2004 LFS, around 5 percent o f children aged 7-1 7 (more than half o f them aged 15-1 7) were employed inthe last three months preceding the interview (7.6 percent o f boys and 2.3 41 percent o f girls). This corresponds to around 30,000 working children. The main reason for working for half o f the children, according to parents' answers, was acute financial shortage o f the family, while in 38 percent o f the cases it was the child's own initiative and willingness. However, the majority o f children were engaged in housekeeping work: 72 percent o f children reported that they are involved in cleaning, cooking, washing, ironing, and food purchases (housekeeping responsibilities were shared almost equally among all the age groups between ages 7-17); 3.6 percent took care o f younger siblings or ill family members, but 20 percent o f children aged 7-17 worked on family plots and did other household work. Only one-quarter o f children reported that they were not engaged in housekeeping work. (NSS 2005b). It is worrisome that quite a significant share of working children-more than one- fifth-work more than 35 hours per week. One-third o f the working children thought that they had performed hard manual work. More than halfo f the children were paid for the work performed, but while most boys (56 percent) earned money for their work, most girls (89 percent) worked without remuneration. Figure2.17: HoursWorked by Employed Childrenina Week, 2004 LFS 30 25 20 15 10 5 0 Up to 10 10-14 15-24 25-34 35-41 42-28 49-55 56 and hours hours hours hours hours hours hours more hours Source: NSS (2005b). 2.7. Gender Segregation Especially since the early 1990s, there have been significant gender differences in the participation of males and females in the Armenian labor force, which reflects the duality of the labor market. According to the ILCS 2004 data, the employment rate o f females aged 15-64 i s 45.0 percent compared to 67.9 percent for males; for prime age adults (aged 25-54) the rates are 53.1 percent for females and 76.2 percent for males. Coping strategies and social protection systems based on the model o f the male breadwinner, given the rapid drop in fertility rates and gender differences in longevity, may result inmany older women receiving drastically insufficient pensions. Today women account for almost 60 42 percent o f those over age 65, but the number o f females at that age will almost double from 185,000 in2001 to 342,000 by 2030. Both horizontal and vertical segregation are characteristics of female employment. Horizontal segregation i s indicatedby the highconcentration o f female labor insome sectors and professions that pay worse than the average. More than two thirds o f the employees in health care, social security, and education are women, which i s partly explained by the low wages in these sectors, and partly by working conditions, which are more favorable to reconciling family and work. (Annex Table 1). Vertical segregation i s indicated by the fact that in the private and public sector, only about one-fifth o f the senior government officials and managers fillings the top positions o f the employment hierarchy are women, and women represent only 16 percent o f employers, but women fill 78 percent o f clerk positions and 60 percent o f technician and associateprofessional positions (Table 2.5). Table 2.5: Employment, by Occupation and Gender, inPercent, 2004 LFS Total Including Total Including Males Females Males Females Legislators, Senior Government 100 78.6 21.4 5.8 7.8 3.O Officials, and Managers Professionals 100 38.2 61.8 14.0 9.1 21.0 Technicians and Associate 100 40.1 59.9 9.6 6.6 13.8 Professionals Clerks 100 21.9 78.1 2.7 1.o 5.0 Service Workers and Sales Workers 100 56.3 43.7 12.5 12.0 13.1 Skilled Agricultural and Fishery 100 52.3 47.7 27.3 24.5 31.4 Workers Craft and Related Trade Workers 100 86.4 13.6 12.9 19.0 4.2 Plant andMachine Operators and 100 96.7 3.3 5.6 9.2 0.4 Assemblers Workers without Qualification 100 65.9 34.1 9.7 10.9 8.0 Total 100 58.5 41.5 100 100 100 Source: NSS (2005e). For women, the greater responsibility for housework represents a barrier to entering and participating in the labor market. While out o f work, females look for a job for a longer period o f time: more than 68 percent searched for a job for 12 months and more compared to 58 percent among unemployed males. Women take on more responsibility for work relating to the home and family. Men have greater freedom to allocate time to work when necessary. According to the Armenian Time Use Pilot Survey conducted in July 2004, 65 percent o f men and 34 percent o f women are gainfully employed on an average day (Annex Table 13). This means that the average working day for men was almost 8 hours and 12 minutes, but only 5 hours and 4 minutes for women.23 Housework, or unpaid work, also differs considerably between the sexes, with almost 6 hours for women and 1 hour and 45 minutes for men. 23According to the August 2004 LFS, the average daily working hours for men were 6 hours and 18 minutes, and for women, 4 hours and 48 minutes (NSS 2005b). 43 2.8. Urban-rural Dimension and RegionalDisparities Inanalyzing labor markets and designingeffective labor market policies, itis important to note that internal labor markets in Armenia can hardly be considered homogeneous. Rather, they are heavily localized and have individual and specific characteristics. The varying speed o f reforms across the regions, unequal investment activities, geographical patterns o f demand for goods and services, and other factors have led to regional disparities inemployment andunemployment. Withinthe regionalmarkets, more specificterritories can often be identified in terms o f urban-rural labor markets, or the characteristics o f the workforce. Demand for labor was also unequal across regions; labor force participation varies significantly by region, includingby urban andrural area. One of the dimensions of a dual labor market in Armenia i s that labor force participation and employment rates in rural areas are much more favorable than in cities. Employment rates of rural population were almost 30 percentage points higher than in urban areas, 67.2 percent and 38.4 percent, respectively, for population aged 16 and older (Annex Table 8). However, these data may not reflect the real dimensions o f the disparities. Although formally rural areas have better employment and unemployment rates, the jobs are mostly low paid and seasonal (see Chapter IV). For the entire period starting in 1996, the urban unemployment rate has been significantly higher than the rural unemployment rate. The urban unemployment rate in 1996 was three times higher than the rural unemployment rate. The largest gap between the two indicators was recorded during 1998-99, when, according to the results o f household surveys, the urban unemployment rate was around 8 times higher than the corresponding indicator for rural areas. Significant differences between urban and rural unemployment rates were recorded by the 2001 census. According to the census, urban unemployment was at 48.4 percent, while the rural unemployment rate was only 17.4 percent, or 2.8 times lower. According to ILCS 2004 data, the unemployment rate inrural areas was 6.7 percent, while in Yerevan and in other cities the rate was close to 30 percent (population aged 15 and older) (Figure 2.18; Annex Table 8). 44 Figure 2.18: Participation, Employment, and Unemployment Rates inUrban and Rural Areas, According to the 2004 ILCS, Population Aged 15 and Older 80 70 60 50 0Participationrate 40 nEmploymentrate 30 Unemploymentrate 20 10 0 I Total Yerevan Other urban Rural Source: World Bank andNSS (2006). Data on labor force participation, education levels, and wages show considerable regional disparities by marz (district). According to the 2001 population census data, the employment rate in Yerevan o f 34 percent for those aged 15-64 was much below the national average o f 45.6 percent, but also lower than inany other marz inthe country. At the same time, the level o f unemployment inYerevan was the highest, at 32.4 percent, or almost equal to the employment level o f the able-bodied population, despite the fact that the capital has the highest concentration o f professionals with higher or secondary specialized education. Regional differences in labor market outcomes by marz are also enonnous: 1.8- fold in employment rates, 2.2-fold in unemployment rates (according to 2001 population census data), and 1.7-fold inaverage wages (Table 2.6). There i s no compelling evidence that regional wages are sensitive to regional unemployment rates (the so-called "wage curve"), as inmature market economies. The highest average wage levels are inYerevan and Kotaik- the regions with the highest unemployment rates. Wage setting might still be less competitive, largely due to the still-dominant role played by the "old" sector, including public and privatized enterprises, and by public and civil services with their unified wage scales. Interregional labor mobility, including commuting, i s still restricted. This i s due to the deficiencies o f the transport infrastructure, especially at the level o f microregions, and the extremely hightransportation costs. Moreover, the shortage o f low-rent flats, the highduties on purchasing and selling flats, and the significant regional differences o f real estate prices limit the opportunities for domestic migration. Regional disparities are also caused by the fact that the supply o f vocational training i s not adjusted properly to the economy o f the regions. The retraining and vocational training providing more advanced skills and adjusted to the specific economic and social environment o f the various regions are often not availab1e. 45 2.9. Education Level and Labor M a r k e t Outcomes Growth inlabor productivity will also require a highly qualified workforce. With regard to the education status o f the population, Armenia enjoys impressive educational outcomes for a country with its income level. Armenia inherited relatively good education indicators compared to countries with similar gross domestic product (GDP) per capita. It also has high enrollment rates at all levels o f education regardless o f gender, poverty level, or geographic location. The high levels o f educational attainment can be confirmed by the 2001 census data, according to which, in the most active working-age groups o f 26-49, 23.4 percent o f the population had higher and incomplete higher education, and almost the same percentage (23.8 percent) had secondary specialized (professional) education (NSS 2003). While the number of graduateswith tertiary education has significantly increased, part of the workforce i s quickly losing its skills, or the existing skills are becoming obsolete. Also, the new labor market entrants have lower levels o f education. Continuing investment in education will be required to develop a skilled and well-educated labor force. Typically, real wages are higher for more-educated than less-educated workers, and highly educated individuals have a lower rate and duration o f unemployment than less-educated workers. The available official data suggest that an increasing number o f vocational and technical graduates are more likely to become unemployed in a changing labor market (see Chapter VI. 46 There appears to be a positive correlation between level of education and economic activity. The 2001 census data confirm that individuals with higher education have significantly higher employment levels-73 percent-compared with 59 percent for the population with secondary and primary professional education, and 63 percent for the population with secondary general education (population aged 15 and older). However, in OECD countries, the relationship between level o f education and employment is clearer; that is, more educated people have higher employment rates.24 Figure 2.19: Unemployment Rate by the Levelof Education, inPercent, 2004 LFS Generalbasic Secondarygeneral Vocational Secondary Tertiary Total specialized Source; NSS (2005b). Table 2.7: HaveYou Worked since Graduation? (Respondents aged 23-33), Percent Education Yes No Incomplete secondary 21.1 78.9 Secondary 31.4 68.6 Secondary vocational 43.5 56.5 Higher 39.8 60.2 Total 34.8 64.9 Source: UNDP 2002. Nevertheless, even university graduates face serious problems in Armenia's labor market. For example, unemployment i s 38 percent among lawyers, 46 percent among agricultural specialists, and 31 percent among doctors (UNDP 2002). Among respondents aged 23-33, only one-third reported having had any employment since graduation. One of the explanations i s that curriculums do not correspond to market needs. By the UNDP 24InOECD countries in 1999, on average, the employment/population ratios for people aged 25-64 were 64.0 percent for workers having less than upper-secondary education, 77.0 percent for those with upper-secondary education, and 85.1 percent for those with tertiary education. The labor force participation rates were 68.3 percent, 80.9 percent, and 87.6 percent, respectively (OECD 2001). 48 Education, Poverty, and Economic Activity Survey, respondents aged 22-24 believed that only about one-third (35 percent) o f the knowledge and skills obtained duringtheir education can be used in work (the economists reported a figure as low as 15 percent). Slightly older respondents (aged 25-33) valued their university education more and indicated that 43 to 48 percent o f the knowledge and skills obtained during education could be used at work. Data on computer skills and knowledge of foreign languages especially bear witness to the inadequacy o f youth training for entry into the labor market. 49 CHAPTER111: WAGE DYNAMICSAND UNIT LABOR COSTS The wage-setting process influences the level of labor market flexibility, social security, and economic stability. The deregulation of salary policies in Armenia in general had a positive influence on stimulating highly productive and quali5ed work, in differentiating the qualiJication level of jobs, in influencing investments in human capital, and in labor movements. Real wage levels are gradually catching up to the pre-transition level but the purchasing power of wages tends to be higher than the official records indicate. According to suwey data, the return j-om education has suflciently increased but wage inequality largely represents the emergence of marketfactors, because education explains only a small part of wage inequality. 3.1. Wage Dynamics In the 1990s, the dynamics of real wages in Armenia had two clear phases. In 1994, following the fall in output and labor demand, and reflecting an erosion of wage levels by rapid inflation, real wages dropped to 7 percent compared to 1990.25The decline in real wages in absolute terms (and relative to output declines) was sharper in Armenia than in every other Commonwealth o f Independent States (CIS) country (except Tajikistan and Azerbaijan; CIS STAT 2005). Inthe early 1 9 9 0 ~wages (and working hours) were the main ~ mode o f labor market adjustment, instead o f employment. In 1995, wage levels in Armenia started to rebound at a relatively highpace, but the measured real wages are still muchbelow the pre-transition level: in2004, average (real) wages equaled 35 percent o f the wage levels in 1990 and 68 percent of the level in 1991, reflecting productivity gains in the economy (Figure 3.1). In purchasing power, real wages have probably declined less than wage statistics suggest.26Given that real wages and labor productivity are increasing fiom a very low base, it will require considerable growth for Armenia to close its wage and labor productivity gap with fast-reforming transition countries. Labor and employment policies in productive sectors in the early 1990s may be characterized as a building up of hidden unemployment by dramatic cuts in wages and salaries, allowing part-time employment and mass administrative leaves, when people remained on the payroll list, but actually did not get paid for a long time.27 In public 25 Consumer price inflation inArmenia was 174 percent in 1991, 729 percent in 1992, 1,823 percent in 1993, and 4,962 percent in 1994.Prices stabilized to single-digit levels starting in 1998. 26 The purchasing power o f wages under central planning was overestimated because prices o f many consumer goods were set below the equilibrium level, as attested to by widespread shortages. Therefore, the fall in real wages during the initial stage of the transition, when prices were liberalized, was most probably overestimated (World Bank 2005b). 27 According to the National Statistical Service survey conducted in 1996 o f 1,500 economic entities, about 25.7 percent of the personnel of these enterprises were on administrative leave, and 20.6 percent were not paid during the leave. Inprivatized enterprises the personnel on administrative leave comprised 33.7 percent o f total 50 service sectors (health, education, and so forth), the policy was oriented toward the conservation o f employment via very low salaries and savings on maintenance costs. In 1994 employment in the social infrastructure sectors (education, culture, art, health, research and development) comprised 85.2 percent o f the 1990 level, compared with 73.3 percent for all nonagricultural sectors. For 1997, corresponding indicators amounted to 75.3 percent and 60.1 percent, and for 1999,71.7 percent and 54.9 percent, respectively. In 1990, wages inthe health sector comprised 64.3 percent o f the economy wide average wage, whereas in 1998 they comprised 49.7 percent. In education, corresponding indicators amounted to 68.8 percent and 42.3 percent, respectively, Figure 3.1: Dynamics of GrossDomesticProduct, Employment,RealWages, and Labor ProductivityinArmenia, 1990-2004 (Indexes, 1990 = 100) 180 160 140 120 100 80 60 40 20 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 0Laborproductivity - - - Gross DomesticProduct- Employment Wages Source: National Statistical Service and authors' calculations. Large-scale privatization resulted in massive job cuts, especially in industry and construction, and in "vacant jobs," that is, in jobs that were not filled. This policy also resulted in the marked deterioration o f labor productivity and unit labor cost indicators in industry and other nonagricultural sectors of the economy. It was made possible by an absence at that time o f effective contract enforcement mechanisms and bankruptcy procedures, facilitated by widespread barter trade, and a build-up o f tax and mutual nonpayment arrears, mostly to the energy sector. As a result, the marked quasi-fiscal deficits were generated via indirect subsidizing of mainly state-owned medium- and large-sized enterprises.28 personnel. In medium-sized and large enterprises the personnel on administrative leave comprised about 40 percent o f total personnel. 28However, Armenia was more successful in managing quasi-fiscal deficits in the energy and utility sectors than most of the CIS countries. The volume o f the hidden deficit inArmenia declined from 8.2 percent o f GDP in 1995 and 4.9 percent in 1998 to 1.1 percent of GDP in2001, Meanwhile, total subsidies (including hidden 51 Changes in the structure of the economy, and thus changes in relative demand for different types of labor, have given rise to various wage premia to worker and firm characteristics (such as education and skills, occupation or industry) that were not widespread during the socialist period. Also, deregulation o f wages and wage policies in the transition countries was a part o f the overall liberalization process. This has led to the better aligned with productivity differential^.^' Wage inequality inArmenia, measuredby the inequality in dispersion of earnings, but also to better market efficiency, as wages become Gini coefficient, is very high but declined substantially between 1998/99 and 2004-fiom 0.438 to 0.37 (Table 3.1). (See World Bank and NSS 2006 for the details.) The wage gap between top decile workers and bottom decile workers (P90/P10) also decreased over 1998/99-2004. Wage inequality measuredby this ratio i s more comparable to CEE countries than to CIS countries, where inequality was highest (World Bank 2005). At the same time, 23 percent o f all employees in Armenia earned less than two-thirds o f the median (low pay), which means that the incidence o f low pay i s quite high. Inthe European Union, the decile ratio varies between 3 and 4. Table 3.1: Armenia: Summary of EarningsDistribution, 1998/99and2004 Monthly wages, 1998199 Monthly wages, 2004 Hourly wages, 2004 All Public Private All Public Private All Public Private PlOP50 0.50 0.48 0.35 0.43 0.44 0.38 0.43 0.42 0.43 P9OP50 3.00 2.50 3.00 2.33 2.22 2.13 2.33 2.26 2.23 P9OP10 6.01 5.21 8.57 5.38 5.00 5.67 5.44 5.36 5.20 Gini coefficient 0.438 0.405 0.498 0.370 0.356 0.357 0.382 0.360 0.405 Standarderror (1.2) (1.0) (3.4) (0.7) (0.6) (1.2) (1.1) (0.6) (2.3) Incidence o f low and highpay L o w pay, % 26.7 28.9 9.8 23.3 29.7 12.8 28.1 31.0 22.4 Highpay, % 28.7 25.4 54.5 29.8 22.3 41.2 26.1 24.3 29.2 Source: ILCS 1998/99 and 2004. See World Bank and N S S (2006). Notes: PlOP50 (P90/P50) denotes the ratio o f earnings of the bottom (top) decile relative to the median. Decile ratio is the ratio o f the top decile to the bottom decile, i.e. P9OP10. L o w pay is defined as earnings below two-thirds o f the median. Highpay is defined as earnings over 1.5 times the median. The incidence o f low (high) paid workers i s a percentage of low (hgh) paid workers inall wage and salary workers. As noted, regional disparities in average wages are also enormous. Wage levels in Yerevan and in Kotaik marz-the regions with the highest unemployment rates-are on average 1.7-fold higher than inthe lowest-paidregions inArmenia (Table 2.6). Wage gaps in regions are mainly explained by differences in human capital and job characteristics, while the variation in the wage due to the variation o f the unemployment rate i s relatively weak. Typical o f other countries in transition, average wages in the highest-paid financial intermediation sector much exceed the average, and are almost eightfold higher than the averagewages inpredominantly budget-financed culture and arts (Annex Tablel). According and quasi-fiscal subsidies) also declined from 9.4 percent o f GDP in 1997 to 4.2 percent in 2001 (see World Bank 2003a). 29 During the socialist period, wages were set according to a centrally determined wage grid, or the unified Tariff Schedule of Wages and Salaries. Wage differentials were very small and earnings structure was compressed. 52 to the 2004 LFS, average monthly wages o f employees in the state sector of the economy were 29,219 drams, while inthe non-state sector they were 40,646 drams (NSS 2005b). Despite the higher education level of women, the gender pay gap in Armenia, controlling for other individual characteristics, i s significant in both the public and private sectors and indicates that females are paid less than their male counterparts. This gap appears lower inthe public sector than inthe private sector. Women inthe private sector earned on average 32 percent less than menwith similar characteristics, while women inthe public sector earned 18 percent less than menwith similar characteristics. The gender pay gap inArmenia i s comparable with other CIS countries (see World Bank andNSS 2006). There i s a correlation betweenjob segregation and wage gaps: the concentration o f female employees i s higher in the public sector, which traditionally offers lower earnings. Pay differentials between males and females are in general significant, but they also depend on other factors, such as occupation, position, and differences in working hours. At the same time, due to the lower salaries, the public sector is less attractive to men, and this further increases the concentration o f women. Figure3.2: Average Wages of Employeeswith Different Levels of Education, According to 2004 LFS Data, inDrams 70000 60000 50000 40000 30000 20000 10000 0 k s t - Tertiary and Secondary Vocational General General graduate incomplete specialized secondary basic and tertiary primary Source: NSS 2005b. To close the gap, it is necessary that both public and private sector employers acknowledge the principle of equal pay for equal work, regardless of gender. Inthe past few years, salaries increased significantly in the public sector. This considerably improved the wage situation o f several traditionally female occupations, and reduced the gender wage gaps in the whole economy. Due to the increased earnings, the public sector has become an attractive option for a larger number o f men, thus also reducingjob segregation. An increase ineducationalwage premia has been an important factor behindthe rise in wage inequality. In relative terms, wages and salaries o f well-educated and highly skilled workers have increased, while wages o f less-educated workers have decreased (Figure 3.2). These developments have led to a substantial increase in returns to education, especially 53 university education. By the ILCS 2004, while the private sector offers a premium to special secondary education and tertiary education, the public sector offers a premium to tertiary education only. In the private sector, employees with tertiary education earned 63 percent more than those with general secondary education or below, keeping all other characteristics constant. In contrast, in the public sector this premium was 48 percent. (World Bank and N S S 2006). The wage gap would be even higher ifthe budget-financed social services sector, which traditionally employs the most-educated part o f the labor force, had not been the lowest-paid field o f economic activities, at least according to the official (measured) wage statistics. Table 3.2: Average Annual Earnings and Returnper Unit of Cost of Education, 16-33 Year- olds with Jobs, by Gender and Level of Education, In ,000 Drams; 2001 1.Average 2. Impact on 3. Annual 4. Total Cost 5. Return per Annual Earnings of cost to of this Level Unitof Cost Earnings thisLevel Student of (2 +4) Male Incomplete 240.4 47.4 Secondary 312.5 +72.1 141.0 300.3 24.0% Secondary 387.1 +146.8 178.4 751.O 19.5% Secondary 669.7 +357.2 582.7 2924.9 12.2% Vocational Higher Female Incomplete --- 47.4 Secondary 144.8 141.0 300.3 Secondary 160.9 --- 178.4 751.0 --- Secondary 461.0 +3 16.2 582.7 2924.9 10.8% Vocational Higher Both Sexes Incomplete --- 47.4 Secondary 253.0 141.O 300.3 Secondary 251.4 --- 178.4 751.0 --- Secondary 542.6 +289.7 582.7 2924.9 9.9% Vocational Higher --- Number o f replies from women too small for statistical significance. = Note: The returnper unit o f cost in column 5 is based on the "shortcut" formula for rate o f return when data on earnings are available for only a single year. Impact on earnings and total cost o f both secondary and secondary vocational are measured in comparison with incomplete secondary, and o f higher education compared with secondarv. Source: UNDP2000. The impact on labor-market outcomes can be analyzed within the framework of rate- of-return analysis, which focuses on the impact of education on earnings in relation to cost. Based on the results o f the Education, Poverty and Economic Activity Survey, the returns to education in relation to its cost are quite high for young males who get jobs (though they diminish as the level increases) (UNDP 2002; Table 3.2). Young women o f all 54 educational levels earn substantially less than men, but also gain a reasonable rate o f return on their higher education. For internationalcomparisons, a preferred approach might be to follow wage dynamics and levels in U.S. dollar terms (that is, industrial wages valued in foreign currency by the contemporaneous market exchange rate), using these as indicators of the results of the reform strategies and other factors. Prior to the transition, ruble wage levels inthe CIS states were rather equalized. After some 15 years o f reforms, differentials inmonthly wages (inU.S. dollars) are enormous: average wages differed in2004 from US$22 inTajikistan to US$237 in Russia and US$207 in Kazakhstan, in which the economic growth is based on extraction o f natural resources (mainly oil and gas). Armenia i s somewhat in the middle o f CIS states by the wage level inU.S. dollar terms, and average wages equaled US$79 in2004, a substantial increase compared to US$17 in 1995 (Annex Table 12; CIS STAT 2005). The influence of government on rate and dynamics of wages i s expressed in setting minimum labor cost, time schedule for payment of salaries to the public sector, and change of direct taxes and social benefits (through parliamentary procedures). The labor market impacts o fminimum wages depend heavily on the level at which they are set and how well they are enf~rced.~' Armenia, minimum wages are raised on an ad hoc basis, and In since 1997, they were increased only three times-in 1999, 2004 and 2006 (Table 3.3).31 Based on the 2004 LFS, only 10.6 percent o f the employees received salaries below the level o f minimum wages, and many o f them might be part-time workers. On the other hand, workers with earnings close to a minimum wage are likely to fall into poverty. Over the transition years, minimum wages in Armenia tend to be at a level below subsistence wages, and together with weak enforcement, the level i s too low to bebinding(that is, to affect wage and employment decisions). A t its current level, the minimum wage can hardly be characterized as an instrument of wage policy. Its role inalleviatingpoverty i s limited, and it seems to provide limited adverse effects on unskilled workers. Table 3.3: The Dynamicsof Average Monthly Wages and MinimumWages InArmenia, 1997-2006, Drams 1997 1998 1999 2000 2001 2002 2003 2004 2006* Average Monthly 13,600 18,000 20,200 22,700 24,500 27,300 34,800 43,400 60,600 Wages MinimumWages 1,000 1,000 5,000 5,000 5,000 5,000 5,000 13,000 15,000 Ratio o f Minimum 7.4 5.6 24.8 22.0 20.4 18.3 14.4 30.0 24.8 Wages to Average Wages, % * -January-July Source: NSS (2001); NSS (2005c), www.armstat.am 30 For a concise review o f the theory regarding the employment impacts of the minimum wage, see OECD (1998). 31Starting in 2004, according to the relevant provisions of the Poverty Reduction Strategy Paper, the minimum wage is defined in accordance with the general poverty line. In 2006, the minimum wage equals 15,000 drahms. 55 3.2. UnitLabor Costs The attractiveness of Armenia to foreign direct investment lies not in lower relative nominal wages per se, but inlower unit labor costs?2 Unit labor costs capture the change in the nominal wage in relation to the trend in labor productivity. A moderate rise in unit labor costs, implyingnominal wage growth approximately in line with labor productivity, i s essential to maintainingthe competitive cost advantage o f the economy. The systemic crisis of the first half of the 1990s,combinedwith the surplus employment, resulted in a sharp decline in unit labor cost. The return o f economic growth has somewhat improved this ratio. During2000-03, there has been a clear trend o f decline inunit labor cost inthe nonagricultural sector, which can be attributed to growing real exchange rate undervaluation(Figure 3.3; Annex Figure 9). This trend reversed in2004 with change of real exchange rate pattern and rapid appreciation that pushed US$- nominated wages up. In 2004 and 2005 nominal wages increased by 36 and 40 percent in USD terms, which fully offset unitlabor cost reductionof early 2000s. Figure 3.3: Dynamics of UnitLabor Cost inArmenia, 1990-2004 (1990 = 100) 100 80 60 40 20 0 1990 1994 1995 1996 -1 Unit labor cost, overall economy- 1997 1998 1999 2000 2001 2002 2003 2004 Unit labor cost, agriculture Unit labor cost, non-agriculture Source: National Statistical Serviceand authors' calculations. Figure 3.3 shows that the agricultural sector, although substantially more competitive than during the Soviet period, i s becoming less and less competitive compared with the nonagricultural sector, and the productivity gap between them i s gradually deepening. Currently, unit labor cost in agriculture i s the highest compared with other sectors. Labor productivity inagriculture in2004 was about 37 percent o f the labor productivity level inthe nonagricultural sector, whereas income from labor in agriculture comprised 55.5 percent o f 32 Calculated as the ratio o f annual wage (a proxy for labor cost paid bythe employer) to the GDP producedby one worker (labor productivity, or output per person employed). According to this indicator, sectors o f the economy in which unit labor cost is lower have higher levels o f labor productivity, and consequently higher levels o f competitiveness. 56 wages in the nonagricultural sector, that is, unit labor cost in agriculture in 2004 was about two times higher than in the nonagricultural sector. The obsolete machinery, equipment, and so forth, and fixed capital in the rural economy resulted in much more limited possibilities for investment generation from internal revenue than in the other sectors o f the economy.(World Bank 2004b). In addition, agriculture during 1999-2004 received a disproportionately low share o f public and private investments in fixed assets, including investmentsinrural infrastructure. Armenia's unit labor cost (ULC) generally compares favorably to neighboring markets but has been on the increase. This is mainly due to Armenia's comparative advantage of having an educated and skilled workforce and the country's strong tradition o f highly skilled craftsmen. In the mid-1990s real wage increases outperformed increases in labor productivity. This was partially due to wage adjustment from a very low base. Unit labor costs did not rise during the 1990s as the real exchange rate depreciated. The early 2000s were characterized by strong productivity gains, and with the exception o f 2003, productivity outperformed an increase in real wages. Unit labor cost started to increase in late 2002, reflecting both an increase inemployment compensation and real exchange rate appreciation. Inthe late 1990s, unitlabor cost increasedto about 46 percent ofU.S. level (about 44 percent o f the Euro area level), from as low as less than 3 percent in early 1990s. Manufacturingunit labor cost in Armenia i s 0.27, compared to the average o f 0.57 for 67 countries, and i s comparable to Malaysia (0.27), Ireland (0.26), Korea (0.26), and China (0.25) (World Bank 2003b). Inthis respect, Armenia's industryi s fairly competitive, also consideringthe fact that there are possibilities inArmenia for further reduction o funit labor cost. 57 CHAPTERIV: RURALEMPLOYMENT Thepolicies aimed at liberalizing agriculture under the conditions of insuflcient investment and deterioratingproductive infrastructure resulted in excess employment, low productivity, isolation of agriculture and the rural population from the rest of the economy, and widespread subsistence-level farming. Although employment rates of the rural population are high, total income from agriculture for most households is low, and underemployment determined by seasonality of agricultural activities is a serious problem. Theprevailing rural economics are a result of survival strategies accompanying economic reforms. In the long run, an agriculture that is poor in investment and capital can hardly be sustainable and competitive, even in the internal markets. An increase in labor productivity in agriculture, and relevant labor shedding are prerequisites for transforming from subsistencefarming to agricultural business. To maintain the high employment rates in villages, more nonagriculturaljobs in rural settlements should become available. This would lead to better integration of the rural population in the economic life of the country and to diversijkation of the sources of their income. Also, labor mobility of the ruralpopulation should increase. 4.1. Output andEmploymentinAgriculture In Armenia, during 1990-2005, the number of people employed in agriculture and forestry increased from 289,000 to 507,600. (CIS STAT 1994; NSS 200%; www.armstat.am). Like in many other countries in the region, transfer o f a substantial percentage o f the nonagricultural labor force to agricultural activities, following the collapse o f collective and state farming, land privatization, and formation o f the small rural farms, was for many households the main coping strategy for cushioning the blow o f the tran~ition.~~ This resulted in about a twofold increase of employment in agriculture, compared with the pre-reform period.34 Such trends have also been caused by loss o f employment in other activities in villages, and by out-migration from the cities in Armenia and from other Commonwealtho f Independent State countries to rural areas. 33Armenia has implemented one o f the most comprehensive landreform programs inthe Former Soviet Union states. Land reform was initiated in 1991, and by mid-1994, most o f the agricultural land had been privatized (see World Bank 1995a). 34A drop inagricultural employment and a corresponding increase inlabor productivity in2002 compared with 2001 (from 570,000 to 500,800 for employment, and from 57.9 percent to 68.5 percent for productivity, measured as an index in which 1990 = 100) is explained by revisions in employment estimates based on the 2001population census data. The 1990 data also include employment infishery. 58 Table 4.1: Number of Farms andthe Total Area Of their Privatized Agricultural Lands 1991 1992 1993 1994 1995 1997 1999 2000 2001 2002 2003 2004 NumberofFarms, 1,000 165. 238. 298. 312. 316. 321. 335. 332. 334. 334. 337. 338. Hectares (end-year data) 2 3 1 9 4 1 1 6 8 7 9 5 Agricultural LandAreaof 214. 310. 377. 396. 429. 433. 458. 458. 458. 453. 461. 468. Farms, 1,000 Hectares 9 4 1 9 2 8 9 6 6 1 3 6 Average AgriculturalLand 1.30 1.30 1.26 1.27 1.36 1.35 1.37 1.38 1.37 1.35 1.37 1.38 Area per Farm, Hectares AgriculturalValue Added, 23.5 28.2 48.7 43.5 40.7 29.4 27.0 23.2 25.5 23.4 21.5 22.6 % o f GDP Source: National Statistical Service and authors' calculations. Over the last 15 years the output dynamics in agriculture were less volatile, and the drop in agricultural production was the smallest compared to other sectors of the economy. The lowest output was reported in 1993 when the volume o f agricultural value added comprised about 84.3 percent o f the 1990 level. The pre-transitionlevel o f total output in agriculture was already achieved in 1998. As of 2004, the total agricultural output accounted for 142 percent o f the 1990 level. On the other hand, adjustments inlabor income from agriculture were more dramatic. As o f 1994, income fiom agricultural activities, inreal terms, comprised about 5 percent o f the 1990 level; by 2004, labor income still remained relatively low-at 31 percent o f income in 1990-although they have increased about six fold compared to 1994 (Figure 4.1). Figure 4.1: Dynamics of Value Added, Employment, Labor Income, andProductivityin Agriculture, 1990-2004 (1990 = 100) 1990 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 0Laborproductivity - Value addedinagriculture- Employment Labor incomes Source: National Statistical Service and authors' calculations. In 1994, labor productivity in agriculture was only 51 percent of the 1990 level, and in 2004 it was only 81 percent of the 1990 level. These developments, while helping to keep 59 agriculture afloat and making it the largest economic sector in the Armenian economy, in terms o f the number o f people employed, resulted in an increase o f labor intensity and a corresponding decrease o f capital intensity (in terms o f both fixed assets and working capital), but led to a sharp drop in productivity inthe early 1990s.Also, a significant part o f agricultural activities i s subsistence farming, resulting in a dramatic decline o f the levels o f commercialization and monetization o f agricultural production o f small farms, which are currently responsible for more than 95 percent o f agricultural output.35 Agricultural farms are taxed by a land tax and farmers are exempt from other taxes.36 However, within the framework o f Armenia's post-World Trade Organization accession obligations, the existing value-added tax exemptions for agricultural products should cease to exist as o f January 1, 2009. The government i s providing subsidies covering part o f the cost o f irrigation water supply. There i s also a small and diminishing amount o f grant-based provision o f seeds, fertilizer, and fuel to farmers at subsidized prices. (See Annex Table 15). Table4.2: Employment-to-PopulationRatioandEmployment Status by Stratum,Percent Strata Employment-to- Employment Status, YO Population Ratio Total, Employee Employer Self- Self- Collective Other (Proportion of Employed emp- emp- Enterprise population aged Aged 15+ loyed loyed Member 15+, that is in in employed), YO Family Non- Farm farm Busi- ness Stratum 1 77.4 100.0 20.9 0.5 75.4 3.0 0.2 0.0 Stratum 2 74.8 100.0 18.3 0.6 78.6 2.1 0.0 0.4 Stratum 3 79.3 100.0 18.8 0.2 79.2 1.6 0.1 0.1 Stratum 4 81.0 100.0 12.0 0.1 86.6 1.3 0.0 0.0 Stratum 5 85.6 100.0 13.5 0.2 85.3 0.5 0.0 0.5 Total 78.8 100.0 17.1 0.3 80.5 1.8 0.1 0.2 Source: RuralEmployment Survey 2005. The Rural Employment Survey 2005 confirms significantly higher rates of employment inruralareas comparedto the generalp~pulationThe~employment-to-populationratio . ~ ~ 35 According to official data, in 2002, the level o f commercialization of agricultural production was about 56 percent, whereas the level o f monetization (that is, excluding barter trade and incomes inkind) was about half o f the commercialization levels. According to the results of the Rural Employment Survey 2005, the level o f commercialization i s even lower. 36 In2003, the landtax comprised 0.9 percent oftotal tax revenues of the Armenian consolidated budget, or 0.6 ercent of agricultural value added. P7 The RuralEmployment Survey was conducted during June-July 2005 on a sample of2,000 ruralhouseholds, representative on the national level and by agro-economic zones. To ensure the heterogeneity o f the sampled, proportionate stratified sampling was done, and 5 strata o f population were designated according to the degree o f "vulnerability" of the rural population with regard to agricultural production and marketing based on the following characteristicsiindicators o f the country's rural population: (a) altitude above sea level of the population's residence; (b) population residence distance from Yerevan; (c) population residence distance from the capital city of the district; (d) size (number o fpopulation) o f the residential community; and (e) agro-zone o f 60 o f population aged 15 and older was 78.8 percent (Table 4.2 and 4.3). Gender differences in rural employment are somewhat less pronounced, but the employment rate for men i s 9.1 percent higher than for women (83.7 percent for men compared to 74.6 percent for women). Survey data on employment by sector indicate that 81.2 percent o f the employed in rural areas were working inthe agricultural sector, and the predominant form o f rural employment (80.5 percent o f the total) was agricultural self-employment on privately owned lands. Overall, 63.5 percent o f the population aged 15 and over at the time o f the survey were self- employed inthe agricultural sector. Only 1.4 percent o f rural residents aged 15 and over was involved innon-fann self-employment. Analysis of employment by age group also reveals some interesting patterns. Employment-to-populationratios are the smallest inage groups 15-24 and 65 and older (61.6 percent and 64.3 percent, respectively). Employment inage groups 15-24 and 65 and older is predominantly agricultural self-employment on private farms (90.8 percent and 95.6 percent, respectively). Notably, the employment rate o f older workers, including at retirement age, i s very high.Inthe prime age group of 35-54, employment is more differentiated. The share of the employed in this age group is the highest, at 93.8 percent, and agricultural self- employment accounts for around 66 percent o f the total employed. Hired employees had relatively larger representation (25 percent o f total employment at that age group), followed bynon-farmingself-employment (2.2 percent), and employers (0.4 percent) (Figure4.2). Figure 4.2: Employment by Age Group, Percent of Population I I I I I 15-24 25-34 35-44 45-54 55-64 65+ - -X. - including self-employed in f a d y farm Source: RuralEmployment Survey 2005. Due to the seasonal nature of agricultural production, secondary employment i s an important characteristic of rural employment, and 15.7 percent of respondents aged 15 and older also had secondary employment. These were mainly people whose main the residential community. The highest strata reflect the least-favorable farming conditions. (See Annex Table 16). 61 employment was hired employment or non-farming self-employment. The majority o f those with secondary employment (98.3 percent) consider agricultural self-employment inprivate farms as secondary. The remaining employment statuses have a very small representation. Only 0.8 percent of those with secondary employment were involved in non-farming self- employment, and only 0.6 percent were hired workers, indicating that jobs other than in agriculture are very limitedinrural areas. Rural employment i s largely seasonal, temporary, or occasional. Nearly half o f rural residents aged 15 and older are underemployed, due to involuntary factors such as the seasonal nature o f farming, while around 33 percent o f those currently employed would prefer to have full-time employment instead o f their current work arrangements. (Annex Table 17). The unemployment rate in rural areas is low. Given that the survey was conducted in- season (during June-July 2005), the unemployment rate o f the population aged 15 and older was only 0.9 percent. The low unemploymentrate arguably corresponds to the acknowledged (official) methodology for calculating rural employment, according to which working-age people owning usable agricultural land are considered employed. This approach, however, does not take into account the underemployment inherent in agricultural activities. Considering that the employed are mainly involved in the agricultural sector and their employment i s seasonal, the unemployment rate was further recalculated through the use o f the underemployment data, and the adjusted annual average unemployment rate inrural areas amounted to 14.1 percent.38The rate increases in parallel to the increase in vulnerability o f surveyed households constituting 17.1 to 17.3 percent for the most-vulnerable strata, compared to 9.5 percent for the least-vulnerable strata. Table 4.3: Economic Activity and Unemployment by Age Group, June-July 2005 Age Employ- Non-emp- Unem- Economi- Economi- Unemploy- Adjusted Groups ment-to- loyed, YOof ployed, YO cally Active cally ment Kate, Unemp- Population Total of Total Population Inactive Y O loyment Ratio, Population Population (employed Popula- Kate, YO YO +unemp- tion, YO loved). YO 15-24 61.6 38.5 1.o 62.6 37.4 1.7 14.9 25-34 85.4 14.6 1.9 87.4 12.6 2.2 15.4 35-44 93.6 6.4 0.5 94.1 5.9 0.5 13.7 45-54 93.4 6.6 0.1 93.5 6.5 0.1 13.3 55-64 90.2 9.8 0.2 90.5 9.5 0.2 13.4 65+ 63.9 36.1 0.1 64.0 36.0 0.1 13.3 Total 78.8 21.2 0.9 79.7 20.3 1.1 14.1 Source: Rural Employment Survey 2005. 38 In the 2001 population census, 100,800 rural unemployed were registered, or 16.9 percent of the economically active populationaged 15 and older (NSS 2003). 62 4.2. Landownership and Land Use In most rural households, 97.6 percent of the respondents had privately owned, leased, or other useable agriculturalland. The remaining2.4 percent o f households were landless. The average rural household has 1.45 hectares (ha) o f land, o f which 1.3 ha i s privately owned and 0.15 ha i s leased. About 97 percent o f households have privately owned land, while 10.5 percent o f households leased land. The majority o f households with leased lands (95 percent) also have privately owned lands. For the least-vulnerable strata (mainly Ararat valley), the average size o f households' lands i s about four to five times smaller than that o f the most-vulnerable strata. The average size o f agricultural land per household increases in parallel to the increase in vulnerability ranking (correlation coefficient o f 0.96; see Figure 4.3). Rural households cultivate on average 84 percent of their agricultural land, or 1.2 hectares. In the first three strata (least vulnerable), the average household cultivates on average 88 percent o f its agricultural land, compared to 80 percent in the last two strata. Consideringthat the size o f land per household i s significantly smaller inthe least-vulnerable strata, as a rule, small land plots are used more intensively. Households that do not entirely cultivate their land explain the underuse o f land mainly by shortage o f irrigation (21.2 percent), poor quality o f land (14.2 percent), financial constraints (13.9 ercent), and large distances between the agricultural plot and their residence (11.3 percent).3 8 Figure 4.3: Average Size of AgriculturalLand per Household, Hectares 4.0 3.5 3.0 2.5 2.0 1.5 1.o 0.5 0.0 Strata 1 Strata2 Strata3 Strata 4 Strata 5 0 Averagelandareaperhousehold 0 Privatly owend 17Leased Source: Rural Employment Survey 2005. A vast majority of households, 95.8 percent of the respondents, were involved in production of crops, and 64.9 percent in animal husbandry. The share o f households involved in animal husbandry i s larger among the most-vulnerable strata. Data on levels o f marketability reveal that the average household sells for cash or barters 53.1 percent o f its crop production and consumes slightly less than half (46.9 percent) in the household itself. 39 Inthe entire survey cluster, the shortage of irrigation water was mentioned as the causeof underuse of land by only five percent o f households. 63 But levels of marketability of crops vary widely. Watermelon, melon, vegetables, and grape have the highest level o f commerciality (on average 82 percent). Wheat, barley, and fodder crops produced are mainly used for own consumption inhouseholds and farms (average level o f commerciality i s 26.8 percent). Considering that the major part o f the cultivated land in the most-vulnerable strata i s allocated to those crops, it can be assumedthat the average level o f marketability o f crops produced in these strata should be very low. Around 80 percent o f farms involved in animal husbandry had produced some kind o f animal products within the 12 months preceding the survey, the most widespread o f which were milk and eggs, but the marketability o f those products i s very low. 4.3. Income from Agriculture Household total income from agriculture is, in general, low. The total monthly income o f the average household from crop productionamounted to around 22,000 drams, 46 percent o f which was monetary income (Annex Table 18). The total monthly income fi-om crop production per 1hectare of land declines inparallel to the increase invulnerability. Instrata 1 and 2, it amounts to an average o f 32,000 drams per hectare, compared to 13,000 drams in the two most-vulnerable strata. The average monthly income from livestock production per household (including own consumption in household/farm) was 15,300 drams (Annex Table 19). Table 4.4: The AverageMonthlyIncomefromAgriculturalActivity (Cropproduction,animal husbandry, processing)per Household,Drams Average Monthly Of Which: Average Monthly Incomefrom Incomefrom Incomefrom Estimate of Incomefrom Production Sold as Agricultural Production Own Agricultural Share of Total Activity per Sold Consum- Activity per EstimatedIncome, Household ption Hectare ofLand Y O (includingown consumptionin household) Stratum 1 30,897 15,905 14,992 50,65 1 51.5 Stratum 2 33,629 20,763 12,866 49,454 61.7 Stratum 3 40,7 14 13,622 27,092 31,562 33.5 Stratum4 46,000 15,156 30,844 29,299 32.9 Stratum 5 80,632 23,947 56,684 31,012 29.7 Total 44,435 16,583 27,853 36,423 37.3 Source: RuralEmployment Survey 2005. The average total monthly income per household (including consumption within the household) from crop production, animal husbandry, and processing of agricultural products amounts to around 45,000 drams per month. This i s comparable to the average wage inthe country o f 43,400 drams per month in2004. The major part o f the output, or 62.7 percent, i s directly consumed in the household and the farm to meet their own needs (see Table 4.4). The total income o f the least-vulnerable strata i s 2.6 times larger compared to the most-vulnerable strata. In the more-vulnerable strata, the major part o f agricultural products (on average around 70 percent) are used for own consumption, while more than half o f the 64 production inhouseholds o f less-vulnerable strata i s sold or bartered and generates monetary income. Almost half of households (52 percent) were involved in the processing of their agricultural products. The main products resulting from processing are dairy and bread. Bread i s produced equally by 15 to 21 percent of households in each stratum, while dairy production i s more common inthe most-vulnerable strata, Meat processing has an extremely small representation inall the strata, but the level of marketability ofmeat and dairy products i s quite high (73.8 percent and 62.5 percent, respectively), that is, they are produced mainly for sale purposes. The level o f marketability o f domestic wine and vodka i s 50.2 percent, but bread i s almost exclusively produced for own consumption, 65 CHAPTERV: YOUTH EMPLOYMENTAND UNEMPLOYMENT Surveys of unemployed youth and students of primary vocational schools and secondary specialized education establishments confirm that youth unemployment is a serious issue in Armenia. Especially worrisome is the long-term character of unemployment among the youth. The situation is exacerbated by the low level of professional education among the young job seekers, and their unwillingness or inability to invest in upgrading of their skills. The current information base does not allow monitoring of the training system, the labor market, and the link between them, to forecast skills requirements of the labor market and, therefore, training needs. Nevertheless, vocational education and training cannot always guarantee employment. Along with other employment policy measures, the provision of training could be matched with labor demand, including the cyclical and transitional characteristics of the labor market. The eficiency of vocational education and training is directly related to successfully targeting and individually tailoring the proposed training courses to the needs of certain regions, sectors of economy, occupations, and to the individual characteristics of the student or adult seeking retraining and upgrading of skills. VET should fit into the general adjustment of the local economy to the new demand. The country laclcs instruments to assist in the identiJication of present andfuture skills requirements by the local economy during a period of rapid economic change and subsequent uncertainty. This should be based not only on the limited number of currently vacant jobs registered by the employment services, but also on the competences (knowledge and skills) thatpeople need to have. There is a lack of a clear view on what training needs to be provided and what skills need to be developed to serve the needs of the local economy and maximize the chances of the graduates of vocational schools and the unemployed offinding employment. 5.1. The Level and Structure of UnemployedYouth Today, youth unemployment is a global problem. According to the United Nations, in 2003, people aged 16-24, representing one-fourth o f the world's working-age population, accounted for around 47 percent o f global unemployment. An unemployment rate among the age group 15-24 was included on the list o f Millennium Development Goal indicators. As stated in the International Labor Office (ILO) report, "Global Employment Trends for Youth 2004," people aged 15-24 (also referred to as young people in this chapter) are more vulnerable in the labor market. Moreover, international experience reveals that compared to adults, employed young people receive lower pay, and have higher rates o f inclusion in informal sectors and o f underemployment, and lower rates o f stable employment (ILO 2004b). Despite some improvements in Armenia's labor market, it is still characterized by an extremelyhighoverall level of unemploymentand a structuralmismatchbetweenlabor supply and demand. The high unemployment rate, in itself, makes finding productive 66 employment difficult for youth primarily upon first entering the labor market, and significantly increases the relative risk o f unemployment for young people. The problem i s more acute intransition economies with highpoverty and unemployment rates. According to 2001 population census data, in Armenia around 55 percent o f those aged 15-24 were economically active; however, only half were employed. According to the census data, unemployed youth accounted for 27 percent of the total unemployed in the country (NSS 2003). A substantial percentage of youth are in neither work nor school. A more appropriate indicator than the youth unemployment rate could be the youth non-employment rate (the idleness rate), defined on the basis o f a widened definition of the labor force, adding to both nominator and denominator young people who are in neither education nor employment:' According to the 2001 census, the net enrollment rate o f young people aged 15-24 in education in Armenia was 35.8 percent, while 28.7 percent o f the population aged 15-24 were employed. Of those aged 15-24, 35.5 percent were neither studying nor employed. The non-employment rate of youth (U)was 55.3 percent. In Armenia, the proportion of youth dropping out of the basic compulsory education system before obtaining qualifications, as indicated by data on education attainment levels, i s relatively low (Table 5.1). Although the number o f young people who leave school with only basic education i s not large, these individuals have little chance o f finding a job, and still less o f finding one o f reasonable quality. According to 2001 census data, around 13 percent of youth aged 21-25 had completed only general basic (grade S), general primary, or primary education. Training provided by labor offices for such people does not seem to address this problem, since those participating in the courses that exist obtain neither adequate training to actually be qualified, nor a certificate o f qualification, on completion. What i s needed i s to minimize the drop-out rate from schools and to provide more opportunities for those who want to reenter the education system. Recognition o f qualifications obtained outside the formal education system i s also important inthis respect. Table 5.1: School Attendance of De Jure Population, by Age and Type Of Educational Institution, 2001 Census Age Total Higher Secondary Secondary Other Not specialized general studying 15 100 2.0 88.9 0.5 8.6 16 100 9.5 9.0 53.5 0.8 27.2 17 100 21.8 13.3 5.2 1.3 58.4 18 100 22.4 11.0 0.9 0.8 64.9 19 100 24.4 4.8 0.3 0.5 70.0 20 100 21.7 2.7 0.2 0.3 75.1 21 100 13.8 1.4 0.1 1.2 83$4 22 100 7.8 0.7 0.1 1.2 90.2 23 100 4.9 0.4 0.0 1.5 93.1 24 100 3.3 0.3 0.1 1.4 94.9 Source: NSS (2003). 40The non-employment rate may be defined as: U= (young people not employed - young people ineducation) f (youth population - young people in education), or its counterpart: E = (employed young people) / (youth population - young people in education). U= 1- E (O'Higgins 2003). 67 Young people have a tangible participationinArmenia's labor market. According to the 2001 census, young people comprised around 18 percent o f the civil labor force. Such a large share o f young people inthe labor force i s a result of the highrepresentation o f youth inthe working-age population, which, at around 27 percent, i s significantly higher than the average for developed and most transition countries. On the other hand, labor force participation o f young people (ratio o f economically active young people to the total number o f youth) in Armenia i s significantly lower than the participation rate o f adults in their prime age (Table 5.2; Annex Figure 8). As noted, young people in Armenia's labor market, much like their peers in developed and transition countries, are more vulnerable to unemployment than adults (Table 5.3). According to the 2004 Integrated Living Conditions Surveys (ILCS), the unemployment rate among youth was 43.0 percent, a small decline compared to 45.9 percent in the 1998/99 ILCS (World Bank and NSS 2006). The ratio o f youth-to-adult unemployment inArmenia i s 2.5, the same as both developed and transition countries. The absolute number o f unemployed youth inArmenia i s extremelyhigh. Table 5.2: Youth Unemployment Rates and the Ratios of Youth-to-Adult Unemployment Rate Youth unemployment (aged Youth unemployment rate, Ratio o f youth-to-adult 15-24) %o flabor force unemployment rate thousand 1993 2003 1993 2003 1993 2003 Developed countries 10,441 8,609 15.4 13.4 2.3 2.3 Transition countries 4,399 5,051 14.9 18.6 2.9 2.4 Armenia ... 155" ... 43.0** ... 2.5** *200 1population census. **2004; ILCS estimates. Source: ILO (2004b); Armenia: National Statistical Service. 5.2. Sociodemographic Characteristics of Registered Unemployed Young People and their Households As part of the Armenia labor market study, two youth-related surveys were conducted: one among the youth registered at the State Employment Service (SES), and another among the students of vocational education establishments-both at primary vocational and secondary specialized schools. The surveys reveal additional aspects with regard to labor market participation o f youth, and give a representative picture of young (aged 16-24) beneficiaries o f the SES, and students o f vocational education and training (VET) regarding 68 their demographics, level of education, abilities, experience inthe labor market, motivations, and expectation^.^^ Figure5.1: Householdsof UnemployedYouth,byNumberof Working-ageAnd Employed Members,Percent 90 __ 0 -o.o------.--t----c1 None Three and more Working age members Employed members Source: Survey of UnemployedYouth 2005. The surveyed householdswith unemployed youth have a large working potential-on average, 3.5 working-age members per household. Nonetheless, the level o f inclusion o f this potential in the labor market is very low: 46 percent of the surveyed households have only one employed member, 23 percent o f households have at least two employed members, and only 4 percent o f households have three or more employed members (Figure 5.1). A strong reverse dependency between the number o f working-age members o f the household and the number o f employed members i s noted (correlation coefficient r = 0.75). According to the results o f the survey, parents o f registered unemployed young people mainly have secondary general (42 percent o fmothers, and 40 percent o f fathers) or secondary specialized education (33 percent o f mothers, and 27 percent o f fathers). About 15 to 17 percent o f parents o f unemployed youth have higher education (compared to the 22 to 23 percent o f the prime-age population with higher education). Low education attainment o f the above- mentioned households may partially explain the higher unemployment rate among the working-age members o f respondents' households. The majority o f respondents' households 41 The survey of unemployed youth included 600 registered unemployed aged 16-24. The majority-73 percent-were women. More than half of the registered unemployed young women, or 51 percent, were married, compared to the 64 percent o f men who had never been married. Also surveyed were 600 students from 30 primary vocational schools and secondary specialized education establishments in Yerevan in 8 marzes. 69 have a low living standard: 77 percent estimated their household's income at up to 50,000 drams, or around 10,000 drams (US$20) per person per month. The unemployed youth themselves have low levels of education and an absence of financial and other possibilities and intentions to upgrade their level of education. Only 43 percent o f the respondents have completed vocational education o f any level, including 5 percent primary vocational, and 25 percent secondary specialized education. Only 13 percent o f unemployed youth had higher education. There are 1.7 times more people with any level o f vocational education among unemployed women: 47 percent compared to 28 percent for men. The average number of years of study is higher for women-1 1.3-compared to 10.4 for men. Respondents without any level o f specialized education mentioned the following main reasons for not continuing their studies: failure to be accepted by vocational and higher education schools (32 percent), financial constraints and the need to work instead o f studying (29 percent), and lack o f interest incontinuing studies (27 percent). Figure 5.2: UnemployedYouth without Professional Qualification (specialty) By Intention to Continue Education inthe Future, Percent Intended to continue Intendedto continue, if it would Not intendedto continue be possible Source: Survey of UnemployedYouth 2005. Upgrading of skills and qualification levels is not a high priority for many unemployed youth. More than half o f the unemployed youth with no specialty (58 percent) do not intend to enhance their level o f education, and accordingly receive a specialization, and another 38 percent would enhance their level o f education if they had the financial means. Unemployed women are more prone to enhance their level o f education, but will attempt to do so only if the corresponding possibilities are in place (Figure 5.2). Other than studies in educational institutions, the majority o frespondents (57 percent) has not participatedinany other training courses, while 16 and 13 percent, respectively, have participated in such courses before taking up employment, and after losing their jobs. These courses were mainly financed by the currently unemployed themselves. 70 The most frequent professions among unemployed young people with a specialty are those related to medical services: nursing, obstetrics, and dental technician: around one- fourth o f respondents with a specialty had professional education in one o f these professions. The second most frequent specialization i s pedagogy (20 percent), followed, by a large margin, by accounting (8 percent) (Table 5.3). Although, the youth with medical and pedagogical specialties have the highest Unemployment rates, these specialties are still among the most popular areas of study in vocational schools and higher educational establishments. The choice o f specialization was mainly based on the interest o f young people in the given specialty (83 percent) rather than on labor market considerations, but in 12percent o f cases the lack o f alternative options or financial means also played a role inselection o f a specialty. This indicates that vocational school students and their parents do not have a good understanding o f the demand in the labor market. Providing young people with information on labor markets opportunities and payoffs to different levels and modalities o f schooling can allow them to make more educated guesses about their future returns, producing efficiency gains. There i s a lack o f cooperation and information exchange among the various labor market institutions and educational establishments. Table 5.3: Unemployed Youth with ProfessionalQualification, by Area of Specialization (percent) and Average Duration of Being Registered, At the SES (months) YOto Total Average Duration Number of of Being Registered Unemployed inthe SES, Months Youth with Professional Qualification Medical Services(nurses, midwifes, dental technicians) 24.3 19.6 Pedagogy 19.9 16.3 Accounting and Audit, Finance 7.7 17.9 Engineering 4.8 21.2 Economics 4.8 16.9 Law 4.8 22.1 Crafts 4.8 25.6 Computer Engineering and Automated Control Systems 4.4 17.6 A r t s (music, painting, design) 4.4 15.2 Philology 4.0 10.1 Other 4.0 16.6 Transport 3.7 16.1 Mechanics and Technology 3.7 15.8 Applied Sciences(geology, psychology, agriculture, etc.) 2.6 12.9 Building, Construction 1.1 13.7 Natural Sciences 1.1 10.0 Total UnemployedYouthwith Professional Qualification 100.0 19.8 (specialty) Source: Survey of UnemployedYouth 2005. According to the self-assessment, the majority of respondents (51 percent) think that their unemployment i s a result of their lack of relevant qualifications, education, skills, and knowledge. Another 31 percent are o f the opinion that their specialty does not have a 71 demand in the market. In spite o f that, only one-fifth o f them have tried to change their vocational profile through training or to enhance their qualification level. More than half of the unemployed mentioned financial constraints as the reason for not upgrading their skills level through training (Table 5.4). This indicates the need for the employment service to provide job counseling and inform the youth about the employment opportunities, and to organize training courses inaccordance with labor market demand. 5.3. Past Interactionsof Unemployed Youth with the Labor Market The unemployed young people registered at the employment service stay on the roster for a long time-an average of 19.7 months. Almost three-quarters of the registered unemployed lost their last job more than 12 months ago. For the unemployed with a specialty, integration into the labor market i s the most difficult for craftsmen, lawyers, engineers, and medical service providers, with a duration o f registered unemployment, on average, o f 25.6,22.1,21.2, and 19.6 months, respectively. Table 5.4: DistributionofReasonsfor Not ObtainingTrainingOr RetrainingbyUnemployed Youth, Percent Uselessness inTerms of FindingJob 20.4 Lack of Finances 56.3 DoNot Know Where to Obtain Training or Retraining 12.7 DoNot KnowWhat other Retrainingto Take 7.9 Other Reasons 2.5 N o Particular Reason (difficult to answer) 0.2 Total 100 Source: Survey of UnemployedYouth 2005. The survey of unemployed youth confirms that, as a rule, the higher the level of education of the respondent, the shorter the duration of being registered at the employment service. According to the results o f the survey, there i s a statistically significant reverse dependency between the number o f years o f study and the duration o f the registered unemployment (correlation coefficient r = -0.2). Those without a specialty are registered at the service for 21.5 months, on average, compared to 19.8 months for people with a specialty. Clearly, finding ajob i s more difficult for young people with no specialty. In addition, the vast majority of respondents (74 percent) had limited and short-term work experience and managed to have only one job before registering at the employment service as unemployed. More than half o f the respondents (57 percent) have up to two years of work experience, 40 percent have up to up to five years, and only 3 percent have five or more years of work experience. Prior to becoming unemployed, respondents received a salary of, on average, 15,000 drams to 17,000 drams per month, which i s almost three times lower than the average wages in the country. The absence o f 72 adequate work experience and low qualification levels are among the main factors not allowing young peopleto reenter the labor market. Unemployed young people lost their last job mainly due to jobs redundancies in enterprises and organizations (29 percent of cases) and personal circumstances (23 percent) (Figure 5.3). Considering that the majority o f respondents were married young women, we can assume that marriage was the personal circumstance causing loss o f job. Women mentionjob cuts and personal circumstances as the main causes o f losing their jobs, while males lost jobs mainly due to labor shedding, inadequate salaries, and contract termination. Figure 5.3: Unemployed Youth, by Reasons of Losing Last Job, Percent I Bankruptcy or liquidation of company Other structural changes 32.2 Terminated by ewloyer Non-satisfactoryw age Non-satisfactory working conditions S o w personalreasons &piration of fixed-termcontract Inconsistentqualificationand professionalskills = ' O i I I I I Other reasons 0 5 10 15 20 25 30 35 Total Male 0 Female Source: Survey of UnemployedYouth (2005). Informal channels, includingfriends, acquaintances, and relatives, are the main sources of information on vacant jobs (Table 5.5). Part of the unemployed youth are desperate to accept any job offer, and around one-third o f surveyed unemployed youth would take any job, another third are looking for jobs corresponding to their specialty, and the rest o f job seekers are seeking permanent employment on a contract basis. 73 Table 5.5: Unemployed Youth, by Source of Job Search, Percent of Total Relatives, Friends, Colleagues 89.2 Mass Media, Public Notice 49.1 State Employment Service 75.3 Private Employment Agencies 10.0 DirectlyProspective Employers 31.3 Internet 6.5 Other sources 0.2 Source: Survey ofUnemployedYouth 2005. L a c k of employment opportunities i s reflected in the fact that the SES has offered jobs to only 31 percent of respondent unemployed youth. In the majority o f cases where jobs were offered (63 percent), the unemployed have refused to take the job for one or another reason, thus preferring to stay in their status o f unemployed, rather than accepting low-paid jobs or jobs that do not correspond to their qualification. Employers rejected the candidacy o f the unemployed person suggested by the SES in 17 percent o f the cases, and the dominant cause was the absence o f skills and qualifications needed (in 45 percent o f such situations). Employers rejected those having a specialty 2.7 times more often than people with no specialty, and women twice as often as men. We can assume that qualification levels o f those rejected do not correspond to market demand, or their wage requests might be higher. Unemployed youth are passive in their relationship with the SES. More than half o f the respondents hadnot visited the employment service inthe three months precedingthe survey. The most important motivation for registering at the service i s to receive the status o f unemployed, which allows receipt o f unemployment benefits and monetary assistance. This would probably partially explain the rejection o f jobs offered by the employment service to the unemployed youth, most o f which are low-paidjobs. The participation o f respondents in various state employment programs has been mainly limited to the unemployment benefit scheme (for 81 percent o f registered unemployed youth). Only 9 percent o f the respondents have participated in training courses offered by state employment programs, and only one- quarter o f registered unemployed youth have received job search assistance. This demonstrates that services provided by the SES to registered unemployed people are inadequate. 74 Figure 5.4: Reasons for Failureto Take a Job OfferedBy the SES, Percent Refused by employer i I I I Refused by unemployed 69.5 54 2 I Vacancy was already I closed 9.5 14.5 Temporary job (already iI closed) 5.7 i3.6 Difficult to answer 5.7 2.4 I 80 60 40 20 0 20 40 60 80 % ~~ Source: Survey of UnemployedYouth (2005). 5.4. Students inPrimary Professional Schools and inSecondary Specialized Education Establishments Duringthe transition years, the problem of vocational education and training (VET) in Armenia i s not so much that the system of training has deteriorated, but that the world has changed. Inthe centrally planned system, vocational schools trained students according to contracts made with enterprises. As noted, at that time, a job was guaranteed to all graduates, and graduates had compulsory assignments for their first jobs. As a rule, young specialists had to serve at least three years at their first assignment. In a rapidly changing market economy, an individual's occupational future i s uncertain, and the risk o f a mismatch between the training required andthejobs i s very high. The schools are not adjusting to these drastic transformations inthe labor markets, and continue to offer the same specifications of openings year after year. As labor market data indicate, the skills currently offered by vocational and technical schools cannot lead graduates to jobs, and students are leaving school without contracts or jobs. Industries are unwilling or incapable o f hiring graduates o f vocational and technical education institutions. Even when finding a job, perhaps in the informal labor markets, the share o f graduates working injobs unrelated to their training i s growing rapidly. Upgrading of the labor force, or vertical mobility, i s a precondition for rapid structural and technological change in all the transition countries, for their competitiveness in the world market, and inraising the share of high-value-added products and services inthe 75 markets. Armenia i s no exception. Inmany enterprises, there i s a demand for workers who can perform a broad variety o f tasks. Many newly emerging occupations and sectors o f the economy require higher levels o f worker skills, many o f which cannot be acquired in the public vocational schools.42Many new businesses are severely limited in their development due to the lack o f trained specialists. Foreign investors often bring their own training programs into the country. Figure 5.5: Householdsof VET Students by Number of Working-age and Employed Members, Percent None Only one Two Three Morethan three m l o y e d members Working age members Source: Survey o f Students inPrimary Professional and Secondary Specialized Schools 2005. Based on the survey of students in primary professional schools and in secondary specialized education establishments, the students and their families share the characteristics of unemployed youth and their households. The majority o f students inthe surveyed vocational schools (78 percent) are female. Around 60 percent o f respondents' households have only one employed member (Figure 5.5). Most students themselves are/were not employed or are/were not involved inan income-generating activity duringtheir studies (88 percent). Nevertheless, most respondents (71 percent) classified their households as middle class, 2 percent classified their households as very poor, and 18 percent classified their households as poor. Respondents estimated the average monthly income o f their households as up to 50,000 drams (37 percent o fhouseholds), followed by families with total incomes o f 50,000 drams to 100,000 drams (28 percent o f households). Respondents' parents mainly have primary and secondary vocational (46 percent o f mothers and 48 percent o f fathers) or secondary general (36 percent o f mothers and 26 percent o f fathers) education; 18 to 19 percent o f parents o f VET students have higher education.43Compared to the general population, the parents o f students in vocational schools tend to more often be graduates o f VET schools themselves, which may explain why they encourage their children to continue their education invocational education establishments. ~ 42There are also 27 nonstate (private) secondary specialized schools in Armenia with 2,700 students, or 8.8 percent of the total enrollment insecondary specialized education establishments (CIS STAT 2005). 43According to 2001 population census data, among the generation o f parents aged 40-49, 23 percent had higher or incomplete higher education, 30 percent had secondary vocational or primary professional education, and 38 percent had secondary general education (NSS 2003). 76 The majority of students (60 percent) in vocational schools have chosen such educational institutions because they do not have the money to attend universities (which might be in other towns or regions), and 20 percent did not pass the entrance exams to universities, or thought they would not pass them due to the grades they received in general schools. So for most o f the youth invocational schools, their only option for continuing their education and getting a professionwas to attend vocational schools. The mismatch between the enrollment invocational schools and labor demand i s reflected in the fact that the three most fkequent specializations among the students o f primary and secondary vocational schools are those with the highest unemployment rates among the youth: medical services (20 percent of students), accounting (16 percent), and pedagogy (10 percent), followed by computer science (9 percent).44 Respondents' parents had the largest influence on the selection of specialization in VET for slightly more than half of the students (Table 5.6). The influence of the school and the SES has been decisive for only a very small number o f respondents. This indicates that there i s a need for closer cooperation between the educational and labor market institutions to assist the youth inmaking informeddecisions. Table 5.6: Circumstances with a Decisive Influence on the Selection Of Specialization by VET Students, Percent Parents 50.7% Friends 6.7% General School 1.1% Employment Service 1.3% Advertisement o f the Institution 2.7% Close Location o f the Institution 2.6% Decided Themselves 33.7% Other 1.2% Total 100.0% Source: Survey of Studentsin Primary Vocational And Secondary Specialized Schools 2005. The vast majority of vocational school students had very limited contact with the labor market. Only 28 percent o f respondent youth have made steps to find a job, and 12 percent have worked, mainly during summer vacations or in family business. After graduation most are left on their own, and manyjoin the rank o funemployed. 44 Structural mismatch between labor demand and supply is reflected in the national data on the structure o f employment, and the structure of students and graduates in secondary specialized education establishments (SSEE). In 2004, 22.3 percent o f graduates from SSEE majored in education, while their share in total employment was 11.Ipercent and shrinking; 25.5 percent of graduates specialized in health care, physical culture and sports, while their share in total employment was less than 5 percent. Only 2 percent o f SSEE graduates specialized in agriculture and 5 percent in professions in industry and construction, while their share in employment is 47 percent and 15 percent, respectively. The same trend applies to higher educational establishments as well (see CIS STAT 2005). 77 CHAPTERVI: LABOR MIGRATION Especially during the 1990s, labor migration had a major impact on labor supply in Armenia. It may involve a "brain drain" or the emigration of better-educated, higher-skilled workers to other countries, thus reducing the educational and professional potential of the country. As the migration survey confirms, however, this might be "compensated" by easing the tensions in local labor markets with low prospects for alternative employment. I n a situation of economic chaos and political uncertainty, contemporary intra- and interstate movements have created a heavy burden and considerable social stress for many families. Even among the same nationality groups, tensions have increased between the local population and newcomers. The labor market distortions have been compounded, the housing crisis has been intensiJied, and the impoverishment of a large segment of the population has become more intolerable. Inter-republic migration trends have substantially influenced the occupational and professional structure of the labor force. Perhaps the most important negative consequence of massive emigration is the direct demographic loss incurred by the Armenian population, which will have a long-term impact on the labor supply and on economic, social, and demographic developments in general. 6.1. Outflow of Population I t is a common pattern of labor migration in most industrial and developing countries that the labor moves to where the jobs are.45In the transition countries, the situation is very different. Despite highunemployment rates, the geographic mobility o f the labor force i s very low. People are territorially immobile and tied to their current place o f residence. One major reason for the low territorial mobility o f the workforce i s the established cultural and socio-psychological traditions and patterns o f living. People prefer to stay in the region where they have grown up, gone to school, and worked. Most households have roots in the community, such as a plot or a summer cottage built during the Soviet period, or a piece o f property that was owned by their parents. In addition, they highly value the close ties with relatives, friends, and coworkers living in the region. Many people do not consider the lack of local employment opportunities a strong enough motive to migrate. Such traditions and patterns cannot be ignored in the design o f employment, investment, regional, or other policies. Armenia seems to be an exception, and especially in the 1990s, migration flows had a major impact on the dynamics and structure of the population and labor force. Internal, and in the 199Os, external, labor migration has affected labor supply in the country and in certain regions intwo important ways. First,it arises as a response to lack o f employment (or employment acceptable to the migrating individual); and second, it relieves the pressure on the local labor market for those who do not migrate, thus helping to achieve equilibriumin 45There are three kinds o f migration: (a) emigration; (b) long-term temporary labor migration for longer periods as "quest workers" for a few years; and (c) short-term labor migration, including cross-border commuting and seasonal and casual work (IOM 1998).These migration patterns reflect different motivations for migration and different characteristics o fpotential migrants. 78 the local labor market (although those departing may be more likely to possess marketable skills and be more productive workers). Labor migration to Russia and other countries in search o f work, and remittances to support families at home, played an important role in mitigatingunemployment and providing household income. In the initial period of transition, such movements were produced by forced displacement of people. According to the official estimates, 530,000 ecological migrants (internally displaced individuals and refugees) appeared in the aftermath of the disastrous Spitak earthquake in 1988 (by some estimates, 200,000 o f them left Armenia, but in 1989- 1990, around 160,000 returned); 360,000 refugees came to Armenia from Azerbaijan following a four-year conflict betweenArmenia and Azerbaijan over Nagorno-Karabakh, and 72,000 internally displaced persons fled from border zones with Azerbaijan. Armenia gave refuge to another 11,000 o f its nationals escaping conflicts insome former USSRrepublics!6 Thus, before 1994-95, Armenia received around 420,000 immigrant refugees (an equivalent o f 12 percent o f the current p~pulation).~~ The scope of migration, including labor migration, can be assessed using several sources of information. The most reliable could be the population census data, but the 1989 census in Armenia took place under extraordinarily difficult conditions following the major earthquake in 1988, which resulted in an undercount o f population. The 1989 population census indicated the total (de facto) population o f Armenia was 3.288 million. The National Statistical Service (NSS) o f Armenia estimates the actual population in 1989, usingthe 1979 census data and migration and vital statistics, at 3.449 million. This i s 236,000 people more than in the 2001 population census. According to the available data, during 1989-2001, the natural increase o f population inArmenia equaled 367,000. Assuming that the net migration of population between the two censuses was zero, by the end o f 2001, the Armenian population should have been 3.816 million, but it was actually 3.213 million. According to these data, during the transition years the population declined around 600,000 due to the out- migration(negative net migration) o f population. Another source of information on the scope of out-migration can be the official data on entries and exits across Armenian national frontiers during 1992-2006. For 1992-99, only data on entries and exits through airports are available, but the government began to systematically collect data on all transport types starting in 2000. Balances for each year (entries minus exits) show that Armenia has always experienced a net outflow o fpeople, and that this outflow was very large duringthe early 1990s but fell sharply once the conflict with Azerbaijan ended and the economy was stabilized in 1995. The sum of these balances over 1992-2006 gives a very good approximation to the actual net outflow o f individuals from Armenia in that period. The outflow through airports alone exceeds 700,000 people (Annex Table 20). Ifthe proportion o f highway-to-air travel in the 1990s was similar to that during 46 Speech by G. Yeganyan, Head o f the State Department for Migration and Refugees o f the Republic o f Armenia, on the 45th Meeting o f the European Committee on Migration inStrasbourg, March27-28, 2002. 47According to the 1989 All-Union population census (with incomplete data for Armenia), 84,900 Azerbaijans lived in Armenia, and 390,500 Armenians lived in Azerbaijan, including 145,500 Armenians and 40,900 Azerbaijans inNagorno-Karabakhautonomous oblast. According to the 1999 population census in Azerbaijan, 120,700 Armenians were living inAzerbaijan. GOSKOMSTAT ofUSSR 1991c; SC o fAzerbaijan 2001). 79 2000-01, then total net outflow was equal to roughly 1million people. It i s notable that since 2004, the net flow o fmigrants registered at the border has been positive. Many people of non-Armenian nationalities also left the country. According to the 1989 population census, 6.7 percent o f those living in Armenia were o f non-Armenian nationalities. As a result of migration processes, by the beginning o f the 21st century, Armenia had become one o f the most homogeneous states of the Former Soviet Union (FSU), with 97.9 percent of the population being Armenians.48 Indirectly, the scope of migration can be assessed by the data o f the Savings Bank (Sberbank), indicating that by the end o f the mass privatization process (1995), about 663,000 Armenian citizens had failed to pick up their privatizationvouchers, to which every resident citizen was entitled. Table 6.1: The Number of Population of Armenian Nationality Livinginthe USSR and in Some of the CIS States; inMillions (According to the 1989 population census and the latest census data) Number % of The Latest Data by Census Year in 1989 Total Population Census Armenians 4.623 100 --- --- Including Living In: Armenia 3.084' 66.7 3.145 2001 RussianFederation 0.532 11.5 1.130 2002 Georgia 0.437 9.5 0.249 2002 Azerbaijan 0.391 8.4 0.121 1999 Other republics/CIS 0.179 3.9 0.1642 1995-2002 states 1. The All-Union 1989 population census in Armenia was conducted under complicated conditions (an earthquake in the North region of Armenia; events in Nagomy Karabakh) that influenced the accuracy of calculation. 2. Except Uzbekistan, in which in recent years a census has not been conducted, and Moldova, in which census data are not published yet. In 1989, around 51,000 Armenians lived in Uzbekistan and 3,000 in Moldova. Source; GOSKOMSTAT (1991~);CIS STAT (2005). 6.2. Current Labor Migration The size of population outside o f the current place o f living can be estimated by comparing the de facto and de jure population. In October 2001, the de facto population of Armenia was 3.003 million, and the dejure population, 3.213 million, meaning that during the census, more than 200,000 people were not in their place o f permanent residence. Some o f them might have been outside of their permanent place o f residence for family or personal reasons, but part were working outside o f Armenia, and another part were working in other regions within the country. 48The Armenian population also included 40,600 Yezeds, 3,400 Assyrians, 1,500 Kurds, and others who have also lived inthe country for generations. 80 Based on labor force survey data, the scope of external but temporary labor migration i s less significant. According to the 2004 labor force survey, 2.4 percent o f members of households, and 5.5 percent o f the employed that participatedinthe interview, comprised the people who left for work outside the country (for up to a six-month period) at the time of interview.49 Extrapolating the results o f the survey to the whole population, it can be estimated that slightly more than 60,000 Armenian citizens were working outside o f the country for more than six months at a time. The great majority-84.3 percent-ofthose who were absent from Armenia for work had left for the Russian Federation, and 59.6 percent o f the labor migrants worked in the field o f construction. These workers were predominantly male (89.5 percent) and married (66.9 percent). Blue collar workers dominate the current migration flows. Of total labor migrants, 58.4 percent had achieved secondary general education, 17.0 percent had achieved secondary specialized education, and 9.8 percent had achieved higher education. Among the general population aged 18 and older, the percentage o f the population with higher education i s two times higher (19.8 percent, according to the 2001 census data), and the share o f the population with secondary professional education i s also higher than among the migrants (19.7 percent). The specially designed Labor Migration Survey 2005 estimated the number of households that were involved in the labor migration process during 2002-05 at 95,000 to 122,000, or 12.2 to 15.6 percent of the total number of households (AST 2005). Inmost cases (78 percent), one member o f the family had left to work abroad; 15.4 percent o f the families had 2 labor migrants, and in 6.6 percent o f households, 3 or more family members were engaged in labor movements. This allows us to estimate the absolute number o f labor emigrants over the last three years at 116,000 people to 147,000 people, or 3.6 to 4.6 percent of Armenia's de jure population. Ths corresponds to 7.3 to 9.2 percent o f the economically active populationo f Armenia being involved inthe labor migration process. The highest rates of labor migration were recorded in Shirak and Lori marzes, where every third and fifth household, respectively, was involved in short-term labor migration. These regions had the highest unemployment rates in the country, besides Yerevan. The migration study data confirm the results o f the 2004 LFS-that the most popular country destination o f labor migrants was and still i s Russia: 87.6 percent o f labor migrants have worked there at least once duringthe last three years. Prior to departure, unemployment or inactivity rates among migrant labor are relatively high and their income low.'' This largely explains the motives for labor migration. According to the migration survey, approximately half (51 percent) o f the labor migrants were involved in some income activity before their first trip abroad, of which 65 percent had permanent jobs and 35 percent had occasional employment. The average monthly income o f migrants barely exceeded US$lOO per month at the time they were 49 The 2004 LFS was conducted inAugust 2004, and 2,539 individual households (comprising 9,453 household members) were interviewed. A 95 percent confidence interval with a 0.5 percent standard error for data on population employment was observed (see NSS 2005b). 50 The data on characteristics of migrant labor rely largely on the study, "Labor Migration from Armenia in 2002-2005" (AST 2005). 81 working in Armenia (80 percent o f cases), 16 percent of the migrants generated income between US$lOO and US$200, and only 4 percent were earning more than US$200 per month. Labor migrants mostly come from families with average income, rather than fi-om low- or high-income groups. Working conditions of migrant labor in the country of destination are insecure and harsh. Only one in eight workers had a written agreement (employment contract) with their employers in the host country. In most cases (72 percent), workers' relations with their employers were based solely on oral agreements. Moreover, in one o f three cases, the employers didnot adhere to the initial agreement or only partially fulfilled its terms. The vast majority o f the migrants were working full time (8 to 10 hours) or in 12-hour shifts (34 percent and 33 percent, respectively), and 8 percent o f employees worked on average more than 12 hours a day. Only 44 percent o f the labor migrants had regular days off, and one in four migrants had no days off. As for living conditions o f migrant workers, one out o f three Armenian labor migrants lived either at his or her actual workplaces (office, factory, and so forth) or inbarracks, and one infive laborers was living with relatives or friends who already resided inthe country o f the migrant's destination. Earnings of migrant workers were somewhat higher than in Armenia, but given the expenses of traveling, the short duration of the trips, and the costs of living (which in Russia are much higher than in Armenia), net gains are not significant. The majority o f the migrants (59 percent) earned US$250 to US$500 a month, and almost one in four workers earned less than US$250 per month. Only 17 percent o f the migrant workers had an average monthly income of more than US$500 per month. The mean monthly income o f the migrants inthe host country was US$410, with the minimum and maximum inthe sample o f respondents between US$100 and US$l,500. Although in absolute numbers the average income i s four times higher than what the migrants earned in Armenia, given the costs involved, the economic return to labor migration for a large number o f migrants i s not high. The only major incentive might be that in Armenia the labor migrants could not find a job: half o f the respondents mentioned absence o f work in general, and 43 percent noted that the wages offered by available work were too low for normal living. In 2007, the Russian Federal Migration Service is planning to hold an amnesty for citizens of Commonwealth of Independent States (CIS) countries who are working illegally in Russia. It i s expected that about 1 million citizens o f CIS countries will be amnestied. This may improve the status o f immigrant workers, including their living and working conditions. 6.3. Urban-rural Migration In recent decades in the FSU, the most common migration tendency has been the outflow of population from the rural areas. Rural settlements are most severely affected bythe populationout-migrationbecausethe young, gifted, and educated are the first to leave. Within the outflow, young women are dominant because agriculture does not have many employment opportunities to offer them. 82 In Armenia, the common trend has been reversed in favor of rural migration. The outflow from the cities contributes to the growth o f the rural population. Between 1989 and 2005, the urban population declined by 160,000, while the rural population increased by 74,000 (dejure population; NSS 2003 and 2006). These rural population growth rates are mainly attributed to labor migration resulting from the land reform in the early 1 9 9 0 ~and ~ subsequent better employment opportunities in villages. For many families, especially pensioners, life in the villages helped to support their meager incomes by enabling them to produce some o f the food they consumed. Due to a sharp rise in the prices o f inputs for agricultural production, agriculture itselfbecame more labor intensive. Commuting between the cities and rural areas has been one of the ways of maintaining the high level of rural employment and income. Unfortunately, these possibilities are rather limitedtoday. Transportation costs have also made daily commuting on a regular basis too expensive. From a long-term perspective, it i s obvious that rural communities and agricultural employment, in particular, have no ways o f maintaining their current employment level. Rural areas near cities are becoming more popular residential areas, although local employment opportunities are scarce. Therefore, over the long term, a population outflow constituted mainly o f those within the active working-age group i s to be expected, a fact that will further underminethe pillars o f rural life. 6.4. Remittances Armenia has a very large diaspora community and receives large remittance inflows." Remittances include funds fkom those working temporarily abroad, but also from those who have permanently emigrated and become legal residents o f another country. The official estimate o f remittance inflows was U S 2 8 9 million in 2003, but U.S. Agency for International Development (USAID) experts estimated remittance inflows at roughly US$900 million ann~ally.'~ The true importance of remittances to the Armenian economy is much higher than the ratio o f official remittances to gross domestic product (around 10 percent) and could be three times as important as that. Remittances have had an important role in the economic recovery and in mitigating the impact of the transition on the welfare of households. They are a key financing component o f external imbalance that enabled Armenia to run large deficits with the outside world and maintain living standards and investment. Duringthe boom o f the last three years, remittances have grown strongly, at 23 percent a year on average, by official estimates. Their role in financing the trade deficit has also grown. This i s partly due to the fact that exports 51Armenian diasporan populations are estimated at 6 million to 8 million, which is much larger than Armenia's population o f 3.2 million. The diaspora is spread across the world and is largest in the Middle East, the United States, Russia, and Western Europe. 52 Officially reported remittance estimates are derived from wire transfers, and are usually reported in the official balance o f payment or central bank data. These are widely regarded as underestimating actual remittance flows. In its study, USAID also calculated informal remittances, including the stock o f cash, other financial assets, and real property that emigrants who return to Armenia bringback with them USAID2004). 83 have grown so rapidly and the trade deficit has fallen-whereas remittances grew by 117 percent during 1998-2003, exports grew by 146percent. Table 6.2: Estimates of the Average Remittance that Armenian Temporary Workers in Moscow and Rostov Send Home, 2003 Skilled Seasonal Asphalt Layers Builders Laborers Moscow Average amount sent to Armenia per month $15O-$200 $100 $250 Lengthof stay inMoscow (months) 6 6 4 to 6 How muchmoney is taken home at endof trip $3,000-$4,000 $1,500-$2,000 $5,000-$6,000 Rostov Average amount sent to Armenia per month $15O-$200 $100 NA Lengthof stay inRostov (months) 6 6 NA How muchmoney i s taken home at endo ftrip $2,000-3,000 $1,000-$1,800 NA Source: Interviews with temporary laborers conductedon the street, and discussionswith crew chiefs (USAID2004). Labor migration and associated remittances have multiple effects o n the economy and society. Financial remittances provide a stable source o f foreign exchange that ease fiscal constraints and help finance external deficits. They serve as a potential source o f savings and investments for capital formation and development, although typically investments are made in real assets, such as housing, land, and shops rather than formal-sector financial instruments.Remittances ofmigrants facilitate povertyreductionand mitigate social tensions in a society prone to instability, and they reduce inequality, because the households that receive them would otherwise be at very low levels o f income. According to income data reported to the survey, remittances make up 80 percent o f household income on average for households receivingremittances (Table 6.3) (USAID 2004). Table 6.3: The Impact of Remittances on HouseholdIncomes inArmenia (drams) Average Monthly Income For: Drams Households not receiving external transfer 37,323 Households receiving external transfer: with transfer 85,932 Households receiving external transfer: without transfer 20,604 Average monthly transfer size 65,328 As percent oftotal monthlyincome 76% Source: USAID(2004). Remittances do appear to be going to some of the most vulnerable households in Armenia. The study found that the same percentage o f urban and rural households received remittances, but that rural households received relatively more remittances from CIS countries and relatively less from the United States and Canada. In particular, in Armenia, where child labor i s not apparently an important issue, remittances are more likely to improve the qualityo f a child's education rather thanthe quantity (see USAID2004). 84 Table 6.4: EconomicBenefits and Costs of Remittancesto a Receiving Country PotentialBenefits Potential Costs Are a stable source o f foreign exchange that ease Ease pressure on governments to implement reforms foreign exchange constraints and help finance and reduce external imbalances (moral hazard) external deficits Are a potential source o f savings and investment for Reduce savings o f recipient families and thus capital formation and development negatively impact growth and development (moral hazard) Facilitate investment inchildren's education and Reduce labor effort o f recipient families and thus human capital formation negatively impact growth and development (moral hazard) Raise the standard o f living o f recipients through Migration leads to "brain drain" and negative increasing consumption impacts on economy that are not fully compensated byremittance transfers Reduce income inequality Increase income inequality Reduce poverty Source: Russell (1986). Remittances may have some potential costs. Such transfers may reduce labor efforts o f recipient families and thus negatively impact growth and development (moral hazard). Remittances may also increase income inequality by widening the gap between those households that receive remittances and those that do not (Table 6.4). 85 CHAPTERVII: BUSINESSENVIRONMENTAND LABOR MARKETS A favorable business environment is essential for promoting creation of new businesses, growth of established firms, and job creation. While various surveys indicate some improvement in the business environment in Armenia, it still has features that hinder business establishment and growth. Most complaints of local businessmen are related to factors that can be classiJed as the business environment, such as tax rates, regulatory policy uncertainty, or customs and trade regulations. Starting in 2000, developments in the private sector of export-oriented and import substitution clusters of industry, construction, transport and communications, trade and catering, public utilities, and other services became the biggest source ofjob creation in the Armenian economy. Further improvement of the business and investment climate and facilitation of the access of the private sector (especially small enterprises and newly created businesses) to credits and other financial resources will be crucial for the emergence of new private businesses and substantial job increases to ofset the possible job cuts in other sectors of the economy. This is especially important for small businesses, which are generally labor intensive, and therefore an important means of creating newjobs, even in periods of recession. There are many reasons for this, but theprimary one is that small businesses areflexible and therefore adapt well to the rapidly changing technological and competitive landscape. Worldwide, small businesses are producing more innovations, more jobs, more new developments, and adding more to economic growth than the equivalent number of big businesses. 7.1. Importanceof the BusinessEnvironment The creation of a favorable business environment i s one of important prerequisites for economic growth and private sector development. Barriers imposed by rules, licensing procedures, taxes, and other constraints to investment and job creation hinder the establishment o f new firms and the development o f existing ones. Thus, one o f the priorities o f the national economic policy should be to eliminate financial and organizational barriers to development of entrepreneurship, most o f which are bureaucratic. Armenia ranks quite favorably in terms of the level of economic freedom. The 2006 rankings of the Index o f Economic Freedom, prepared by The Heritage Foundation and the Wall Street Journal, placed Armenia 27`" with a score o f 2.26 (from 42ndin2005 with a score o f 2-58), or mostly free, among the 161 countries surveyed. Armenia was ahead o f Japan, Norway, Spain, and many other highly developed OECD countries.53The Index indicated 53 The 2006 Index o f Economic Freedom measures 161 countries against a list o f 50 independent variables divided into 10 broad factors o f economic freedom, including fiscal burdens and government regulation, property rights, inflation, competitive banking and fiiance laws, and monetary and trade policy. Low scores are more desirable. The higher the score on a factor, the greater the level o f government interference in the economy and the less economic freedom a country enjoys. Among the transitioncountries, Estonia was listed as the 7*-freest economy in the world, the Czech Republic was 21") and Lithuania 23rd. (See: http://www.heritage.org/Research/features/index/.) 86 that the freest economies show a per capita income o f US$30,997, more than twice that o f the "mostly free," at US$13,531, and more than seven times that o f the "mostly unfree." The worst ratings for Armenia were scored in the categories o f property rights, regulation, and informal market. According to the Economist Intelligence Unit, the business environment i s riskybecauseo f weak enforcement o f the country's otherwise favorable business legislation, but also because of the corruption that stems from the excessive role of the State in the economy. Also the judiciary i s underdeveloped and corrupt, which i s a substantial impediment to the enforcement o f contractual rights and obligations, thereby keeping business risk high. In the last few years, Armenia has witnessed certain improvements in some key characteristics of the business en~ironment.'~ By the World Bank Doing Business 2007 data, interms o f Ease-of-Doing-Business ranking, Armenia moved from 37thplace in2005 to 34th in 2006, being the best performer among CIS states, and fourth among transition countries. (World Bank 2006). In particular, according to the European Bank for Reconstruction and Development-World Bank Business Environment and Enterprise Performance Survey (BEEPS), during 2002-05 the ratio o f firms indicating concerns about uncertainty in regulatory policies declined from 63 percent to 36 percent, and the ratio o f firms complaining about customs and trade regulations declined from 51 percent to 29 percent (Figure 7.1). There have been major improvements in the infrastructure in the last three years, and the share o f firms complaining about the poor quality o f telecommunications services has dropped from 26 percent to 12 percent; complaints about electricity cuts have fallen from 30 percent to 10 percent, and dissatisfaction with the transportation network has declined from 19 percent to 13 percent. Armenia stands quite favorably compared with other Europe and Central Asia (ECA) and Commonwealth o f Independent States (CIS) countries with respect to corruption and business licensing. Also, labor regulations, and skills and education of workers, are rarely mentioned by employers as constraints to business expansion. In some other aspects, the business environment has not improved much, or improvements in relevant institutions have fallen behind the overall progress in the economy. Despite annual double-digit growth rates in the 2000s, the share of businesses indicating macroeconomic instability as a problem has increased from a high 52 percent in 2002 to 56 percent in2005. Around half o f firms still complain about hightax rates, the poor performance o f tax administration, the high cost o f financing, and the lack o f access to financing (EBRD-World Bank 2005).55Nevertheless, a number o f economic issues, such as macroeconomic instability, tax administration and tax rates, competition policy, customs management, and the cost o f and lack o f access to financing, remain key constraints for doing business inArmenia. 54 On the business environment in Armenia, see also IFC-World Bank (2004); EBRD-World Bank 2005 and World Bank (2005a). 55The BEEPS survey reflects subjective views of local employers. For example, the overall tax burden (taxes- to-GDP ratio) in Armenia was 14 percent in2004, which is almost three times lower than in many Central and Eastern Europe countries; still, however, tax rates were indicated as a problem by as many employers in Armenia as inother parts of the region. 87 Figure7.1: ProblemsofDoingBusiness(percent of firms indicatinga problem) Labor regulations Skills and education of workers Armenia 2002 Business licensing and permits Customs and trade regulations Corruption Uncertainty about regulatory policies Access to financing Tax rates Cost of financing Macroeconorric instability Tax administration 0 10 20 30 40 50 60 70 Source: EBRD-WorldBank (2005). 7.2. Challenges to Creating a Business Entrepreneurs still face serious challenges to launching a business inArmenia. They can expect to go through 9 steps to launch a business over 24 days, on average, at a cost equal to 5.1 percent o f gross national income (GNI) per capita, and must deposit at least 3.3 percent of GNI per capita in a bank to obtain a business registration number. This i s at the level or below the level in other transition countries, but the procedures are much more complex compared to high-income countries (Table 7.1). InAustralia it takes only two steps and two days to establish your ownbusiness. The Business Information System introducedinFinland in 2001 is maintained jointly by the National Board o f Patents and Registration and the Finnish Tax Administration. Entrepreneurs only need to fill out a single registration application when establishing a company submittedto one contact point. InItaly, since 1998, in each municipality, "one-stop-shops" have been established for existing and potential entrepreneurs conceived o f as help desks and geared to customer satisfaction. They provide companies (mainly start-ups) with information in order to reduce administrative, time, and financial burdens on business. Similar programs and approaches have also been developed in other countries in Europe (EC 2003b). Because business start-ups internationally account for 88 a large portion (from 25 to 50 percent) o f newjob creation, excessive obstacles to entry risk translate into lower employment growth. Table 7.1: Starting-a-Business Indicators-Measures to Register A BusinessFormally inSome Countries in2006 Number of Time (days) Cost (percent Minimum procedures of incomeper capital capita) (percent of Country incomeper capita) Armenia 9 24 5.1 3.3 Azerbaijan 15 53 9.5 0.0 Georgia 7 16 10.9 3.7 CzechRepublic 10 24 8.9 36.8 Hungary 6 38 20.9 74.2 Poland 10 31 21.4 204.4 Slovak Republic 9 25 4.8 39.1 Europe and Central Asia 9.4 32.0 14.1 53.9 High-income OECD 6.2 16.S 5.3 36.1 countries Best performer 2 (Australia, 2 (Australia) 0 (Denmark) 0 (66 countries) Canada, New Zealand) Source: World Bank (2006). Even more complex procedures are reported with respect to complying with licensing and permit requirements for ongoing operations in Armenia. It takes 18 steps (20 steps in2005) and 176 days (112 days in2005) to complete the process, and costs 43 percent (65 percent in 2005) o f income per capita. Some other business transactions, such as registering property or closing a business, are less burdensome compared to other transition countries but still require significant efforts from and costs to entrepreneurs (World Bank 2006). Access to external financing on reasonable terms is viewed as the major constraint on business growth and increased employment The BEEPS 2002 and 2005 participants mentioned stringent conditions on the financial markets as the one o f the major constraints to expansion: maturities on commercial loans are short; interest rates and collateral required are high; and other financial instruments, that is, micro-credit and small-business financing institutions, leasing, and so forth, are not commonly available. In 2005, surveyed firms in Armenia on average financed 67 percent o f their investment needs from internal funds and retained revenues, which i s some progress compared to 85 percent in 2002. Internal funds as a major source o f financing in 2005 were on average 78 percent for CIS states, and 70 percent for ECA countries (EBRD-World Bank 2005). Taxes and tax administratiodregulationi s the area most seriously hinderingthe growth and operations of Armenian businesses. While a medium-size company in Armenia must pay 53.8 percent o f its gross profit intaxes, compared to 50.2 percent inthe ECA region and 45.4 percent inhigh-income OECD countries, labor taxes inthe country are modest (EBRD- World Bank 2005). Infact, according to the recent World Bank study on labor markets inthe 89 region, Armenia has one o f the lowest labor taxes among the transition countries. There are two alternative indicators o f wage tax burden: non-wage labor cost and tax wedge on the labor market.56Based on the data on nominal tax rates, in 2004 the tax wedge on average wages o f 43,400 drams was estimated at 24.8 percent o f the labor costs (see also World Bank 2005b). Non-wage labor costs were estimated in 2004 at 19.6 percent.j7 According to the enterprise data reported to the NSS (Form lA), the actual costs to organizations and firms o f social insurance contributions equaled, in 2003, 17.1 percent of total labor costs, and non- wage labor costs amounted to 18.8 percent o f total expenditures on labor (NSS 2004). In OECD countries, the total burden on the labor market i s 51.8 percent among European member countries, and 37.3 percent for non-European member countries. High labor taxes lower labor demand by raising labor costs. But they also lower labor supply by reducing the real consumption wage. The low wage tax burden in Armenia has contributed to employment, but was still insufficient to offset the impact o f other market factors determining highunemployment. The effective average rate (tariff) of social contributions in Armenia has declined from 25.6 percent in 2001 to 20.9 percent in 2004. This can be attributed to the higher share o f taxable income in upper-income brackets with relatively lower tax rates. The mean taxable salary increased from 27,700 drams in 2001 to 44,200 drams in2004, but the average social contributionper person increased "only" from 7,100 drams per months to 9,200 drams for the same period. Also, the number o f active payers o f social contributions has increased fiom 333,400 individuals to 420,100 individuals (SSI 2005). By comparison, inthe new European Unionmember states (EU8), the statutory rates o f social security contributions in2004 varied from 33 percent in Latvia and 34 percent in Lithuania to 48 percent in the Czech Republic and 49 percent in Slovakia (World Bank 2005~).On the other hand, relatively low social contributions and the limited number o f payees (39 percent o f the total employed in 2004 paid social contributions) result in much lower levels o f social transfers (pensions, unemployment benefits, and social assistance) compared to other countries in transition. In 2003, the ratio o f average pension to average wages in Armenia equaled 19.4 percent-the second lowest inthe CIS following Georgia (Annex Table 12). 56Non-wage labor cost is defined as the ratio between payroll taxes and gross wages, The tax wedge on labor is the difference between the labor cost to the employer and the take home pay (net wage) expressed as a percentage o f total labor cost. 57According to the Law on Income Tax o f 1997, income tax rates are 10 percent o f taxable incomes up to 80,000 drahms, and 8,000 drahms plus 20 percent o f the amount exceeding 80,000 drahms. According to the Law on Mandatory Social Insurance Contributions o f 1997 (as amended), rates o f social contributions for each hired employee are 5,000 drahms for taxable incomes up to 20,000 drahms, 5,000 drahms plus 15 percent o f the sum exceeding 20,000 drahms o f income between 20,000 and 100,000 drahms, and 17,000 drahms plus 5 percent of the sum exceeding 100,000 drahms. 90 Figure 7.2: Underemployment Dueto Labor Regulations inArmenia, andECA, EU8, and CIS States, Percent, by the BEEPS2002 and 2005 ECA EU8 cis Armenia 0 5 10 15 20 Source: EBRD-WorldBank (2005). 7.3. The Role of Labor Regulations There are relatively few complaints among the employers in Armenia with respect to rigidity of labor legislation. However, there i s a potential for employment growth through various labor market policy measures that the government can undertake, in addition to facilitating the general macroeconomic and business environment (see Chapter VIII). The 2005 BEEPS survey o f local employers indicates that employment in the formal sector can be increased by more than 16 percent if labor regulations were eased. Surprisingly, despite the recent positive steps inthis direction associated with the approval o f the new Labor Code in 2004, in the last three years the ratio of employers indicating labor regulations as a constraint to job creation has slightly increased. This points to the necessity o f adjusting labor relationsinparticular to reflect the changes inthe overall economic situation. In practice, for many firms and workers, labor regulation is completely bypassed, so that the labor market i s virtually unregulated. As noted, according to the 2004 LFS, 23 percent o f employees in firms and organizations worked based on an oral agreement, evidence that it i s necessary to deepen the legal and contractual regulation o f labor relations (NSS 2005b). The weak regulation o f the labor market may have helped employment flexibility, but it also appears to have had tradeoffs in promoting informalization o f the economy (through its avoidance), lower worker productivity, and reduced worker welfare, such as low wages, growth o f in-kindsubstitutes, and wage arrears. 7.4. The Role of SkillLevel andEducation Employers in Armenia are relatively more satisfied with the skills and education of their workers compared to most other countries in transition. According to the BEEPS 91 2005 survey, only 13.5 percent o f surveyed firms indicated this factor as a constraint to business expansion. Ths i s much lower than in other ECA countries, and especially compared to new EUmember countries, indicating also that labor inArmenia i s still largely employed in low-skilled but labor-intensive sectors o f the economy. According to the World Bank Investment Climate Survey, in Armenia 65 percent o f firms offer formal training to their employees, which is much less than the average for ECA countries of 75 percent of firms (including more than 90 percent o f firms in the Czech Republic, Estonia, Poland, Slovakia, and Slovenia). Also, only about 17 percent o f permanent skilled workers received training in the last year, compared to 26 percent in ECA countries (and up to 72 percent o f workers in Sl~vakia).~' Firms spend a relatively small amount o f total labor costs on training and retraining o f their employees-0.4 percent on average between 1998 and 2003 (NSS). "Savings" on upgrading o f the labor force may have a negative impact on further improvements in labor productivity and in labor market mobility (the movement o f labor from low-productivityjobs to high-productivityjobs). Figure 7.3: Skills and Education of Available Workers As a Constraint to Business Expansion ECA RJ8 Poland Lithuania Latvia Hungary Estonia Czech R Armenia 0 10 20 30 40 50 60 70 Source: EBRD-WorldBank (2005): BEEPS 2002 and2005. On the other hand, there is a demand especially for skilled workers. The majority of 300 enterprises surveyed by the FIAS 2003 study (62 percent o f firms for skilled workers and 64 percent for unskilledworkers) believe that the level o f employment at their companies i s at a satisfactory level for the normal operations o f their businesses (Figure 7.4). That is, 28 percent o f firms believe that the number of skilled employees i s low, 11 percent believe the number of unskilled employees is low, and 2.3 percent believe the number of skilled employees i s high, and 7.3 percent believe that the number o f unskilled employees i s high. The demand for highly skilled workers tends to be higher inlarge enterprises, and in firms in Yerevan. Among the companies in which the employment level is believed to be too high, 58See: http://m.worldbank.org/InvestmentClimate/. 92 the management states that the main reason for not firing employees i s the belief that it i s wrong to fire people (52 percent o f firms) (IFC-World Bank 2004). Enterprise surveys indicate that there is a potential for employment growth. Average capacity use in Armenian firms (as surveyed by BEEPS 2005) equals 82 percent, indicating that new jobs can be generated usingthe available capital stock. Moreover, employers point out that the optimal employment level intheir firms compared to the current level implies an expansion of the workforce by 12.1 percent. This i s more than the average inthe ECA region o f 8.6 percent o f potential new employment in existing firms. Firms are actively involved in innovations: 72 percent o f surveyed firms in Armenia indicated that they have developed a major new product line, upgraded an existing product line, introduced new production technology, or obtained a new licensing agreement in the last three years (World Bank 2005a). This shows that most firms are actively adjustingto market realities. Figure7.4: Current Level of Employmentin SurveyedFirms(inpercent); FIAS 2003 Study Employmentlevel Do not High by High by 10 High by5 - Just about Low N/A employ more than - 20% 10% right this kind of 20% worker Source: IFC-World Bank (2004). 93 CHAPTER VIII: LABORLEGISLATIONAND LABOR MARKET INSTITUTIONS Since the approval of the new Labor Code in 2004, Armenia has had modern labor legislation. However, due to weak capacity of the Labor Inspectorate and other labor market institutions, including socialpartners, the enforcement of labor regulations is weak, and they are ineffective in a large informal sector. Funds currently allocated to labor market programs are very limited, which, naturally, will have a negative impact on the quality and effectiveness of services provided. A comprehensive assessment of the results of state employment programs implemented by the State Employment Service in Armenia has not been conducted, which otherwise could have been used to adjust the policy, as necessary, and redistributefinancial allocations to various sub sectors and activities within the sector. Improvement of the existing information system is a key element in enhancing services for supporting job seekers in their job search, including information on vacancies, programs, and availablefunding. 8.1. State Intervention inthe Labor M a r k e t State intervention in the labor market in all countries originally aims to create conditions that are more conducive to job creation, protecting employment, and improving working conditions. Nonetheless, as international experience shows, increased state intervention does not correspond to better labor market results. Moreover, numerous empirical studies o f the main indicators o f labor markets and various forms o f labor market regulations in different countries prove the opposite. The high level o f employment protection mandated by legislation reduces the flexibility o f the labor market and increases direct and indirect costs for employers o f hiring or firing workers, thus protecting the interests of those who have permanent employment, and at the same time reducin the chances o f job seekers o f finding employment. In addition, rigid labor legislation5 f can significantly reduce the possibilities for expanding employment, especially formal employment. The effectiveness of state labor market policy i s often evaluated according to five main areas of state intervention (Figure 8.1). Each, by definition, has a distinctive role in the labor market: Employment protection, the main objective o f which i s to ensure the stability o f labor relationships, and refers to the legislatively defined conditions and procedures for hiringand firing employees, andthe limitations for fixed-tern employmentcontracts. Working conditions, which have a significant impact on labor force supply and the flexibility o f the labor market, and mainly refer to legislatively defined norms for 59As used in this document, the word "legislation" includes laws, decrees, ordnances, regulations, and other bylaws. 94 working hours and work remuneration. 0 Active employment policy, which aims to reduce the structural mismatches between supply and demand formed in the labor market, mainly by increasing employment amongjob seekers. Passive employment policv, which aims to mitigate the drastic changes in living standards resulting from unemployment, but which can also lead to quantitative and structural changes inlabor force supply. 0 Taxes and compulsory social security payments levied on salaries, which serve the objectives o f fiscal or social policy, might, nevertheless, have a significant impact on the labor market, particularly with regard to the ratio of formal to informal employment. In addition, in a market economy, collective bargaining can play an important role in determining wages (and other conditions of work). Intransition countries, the State also plays an active role insocial dialogue. Figure 8.1: Government Interventions and Regulation inthe Labor Market Taxes on Labor 95 8.2. Employment ProtectionLegislation A flexible (adaptable) labor market facilitates economic restructuring, promotes productivity growth, contributes to the economy's competitiveness, and helps the economy absorb various supply and demand shocks. Labor market regulations, as one o f the key state interventions in the labor market, aim to protect workers from arbitrary, unfair, or discriminatory actions by their employers while addressing potential market failures stemming from insufficient information and inadequate insurance against risk. Employment protection regulations, especially on hiring and termination, are often hotly debated not only because they define the job security of workers but also because they have potentially important implications for economic and labor market performance.60 Typically employment protection legislation (EPL) makes it harder for certain groups, including women and displaced older workers, to enter or reenter the labor market, at least on an open-ended contract; increases long-term unemployment; and seems likely to strengthen labor market "insiders." In addition, by discouraging hiring and firing, EPL may slow adjustment to shocks and impede the reallocation o f labor, with potentially negative implications for productivity growth and adaptation to technological change. But stricter EPL makes layoffs more difficult for employers, and makes the process more lengthy and costly for employers. Enterprise management should thus be forced to look for alternative solutions to dismissals, such as the enhancement o f functional flexibility o f personnel through better human development policy and stronger motivation o f workers in the framework of, for example, enterprise restructuring, technological upgrading, and improved marketing strategy. Recent studies also indicate that greater flexibility in the host country's labor market relative to that inthe investor's home country i s associated with larger foreign direct investment flows (see Javorcik and Spatareanu2004). Less is known about the impact of EPL in transition countries. Based on the analysis o f EPL in eight transition countries, Cazes and Nesporova (2003) concluded that the impact o f EPL on labor market performance and labor market flows intransition countries seems to be rather modest. Based on the analysis o f labor data, Haltiwanger and others (2003) found that employment protection legislation i s associated with lower employment rates and a higher long-term unemployment share. However, these relationships differ between OECD and transition economies. Inparticular, the relationshipbetweenthe employment rate and EPL i s weaker (less negative) for transition economies, but the relationship between EPL and long- term unemployment i s stronger (more positive) in transition economies. However, these conclusions are based on country comparisons o f particular economic and labor market indicators. The evidence i s sketchy as to whether the reforms inlabor legislationper se inthe region have had a major impact on the ground. Besides, employment laws are often ineffective because o f evasion, weak enforcement, and failure to reach the informal sector. Nevertheless, higher EPL tends to be associated with lower excessjob reallocation, lower job creation, and lowerjob destruction, and thus the analysis o f labor legislationi s o f a particular importance for Armenia. 6o See, for example, two of the latest studies on the topic,OECD (2004b) and Young (2003). 96 Until 2004, the labor market in Armenia was regulated by the Labor Code of the Armenian SSR adopted as early as 1972 during the Soviet period. The Law on Employment Promotion, adopted in 1996, and the Law on Work Remuneration of 2001 were other major legislative acts on labor. Since 2005, labor relations in Armenia's labor market have been regulated by the new Labor Code adopted by the National Assembly o f Armenia inNovember 2004.61The new Law on Employment and Social Protection in Case o f Unemployment was adoptedby the National Assembly inOctober 2005. There are two main methodologies for assessing the rigidity of EPL, one developed by the World Bank and another by the OECD. The basis o f the two methodologies i s the assumption o f the general theory o f equilibrium that a highlevel o f labor market regulation and limitations reduces labor market flexibility (responds to changes in supply and demand with more difficulty), thus reducing the possibilities for expanding employment. Of course, the rigidity o f legislation does not yet mean that the labor market i s by definition less flexible. Other factors also have an impact on labor market flexibility, which might be more consequential.62 Also, the experience of European countries shows that labor market flexibility largely depends on how widespread and well-organized the trade unions are. Undoubtedly, active and passive employment programs implemented by the State also can have a significant impact on the flexibility o f the labor market. Figure 8.2: Labor Regulations as a Problemof DoingBusiness (percent of firms indicating labor regulations as a problemdoing business, 2005, Accordingto the BEEPSsurvey) I II2005 ECA 0 2002 EU8 Armenia I 0 5 10 15 20 25 30 35 40 BEEPS=World Bank Business Environment and EnterprisePerformanceSurvey. Source: EBRD-World Bank (2005). Untilthe adoption of the new Labor Code of Armenia in November 2004, Armenia's labor law, in accordance with the Doing Business methodology developed by the World Bank, could be classified as less rigid among the transition countries (Figure 8.2). However, some labor market indicators in Armenia are more rigid than the average for 61 Inorder for the new Labor Code to become fully operational, more than 60 bylaws needto be adopted. 62 Simply evaluating the law and institutional set up is not adequate for understanding the Armenian labor market, because current practice often diverges markedly from formal arrangements. The actual arrangements for a specific aspect of the regulatory regime can be very difficult to capture. 97 transition countries, particularly those related to the flexibility o f firing.63The EBRD-World Bank BEEPS survey among 200 employers in Armenia also confirmed that labor relations are less problematic for doing business in the country compared to most other transition countries. Surprisingly, despite clear improvements in the 2004 Labor Code, the share of local entrepreneurs who are dissatisfied with EPL has increased (Figure 8.3). This indicates that labor regulations need constant adjustments depending on the country's economic situation. Figure 8.3: Labor Market Regulation inArmenia (before the approval of the new Labor Code) and inSelected Transition Economies FiringCosts (weeks) Rigidity of Employment Index Difficulty of Firing Index Rigidity of Hours Index I 50.4 Difficulty of Hiring Index 0 10 20 30 40 50 60 Average for group of countries (CEE, CIS, and Baltic States) 0Armenia Source: World Bank DoingBusiness Database. The methodology for assessing the level of protection o f employment developed for OECD countries in 1999 and modified in2004 was applied for drawing the comparison o f labor market r e ulation between Armenia and countries with a longer history of a market economy.64g Employment protection in Armenia, assessed by using the OECD methodology, reveals that before adoption o f the new 63 The original World Bank Doing Business methodology was developed by Juan Botero, Simeon Djankov, Rafael L a Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer. This methodology, which is a system for evaluating indexes calculated based on 4 separate groups o f indicators and 17 individual indicators, was used to assess labor market regulation inArmenia. The main advantage o f this system is that it allows for comparisons to be drawn among the levels o f labor market regulation invarious countries. That is probably the reason behnd its being fairly general, and it should be approached with certain reservations when it is applied to the assessment o f legislative changes ina given country. 64See OECD 2004b. The OECD methodology uses a more sensitive evaluation scale. 98 a 5 m 5hs5 a 3 8 vi0 .- d 8a0 w to hl m 3 m a a h 0 m s"a m 5 m 3 42 m 0 d g m 0 hl 3 a, E Y 8$ d * 3 3 d 8 a, 3 d i.4 m cd 3a, 2 a5 a, 2 2 3 a h 38a0 28 00 hl 4 Labor Code, Armenia could be classified among countries with strict labor regulations. According to our calculations, the EPL rigidity rate equaled 3.2 on a scale o f 0 to 6. Compared with 29 OECD countries) the rate was higher in only three countries with even more rigidlabor legislation-Mexico, Portugal, and Turkey (see Figure 8.4). Figure 8.4: Assessment of Labor Protection Legislation inArmenia (for old and new Labor Codes) and Comparisonswith OECD Countries S`ource: For Armenia: authors' estimates; for OECD countries: OECD2004b. The new labor law enabled significant progress in ensuring labor market flexibility. The new Labor Code of Armenia can be considered as one o f the most flexible labor laws among the transition countries. The Rigidity-of-Employment Index for the new legislation evaluated using the World Bank methodology is 22 compared to the former 36. The level of labor market flexibility in Armenia increased significantly while using the OECD methodology. The overall index for rigidity of employment protection legislation declined to 2.2 compared to the former 3.2, which, compared with OECD countries, classifies Armenia as having average rigidity o f employment protection legislation. In assessing individual indicators o f the rigidity of EPL in Armenia) fixed-term employment arrangements received the highest coefficients o f rigidity. The main reason for this i s that the institution o f so-called "temporary work agency" i s not defined in either the old or the new legislation, although no direct restrictions are provisioned either. Therefore, it can be assumed that the new Labor Code o f Armenia will primarily have a positive impact on the labor market, particularly with regard to employment opportunities for young, first-time job seekers. Armenia needs to develop modern Labor Inspectorates to effectively enforce core worker rights, supervise the implementation of labor regulations, and to provide technical assistance and advisory services to enterprises, and perhaps to participate in 100 mediation of labor disputes. The Law on Labor Inspectorates was adopted in2004, and the Inspectorates and their local offices are in the process of recruiting their staff. A detailed legislative framework regulatingtheir daily activities i s still absent. 8.3. The Capacity of the State Employment Service Employment services fulfill important brokerage functions, matching available jobs with job seekers, but the employment service itself cannot create jobs. It has no direct influence on either labor supply or demand. On the other hand, labor market interventions have a big potential for improving labor market performance. Whether or not this potential materializes depends on a number o f factors including external factors, such as stable macroeconomic conditions, a favorable investment climate and an enabling business environment, and a competitive product market. In addition, there are important internal factors-policy design and implementation. Since 1991, Armenia has established the two main programs found in OECD countries and other transition countries: (a) an unemployment benefit (passive) program, providing temporary cash assistance to the unemployed; and (b) active labor market programs (ALMPs) including training, job counseling, and public works. State regulatory functions in Armenia in the sphere o f employment are delegated to the State Employment Service (SES), which i s a nationalbody o f the executive branch operating under the auspices o f the Ministry o f Labor and Social Issues. The main functions of the State Employment Service are: Collection, analysis, and dissemination o f informationon the labor market; Collection o f information from employers on availablejob vacancies; Registrationo fjob seekers; Granting the status o f unemployed to unemployed job seekers, and allocation o f unemployment benefit and financial assistance (suspensiodtermination o f those payments); e Advisory services for the professional orientation o f job seekers, provision o f information on available vacancies, and support infindingjobs; Organizing training and requalificationcourses for the unemployed; Financial support to the unemployed for business start-up and creation o f additional jobs inenterprises; Ensuringthe participationo fthe unemployed inpaidpublic works; and Devisingmeasures for improving employment o f disabled persons. The State Employment Service (SES) has an extensive network of regional employment centers (RECs), which covers the entire country.6' The structure o f the SES includes 51 65 Private employment agencies are currently active mainlyinYerevan, andthey havevery limitedoperations in rnarzes. In the near future, regional centers of the SES will continue to be the main service providers in the 101 regional centers41 inmarzes and 10 inYerevan city. Marz centers are basically located in former district centers (each marz has 3 to 5 district regional centers), and Yerevan (having the status o f a marz) consisting o f 12 municipal communities. Services provided by the SES are easily accessible for beneficiaries and are provided free o f charge. However, the workload o f territorial employment offices i s very unequal. In particular, as o f December 2004, the number o f registered job seekers per one employee o f the SES varied from 56 (Talin regional center) to 1,132 (Gyumri regional center), while the number o f registered unemployedvaried from 23 (Ashotsk regional center) to 911(Gyumriregional center). Figure 8.5.: Annual Expenditure on PublicEmployment Services and Administration per PES Total Staff: InternationalComparisons (2002,2003; U.S. dollars) Armenia Romania Latvia Estonia Slovakia Lithuania Slovenia Spain Finland Germany Austna Denmark Netherlands 77,778 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 Source: State Employment Service of the Republic of Armenia; ILO 2003. Compared to other transition countries, the SES in Armenia i s understaffed. (Annex Table 21; Annex Figure 1l).Territorial and local employment offices are ill-equipped and motivated to provide modern employment services. According to the assessment o f directors o f regional and local employment services, in around 40 percent o f offices the working conditions are very poor or poor, and only in 6 percent o f offices are the conditions good or excellent. Only twelve percent o f regional centers maintain information mainly in electronic form, while 88 percent still continue to manage information in paper-based forms. The majority o f employment centers (72 percent) have one computer per 3 to 7 employees, and 20 percent o f offices are even less equipped with modern information technologies. The employees are not sufficiently motivated, having an average monthly wage o f around 37,000 drams (US$70) in2004. sphere of employment. 102 8.4. Active Labor Market Programs In Armenia, exi enditures on labor market programs are limited in both absolute and relative terms.6 Inrecent years, allocations to the labor market programs amounted to 0.4 to r 0.5 percent o f the current expenditures from the consolidated budget, and their share o f gross domestic product (GDP) was less than 0.1 percent. By comparison, the average expenditures on labor market programs in OECD countries in 2002 were 1.8 percent o f GDP, which i s usually even higher in European countries (Annex Table 23; Annex Figure 9). Expenditures inArmenia are also significantly lower than the averages for Central and Eastern European countries, where expenditures on labor market programs in late 1990s amounted to 0.4 percent o f GDP (Betcherman, Olivas, and Dar 2004). The annual expenditures on employment services are very limited, and in around 80 percent o f centers amounted to less than 5,000 drams perjob seeker in2004, while the average costs for all regional centers were less than 2,000 drams (less thanUS$4) perjob seeker (Figure 8.6). Figure8.6: Distributionof RegionalEmploymentCenters byBudgetedAnnual Expenditure per OfficiallyRegisteredJob Seeker (2004) Morethan 10,000 drams 5,001-10,000 dram 2,000-5,000 drams i I I I I I Less than 2,000 dram 0% 10% 20% 30% 40% 50% 60% Source: Survey of RegionalEmploymentCenters 2005. Another indicator used to evaluate public expenditures on labor market programs i s the ratio of total expenditures to GDP per 1 percentage point of registered unemployment. According to this indicator, Armenia lags far behind both developed and Central and Eastern European countries. The picture becomes even worse if unemployment rates from the labor force surveys are used as the basis for calculations (Figure 8.7). 66During 1998-2003, expenditures on public employment service and administration and expenditures on public works were financed fi-om the state budget, and the expenditures on the remaininglabor market programs were financed from the State Social Insurance Fundof the Republic of Armenia. 103 Table 8.2: Armenia: Public Expenditure on Labor Market Programs, 1998-2003 1998 1999 2000 2001 2002 2003 Million Drams 1. Public Employment Services and Administration 211.2 195.6 187.5 182.6 185.0 243.5 2. Labor Market Training 22.4 29.9 38.8 30.3 8.0 26.3 3. Subsidized Employment 24.1 29.8 36.0 256.4 408.8 456.4 Including: Support o f UnemployedPeople Starting Business 24.1 29.8 36.0 0.0 0.0 0.0 Public Works 0.0 0.0 0.0 256.4 408.8 456.4 4. Measures for the Disabled 0.5 10.8 0.0 15.0 0.0 3.5 5. Unemployment Compensation 714.7 1,323.8 543.5 349.3 450.8 397.4 TOTAL 972.9 1,589.9 805.8 833.7 1,052.7 1,127.1 Active Measures (1-4) 258.2 266.1 262.3 484.4 601.9 729.7 Passive Measures (5) 714.7 1.323.8 543.5 349.3 450.8 397.4 \ I % of TotalExpenditure TOTAL 100 100 100 100 100 100 Active Measures (1-4) 26.5 16.7 32.5 58.1 57.2 64.7 Passive Measures (5) 73.5 83.3 67.5 41.9 42.8 35.3 % of TotalExpenditure onActive Measures Active Measures (1-4) 100 100 100 100 100 100 Public Employment Services and Administration 81.8 73.5 71.5 37.7 30.7 33.4 Labor Market Training 8.7 11.2 14.8 6.3 1.3 3.6 Subsidized Employment 9.3 11.2 13.7 52.9 67.9 62.5 Measures for the Disabled 0.2 4.1 0.0 3.1 0.0 0.5 % of GDP TOTAL o.io 0.16 0.08 0.07 0.08 0.07 Active Measures (1-4) 0.03 0.03 0.03 0.04 0.04 0.04 Passive Measures (5) 0.07 0.13 0.05 0.03 0.03 0.02 Ratio of GDP Expenditure to UnemploymentRate TOTAL 0,011 0.014 0.007 0.007 0.007 0.007 Active Measures (1-4) 0.003 0.002 0.002 0.004 0.004 0.004 Passive Measures (5) 0.008 0.012 0.005 0.003 0.003 0.002 Memorandum Items Total Expenditure (1-5) % of Consolidated Budget TotalExpenditure 0.37 0.61 0.31 0.32 0.40 0.43 % of Consolidated Budget Total Current Expenditure 0.47 0.77 0.39 0.40 0.51 0.54 'Calculatedbased o n officially registered unemployment rate. Source: National Statistical Service; State Employment Service; Ministry of Finance and Economy. As far as the expenditures on active and passive policies are concerned, the most recent period, 1998-2004, can be divided into two parts. During 1998-2000, allocations to passive policies (unemployment benefits and financial assistance) accounted for around 75 percent o f total expenditures on labor market programs. During 2001-04, expenditures on active policy measures constituted a larger share o f public spending on labor market programs, an average o f 60 percent out o f total costs during the period. Such change in the structure o f expenditure resulted from the launch o f a large-scale public works program ("Benefit for Work") in 2001. In Armenia, the share o f allocations to paid public works in ALMPs is larger, but expenditures on training are much less compared to OECD member countries and to a number o f transition countries (see Martin and Grubb 2001). In 2004, fewer than 600 unemployed were trained or retrained using SES funds, and professional 104 training o f 210 job seekers was provided using donor and nongovernmental organization funds. Figure 8.7: PublicExpenditureon Labor Market Programs per Officially Registered Job Seeker (Drams) 10,000 -9,059 8,000 E 6,000 h TJ 4,000 2,000 0 , - - 1998 1999 2000 2001 2002 2003 Total Active masures =Passivemeasures Source: National Statistical Service; State Employment Service. Public works programs are the main activity of ALMPs. In 2004, 11,800 job seekers participated in the "Benefit for Work" program, and 555 projects were implemented. As stated in the recent Armenia Poverty Assessment Report (World Bank 2003a), while public works have been, and continue to be, an important tool in the government's employment policy, one needs to be cautious about the cost-effectiveness o f public works programs, and one needs to be clear about the objective of such programs: are they meant to transfer income to poor households, build infrastructure, or do both? It may well be that from the point o f view of poor families facing income and consumption constraints, public works schemes represent a diversion o f resources away from their direct welfare needs. The rate of participation of registered unemployed in active policy programs (measures) implemented by the Armenian SES i s high and has shown an upward trend. In 2003, 35.5 percent o f officially registered job seekers were involved in active policy programs compared to 25 percent in 2000 (Table 8.3). At the same time, individuals receiving counseling services dominate among the participants o f ALMPs. The share o f recipients o f counseling services as percent o f officially registered job seekers increased by about 5 percentage points-from 24.6 percent in 2000 to 29.3 percent in 2003. Meanwhile, around 40 percent o f those asking for counseling services were interested inissues related to labor and employment legislation. The number o f people participating inpaid public works during2001-03 variedbetween 5 and 7.5 percent o f the total number o f officially registered job seekers, while very small numbers-below 0.3 percent o f the total number o f officially registeredjob seekers- participatedintraining programs. 105 Table 8.3: Involvementof OfficiallyRegisteredJob Seekers inActive PolicyMeasures 1998 1999 2000 2001 2002 2003 IndividualParticipants Counseling Services 49,688 78,525 46,801 33,294 44,3 18 OfWhlch: Related to the Following Issues: Selection o f Specialization and Professional 6,290 8,280 5,822 6,831 7,858 10,100 Orientation Professional EducatiodTraining 3,586 4,997 4,539 3,166 4,786 8,179 Changing Current Job 2,351 3,992 2,065 2,352 1,927 1,799 Labor and Employment Legislation 24,652 44,546 20,117 14,017 12,434 16,257 Other Issues 12,809 16,710 14,258 6,807 6,289 7,983 Labor Market Training 408 422 573 503 184 479 Support o f Unemployed People Starting ... ... ... ... ... ... Business Public Works 0 0 0 8,781 11,700 8,874 Measures for Disabled 10 60 0 150 0 37 % of Officially RegisteredJob Seekers Counseling Services 29.7 44.7 24.6 19.4 21.0 29.3 OfWhich: Related to the Following Issues: Selection o f Specialization and Professional 3.8 4.7 3.1 4.0 5.0 6.7 Orientation Professional EducatiodTraining 2.1 2.8 2.4 1.9 3.0 5.4 Changing Current Job 1.4 2.3 1.1 1.4 1.2 1.2 Labor and Employment Legislation 14.7 25.4 10.6 8.2 7.8 10.8 Other Issues 7.7 9.5 7.5 4.0 4.0 5.3 Labor Market Training 0.2 0.2 0.3 0.3 0.1 0.3 Public Works 0.0 0.0 0.0 5.1 7.4 5.9 Memorandum Item: Officially Registered Job Seekers (annual 167,219 175,514 190,505 170,785 158,437 151,225 average), Individual Source: National Statistical Service; StateEmployment Service. I t is a common problem worldwide that employers register only a minor part of available vacancies with the SES. By end-2006, only 1,167 vacancies were registered with the SES (940 vacancies for workers and 227 for employees), and 1,322 were registered by mid-2005. Considering the education level o f unemployed people-around 14 percent have higher education and 79 percent secondary vocational and secondary general education-and the large number o f registered unemployment, these data confirm that the SES can have only a limited impact o fjob placements. In2004, some 7,900 registeredjob seekers, or 5.4 percent o f the total, found a job through the Territorial Centers o f Employment. The mismatch o f labor demand and supply also reflects social considerations. Very few job seekers are willing to step down to positions lower than the ones they held previously, or accept jobs for which they are overqualified. The job seekers-to-vacancy ratio inArmenia is extremely high. The vast majority ofjob openings are not advertised, and most employment opportunities are hidden. A primary reason i s that most employers do not need to advertise. They have enough applicants without it.Another reason is that employers prefer to hire on areferral from someone they trust. 106 8.5. UnemploymentBenefits The fraction of the jobless that receive unemployment benefit, and the average size of the benefit, are very small. According to the Law on Employment Promotion adopted in 2002, to qualify for the benefit, individuals must have 26 weeks o f paid work in the last 12 months. The replacement rate o f the benefit equals 70 percent o f the amount o f the last wage and can be paid for 26 weeks in a 12-month period. In other cases, the minimum benefit i s paid at a level determined by local authorities. InDecember 2006, the unemployment benefit was received by 10,000 people, or around 11 percent o f the registered unemployed, and the average unemployment benefit was 8,600 drams (about US$20) per month. The benefit replacement ratio was around 57 percent o f the minimum wage, and 14 percent of the average wage (www.armstat.am). The law also allows payment o f unemployment assistance equal to 30 percent o f the base unemployment benefit, and i s funded from the state budget. In 2004,24,300 job seekers received financial assistance. The low level o f benefits and the high uncertainty o f their payment mean that unemployment benefits are likely to provide negligible consumption smoothing. Their adverse work incentive effects are small, ifany. The low number of beneficiaries i s largely explained by the fact that most of the unemployed have exhausted their eligibility for unemployment benefits. In 2006, 71 percent o f the registered unemployed have been out o f work for more than 12 months, and another 15 percent, between 6 and 12 months, so they would no longer qualify for the benefit. Given that access to ALMPs i s also limited, a vast majority o f job seekers have no incentive to register with the SES. Vroman (2002) has developed a summary measure of the generosity of unemployment benefit programs. The two key determinants o f the cost o f unemployment protection are the effective benefit replacement rate (RRate) or benefit level, expressed as a fraction o f average wage, and the ratio o f benefit recipients and the number o f unemployed as identifiedby labor force surveys (NBedUnemp). The former factor reflects the relative value o fbenefits and the latter reflects the relative availability o f benefits, both factors being outcomes o f policy choices. The product o f the replacement rate and the share o f compensated unemployed captures more inclusively the generosity o f unemployment benefit programs. This product i s termed the generosity index (G): Generosity index (G) = 100x RRate x (NBedUnemp) The average generosity index across transition economies i s 11.3 compared to 26.3 for the OECD sample. The most significant exceptions are Hungary and Slovenia, the indexes o f generosity o f which have been comparable to the average o f the OECD sample. InArmenia, the generosity index in 2004 equaled 0.4, or almost 30 times less than the average in transition countries. The new L a w on Employment and Social Protection in Case of Unemployment adopted by the National Assembly in October 2005 proposes a new formula for the benefit replacement rate. Instead o f the average wage o f the beneficiary, it will be tied to the minimumwage, and sets the amount of the unemployment benefit at the level of 60 percent 107 o f the minimumwage (since January 1,2006, the minimumwage i s 15,000 drams). The Law anticipates the minimumduration o fbenefit payment as 6 months, with the possibility o f a 1- month extension for each additional three years o f insurance contribution payments, but not more than 12months intotal. In Armenia passive labor market policy measures are not considered a priority in providing a sufficient standard of living. The existing unemployment compensation system acts as an incentive to reintegrate quickly into the labor market, or for job seekers with low job motivation and no insurance record not to register at all. However, such a low level of benefit i s a serious problem for households in depressed regions, and for those families relyingon such benefits to maintaintheir standard o fliving. The majority of directors of local employment services (88 percent) think that the size of the unemployment benefit i s the main reason job seekers lack the willingness to register at employment services (Table 8.4). Considering the high cost o f passive policies and the budget limitations, the focus should be shifted to active policy measures that have lower unit costs, and the costs will decline further on par with an increase in the number o f beneficiaries. Table 8.4: Reasons for Unwillingness of Unemployedto Register at the Regional Employment Centers inthe Opinionof Directors Percent Ineffectiveness of Programs Implemented 9.8 Low Level of Unemployment Benefit (financial assistance) 88.2 Lack o f Information on Activities o f the Center 7.8 Distance o f the Center from Residential Areas 39.2 Discouragement o f Unemployed 43.1 Number of Required Documents (procedures) for Registration 7.8 Other 2.0 Source; Survey of Directorsof RegionalEmploymentCenters 2005. Directors of the regional centers of the State Employment Service were asked to mention three main factors, from a list of six factors, hindering their operations (giving them the possibility to also indicate other factors not included on the list). Three main factors were mentioned the most: the low level o f technical provisions (mentioned by 78 percent o f directors), low wages (75 percent), and inadequate means of communication (59 percent). Only 16 percent o f the respondents mentioned low skills o f the professional staff as a hindrance inthe provision o fmore effective services. International donor organizations have recently become more involved in Armenia on employment issues, focusing particularly on reforming the state regulations and operations of relevant institutions. Most recently, Sweden i s providing assistance to the Employment Service to train the staff, The Deutsche Gesellschaft fur Technische Zusammenarbeit (GTZ) has arranged training courses for more than 500 individuals in Tavush marz. EUTASIS i s also providing assistancein trainingjob seekers. The USAIDhas provided assistance to many other local employment offices to update their information 108 systems. Integrated social service centers (one-stop-shops) are being piloted with USAID help in Vanadzor. The U.N.Armenian Food Program "Food via Training" project i s being implemented, and around 10,000 job seekers in 37 regions o f Armenia will be involved in small business training courses. In 2004, the World Bank launched the Social Protection Administration Project with the objective o f establishing a modern information system inthe SES and assisting indeveloping contemporary employment services inthe country. Increasing allocations from the state budget in the medium term should be one of the key elements of the policy reform. Such a trend i s already notable inthe MTEF for 2005- 07 and 2006-08 MTEF approved by the government. 8.6. Social Dialogue In a market economy, collective bargaining can play an important role in determining wages (and other conditions of work). It i s well known that, all other things being equal, unions can raise wage pressures, and wages bargained collectively are generally higher and less dispersed than those bargained individually. Also, unionized workers and those covered by collective agreements enjoy shorter working hours and tend to receive more training than comparable nonunionized and uncovered workers (Aitd and Tzannatos 2002). During the Soviet era, virtually all workers belonged to trade unions. However, the function o f unions differed greatly fi-om what they typically do in market economies. Their major role was to provide social services, recreation and culture, housing, consumer goods and services in short supply, and sick pay. Duringthe transition years, their membership and the role o f trade unions has completely changed. The international trend (especially European) i s that, as economies move away from socialism and adopt a more liberal form o f government, the level of trade union membership drops dramatically. The weighted average for union density in 10 new EU Member States i s 22 percent, while the density in 15 "old" EUMember States is 30 percent (Eurostat Labor Force Survey 2002). Inmost CIS states, the union density rate i s still relatively high, for example, around 75 percent in Russia (World Bank 2002d). In Armenia, social dialogue at the national level is carried out by two major umbrella organizations of social partners. Workers and employees are represented by the Confederation o f Trade Unions o f Armenia, and employers are represented by the Union o f Manufacturers and Businessmen (Employers) o f Armenia. According to the Confederationo f Trade Unions, they have 469,000 members, or around 43 percent o f the total number of employed. According to the 2001 Law on Trade Unions and the 2004 Labor Code, inorder to serve as representative in collective labor relations, more than half o f the workers and employees o f the trade unions organization must be members. Currently, 24 branch Unions qualify for this status and are represented in the Confederation. Through its sectoral unions, the Confederation represents 4,338 organizations and firms o f which around half are in the state sector (state and public organizations and state-owned enterprises). The majority o f employees in public services sectors, including civil servants and pedagogical and medical personnel, are members o f their branch Trade Union. 109 The Union of Manufacturers and Businessmen (Employers) predominantly represents the private sector. Currently, the Union has around 500 member firms and organizations. Its main activities are directed at creating an auspicious business environment and protecting the interests o f the business community. The employers' side in Armenia i s definitely less organized, and most o f the entrepreneurs, especially in the nascent private sector, do not consider social dialogue as essential (Box 8.1). Box 8.1: Benefitsof Social Dialogue Social dialogue is one o f the principles underlying the European social model, based on good economic performance, a high level o f social protection and education, and social dialogue. In Europe, the concept o f dialogue between governments and employers' and workers' organizations is generally accepted as part o f good governance, even if its modalities and its extent may differ substantially from one country to another. Social dialogue is not in contradiction to the market economy. On the contrary, it can help sustain its effective hnctioning by dealing with its social aspects. It may prevent or solve unnecessary and violent social conflicts by achieving acceptable compromises between economic and social imperatives, and it may improve the business and investment environment. Social dialogue can help (a) ensure macroeconomic stability through responsible wage negotiations; (b) create balanced labor legislation, enabling employers to adapt to changing markets and thus maintain their competitiveness; (c) implement effective social security systems that protect workers but that do not create an unbearable burden for enterprises and other taxpayers; and (d) introduce cooperative industrial relations, which prevent and solve unnecessary and damaging industrial conflicts (Rychly and Pritzer 2003). As such, social dialogue can be an important instrument in controlling labor and social conflicts and improving productivity and competitiveness. According to the 2004 Labor Code, the General Collective Agreement should be negotiated and signed for three years, with the possibility of an extension for another three years. However, the national collective agreement has yet to be negotiated, and there are very few collective agreements signed at the lower levels.67Decentralized bargaining approaches in collective bargaining should be considered if a centralized approach i s not yielding efficient bargaining outcomes. In many countries in Europe and elsewhere, the majority o f collective agreements are at the sector or company level. Inparticular, except in Slovenia and Slovakia, in other new EU Member States, the dominant level especially o f wage bargainingi s the company (EIRO). 67The 2004 Labor Code anticipates national, sectoral, and territorial collective agreements and collective agreements concluded at the level o f organization or its separate (structural) subdivision. 110 CHAPTER IX: POLICY RECOMMENDATIONS Labor market interventions. The labor market study identified major distortions in the Armenian labor market and key constraints in promoting a legal and institutional environment conducive to job creation. This chapter looks at the scope for policy interventions in Armenia's labor markets to address some o f the crucial policy issues and structural rigidities driving unemployment and slowingjob creation.68Various policy options might be relevant and suitable for Armenia given its macroeconomic and labor market situation. Armenia needs its own Action Plan for Employment, similar to that in European Union (EU) member countries and EU accession countries, which sets up a coherent strategy for employment promotion and its priorities for action, and determines the means and policies to achieve the objectives o f the strategy. The policies should include measures on both the demand and supply side leadingto an increase inthe employment rate. Labor market interventions have a big potential to improve labor market performance. Whether or not this potential i s materialized, however, depends on a number o f factors, and innumerable variables intervene in the final outcome o f their implementation. These include external factors, such as stable macroeconomic conditions, a favorable investment climate and enabling business environment, and a competitive product market. Experience in EU countries confirms that macroeconomic, microeconomic, and employment policies go hand inhandfor deliveringmore andbetterjobs (EC 2005a). Also, institutional rigidities that impede flexibility inhiring and dismissal raise indirect labor costs and thus reduce employment. In addition, there are important internal factors: policy design and implementation. Thus, the positive impact o f labor market policies cannot be taken for granted, which points to the importance o f monitoring and evaluating their outcomes. The new role of employment services. The Government o f Armenia i s currently involved in the reform o f public employment services, with the support o f international donors, such as the World Bank, the United States Agency for International Development (USAID), and the Swedish International Development Cooperation Agency (SIDA). The new Law on Employment and Social Protection in Case o f Unemployment was adopted in October 2005. While employment services provided by the State Employment Service (SES) and private employment agencies play an important role in assisting job seekers in reentering the labor market, based on international experience, the general consensus i s that even under the best scenarios, payoffs to active labor market programs (ALMPs) are modest.69 It i s difficult to address problems o f large-scale unemployment through these programs. They may work for 68For best practices from OECD, EU, and transition countries see, for example, E C 2005a; OECD 2005; World Bank 2005b; and http://europa.eu.int/comm/employment_sociavindex-en.html. 69See the recent World Bank review o f ALMPs around the world, based on 152 scientific (that is, control- group) evaluations (Betcherman and others 2004). It should be noted that these reviews were based almost exclusively on the experience o f industrializedcountries, because very few evaluations exist anywhere else (see also Dar and Tzannatos 1999). 111 specific groups inspecific circumstances-hence, they need to be carefully targeted. Evenin EU countries with much larger budgets for ALMPs and years of experience inprovision of relevant programs, there i s still a highrate o f return into unemployment or inactivity among those who have been targeted by ALMPs. Lessons learned from this phenomenon include the need for early identification o fjob seekers' needs and the tailoring o f ALMPs to these needs. Currently, employment services in Armenia are very limited and the programs are under funded. The country must identify the priorities o f programs that can improve future employment and earnings prospects in a cost-effective manner. The most recent analysis o f the impact o f ALMPs in Organisation for Economic Co-operation and Development (OECD) countries suggests that an increase in the intensity o f spending on ALMPs (defined as the percentage o f GDP allocated to active policies divided by the unemployment rate) accounts for 10 to 20 percent of the total increase inthe employment rate observed during 1997-2002. The results o f an analysis using OECD's expenditure breakdown for ALMPs suggest that the expenditure category with the most significant and positive impact on the employment rate i s spending on public employment services and administration (that is, job search assistance) (EC 2005a). Taking into account the empirical evidence from OECD, Armenia should refocus all its efforts on activation policie~.~'Good activation policies involve: (i) improving personal, social, and vocational skills and competencies, and facilitating social integration; (ii) individually tailored ways to participate in ALMPs, taking into account such factors are the person's age, experience, needs and priorities; (iii) taking advantage o f the resources and strengths o f the beneficiary; (iv) networking with labor market services, social services, health services, housing sector, and communities; and (v) cooperation and interaction between the beneficiary and the agency in the planning, design, and implementation. For that, extensive capacity building o f the State Employment Service-currently understaffed, ill-equipped and poorly motivated-is required. Workers often lack information about and confidence in how to look for a job. Counseling can be crucial inhelpingindividual workers obtain information about education, training, and alternative job opportunities, in making other employment services more demand-oriented, and in better targeting expenditures for other employment services. This involves establishing and developing good relations with employers and the gradual transformation o f the SES into a genuine service enterprise (modernization o f the range o f services, creation of specific enterprise and sector desks, improvement o f its public image, use o f information technologies, and so forth).71 There are also a variety o f other strategies that increase the 70 Activation policies encourage certain unemployed individuals to step up their job search after an initial spell o f unemployment, with a later obligation to participate invarious programs. Eventually, the ctivation principle makes the receipt o f benefits conditional on participation in programs, in the process shifting the balance between the rights and obligations o f the unemployed. See EC 2006. 71The recent World Bank Social ProtectionAdministration Project (SPAP) inArmenia focuses on, among other subjects, the improvement o f the effectiveness (including quality, efficiency, user satisfaction, and financial viability) o f the public employment services. The services to be improved include job search skill training programs, employer contact services, job clubs and job and vacancy fairs, staff training o f the SES, and establishment o f a modern management information system. 112 chances for job seekers o f identifying job openings and breaking into the "hidden job market." These techniques include: networking, informational interviewing, direct employer contact, and employment services' electronic bulletin boards. Table 9.1: Impactsof ALMPsBasedon Experienceof EUCountries Intervention Summary of overall Comments impact Job-search Significant positive impact Improved outcomes when: (i) combined with monitoring assistance (job on the transition from and enforcement o f criteria, on which the receipt o f brokerage and welfare to work unemployment benefits is conditioned; and (ii) job counseling) search activities linked with participation inother programs, such as training. (Re)training for Effective for some target Following features enhance programs effectiveness: (i) unemployed groups (adult women), but small scale; (ii)targeted at disadvantaged groups; (iii) not for others (adult men close partnership with local employers; (iv) training and youth) certified, and certificates recognized and valued by the market. Youth measures Disappointing results Interventions such as pre-school facilities, measures to reduce early school-leaving, and improve basic slulls and the relevance of competencies provided by the education system seem to pay better dividends. Employment Positive effects on the Cost effective to target employment subsidies at specific subsidies probability o f future groups (young men with highlevels o f education), and unsubsidized employment combined with counseling services. Direct job creation Positive results rare Shouldbe short duration and targeted at the most disadvantaged groups, if used. Source: EC (2006). Another dimension in the provision o f ALMPs i s that a variety o f services are needed to address three conditions commonly experienced by displaced workers who are trying to reenter the labor market (see, for example, Fretwell 2004): Frictional Unemployment: Displaced workers have marketable skills for which there i s a demand, but they need intensive job placement assistance. These services can have a substantial positive impact on reemployment andhave low unit costs. Structural Unemployment: Displaced workers lack skills, or their skills are not in demand, and need retraining to compete and reenter the labor market. Different types o f retraining, including on-the-job and/or institutional training, are needed inthese circumstances, and can help inreemployment at a moderate unit cost per worker. Lack ofDemandfor Labor: This i s a particular problem in areas o f highunemployment and in single-enterprise communities where many workers are laid off. Programs include small business consulting assistance, business incubators, public works, and micro-loans. These programs tend to be more expensive than other services, and to attract a limited number o f participants, but they can be quite effective. 113 Entrepreneurshipassistance.The promotion of self-employment for the unemployedor for workers threatened with redundancy could be one o f the objectives o f the SES and other government agencies. Support for self-employment i s usually an effective measure helping the unemployed with entrepreneurial abilities (which is admittedly a relatively narrow group) to start their own business and escape unemployment. Special emphasis should be given to entrepreneurship skills training in order to encourage youth and other target groups to start their own income-generating projects. On the negative side, very few among the unemployed typically take advantage o f opportunities for self-employment. In both market and transitional economies, well-designed self-employment programs will reach up to 5 percent o f the unemployed (Wilson and Adams 1994). Those who participate will be primarilymale, better educated, and intheir 30s. The programs tend not to attract women. Relevant programs can involve providing financial and advisory support for start-up, or supporting the operating costs o f small businesses. Small business advisory services usually provide assistance in business planning, including bookkeeping, financial, legal, marketing, and sales services; assistance incontacting local authorities; short-term training (for example, one to two weeks); and other consulting services to unemployed entrepreneurs who intend to start or who have started businesses during the past six months. Often there is "screening" whereby potential beneficiaries undergo an assessment that evaluates their likelihood of success. Public programs to support small business loans can contribute to the removal o f distortions arising from credit rationing. In many transition countries, self-employment programs are supported or are basically micro-credit programs. As an example, the Lithuanian Labor Exchange grants interest-free loans for a three-year period to unemployed who are going to organize their own business after completion o f compulsory business organization courses, and who have property eligible to beprovided as collateral, or a warrantor (g~arantor).~~ Youth employment programs.One ofthe critical tasks for labor marketpoliciesinArmenia i s to enhance the employability and to increase the employment rates o f the youth currently unemployed or inactive. Internationalpractice confirms that a combination o f preventive and curative policies i s needed to try to prevent the emergence o f youth employment problems and to deal with those that, nevertheless, do emerge. The OECD Employment Outlook for 2002 pointed out that "the biggest payoff for disadvantaged youths comes fi-om early and sustained interventions" (OECD 2002). This i s likely to be a more effective long-term policy than attempting to cure youth employment problems after they have emerged. Policies should emphasize basic education and basic skills and promote innovative approaches for improving the employment situation o f young people in the informal sector. Among preventive policies, in addition to achieving a healthy rate o f increase in the overall demand for labor, or relaxinglabor regulations (especially setting minimumwages at a lower level for young people or eliminating it altogether), improvements in the national education and training system, and especially improving functional literacy among 15-24-year-olds (for example, more broadly defined literacy skills), seem to be key components to reducing l2See: http://europa.eu.int/comm/employment~social/emplo~ent~strategylO4~national~en.htm. 114 unemployment among youth (Godfrey 2003). In many European countries, apprentice contracts have also proven to be an effective tool in making youth labor attractive to employers. As part o f the World Bank SPAP project, a Youth Professional Orientation Center is being opened in Yerevan to provide specialized job counseling, job placement, and professional orientation services to unemployed youth. Relevant career counseling services should be made commonly available to all students in general, vocational, and higher educational establishments, including free-of-charge vocational guidance; career counseling services should be tailored to specific target groups; access should be provided to information about various opportunities ineducation and continued education; assistance should be provided in human resource management, recruitment, retraining; and psychological support should be provided. Improving the efficiency of training. In addition to vocational education and training (VET) programs provided by specialized education establishments, training is one o f the most important employment services for increasing the competitiveness o f the registered unemployed in the overall labor market and improving their chances o f reemployment. Relevant programs offered by the Public Employment Services (PES) include vocational training and retraining, and courses on how to adjust to the ever-changing labor market through learning about how to communicate, usingthe current "language" o f the workplace, and so forth. The activities o f the PES ingeneral are often evaluated on the basis of coverage of the unemployed by these courses, and on the cost-benefit effects o f programs. However, evidence suggests that it i s difficult to overcome previous education problems with short- duration training programs. Programs may need to be better aligned with the profile and needs o f the unemployed. A mismatch between the supply and demand for different trades is probably the most common problem o f vocational training around the world, and Armenia i s no exception. It usually results from a lack o f communication between enterprises and schools, and from the absence o f regular flows o f information on the skill requirements inthe workplace and skill shortages and mismatches inthe labor market (Fretwell and others 2001). It is therefore a priority to carefully assess curriculums in connection with an assessment o f likely labor demand, and to link the training with the skill needs o f the private sector. Links to the private sector could be encouraged through greater use o f on-the-job training programs targeted to specific disadvantaged groups among the unemployed, such as youth or women. Inorder to move from a supply-drivento a demand-driven VET model, the overall state of the labor market should be thoroughly analyzed, including new forms o f employment, underemployment, and unemployment, components o f labor force growth, age structure, industry and occupation structure, labor market segregation (for example, by age or gender), and regional imbalances. These allow not only the identification o f the general trends in the labor market, but also the making o f projections on labor supply and demand. 115 In addition to good general labor market information, there exist several other statistical instruments, including employer-based surveys of current and projected labor market condition^.^^ Such surveys are focusing on actual and planned job creation and job destruction, and on key determinants o f hiring and firing. The objective o f such surveys would be to determine the degree of labor market flexibility, and to prepare projections on changes in employment and unemployment. A tracer survey o f displaced workers would trace changes in labor market status (earnings, employment compared to unemployment, career developments), depending on the educational status o f workers or unemployed individuals. Another approach would be to keep track o f graduates fi-om some years after graduation, as part o f labor market monitoring. As far as adult training i s concerned, there is a need for the development of the principles of and the demand for lifelong learning, which ensures the continuum between initial and continuing training as an essential precondition for rapid economic growth. InArmenia, this concept has not been widely accepted among the population yet. Many transition countries have established tax incentives to encourage adult education. In Lithuania, the Law on Income Tax o f Individuals envisions the possibility o freceivingpartial compensation o f payment for studies. Upon submission o f the annual declaration o f income tax, amounts paid for studies or part o fpaid income tax may be compensated. InHungary, employers pay avocational training contributionequivalent to 1.5 percent of the wage. Employers that spend the same amount on apprenticeships, the training o f their own employees, or on the development o f a vocational training school are exempted fiom this contribution. In-company training, training related to investments creating new jobs, and training aimed at improving the competitiveness o f small and medium-size enterprises and the development o f entrepreneurial skills are supported by grant schemes.74 Older and disabled workers. The Government of Armenia has already taken steps to encourage the employment o f older workers. The retirement age has gradually been extended and the conditions for early retirement have been made more stringent, which should be reflected in an increase in the employment rate among older citizens. Some transition countries have built-inincentives in the pension system to encourage longer job tenure. In Estonia, pensions are increased by 0.9 percent for every month that retirement i s delayed. Thus, it is possible to receive an 11percent higherpensionthrough a one-year delay. Older workers receive less training than workers inother age groups. It i s essential to reverse this trend because the productivity potential o f older workers is not lowered by age, but by obsolete skills. In general, only lifelong learning will provide future older workers with the necessary skills to adapt to changes in the employment market. Other measures include vocationalrehabilitation, adaptation o f working places, further training, retraining, and active 73 Such surveys have already been conducted in Armenia with help fromSIDA, and are plannedas part of the World Bank SPAP project. '4 See: http:lleuropa.eu.intco~employment~socia~employment~strategylO4~national~en.htm. 116 employment services for older workers or workers with disabilities whose vocational abilities have been damaged. Some transition countries have established special employment protection for persons with disabilities. In Hungary, for businesses with more than 20 employees, the law prescribes a mandatory employment quota o f persons with disabilities. If this quota i s not met, the employer must pay a rehabilitation contribution. Those who employ disabled workers above the quota may make use o f wage subsidies. Businesses with a high ratio o f disabled employees might receive special subsidies. The labor market integration o f people with disabilities i s promoted by rehabilitation groups in the labor centers. In addition, comprehensive services are available in the so-called Rehabilitation Information Centers (RIC) in 16 county labor centers. As part o f the World Bank SPAP project, a Job Rehabilitation Center i s scheduled to be opened in Gyumri City with the objective o f providing job rehabilitation, advisory, professional orientation, training, and job placement services for the disabled. Guymri i s located inthe regionmost affected by the 1988 earthquake. Tackling long-termunemployment.Highunemployment inArmenia i s compounded by its long duration. Many o f these jobless have no qualification, and/or a low level o f education. They may have multiple employment barriers, including cognitive and health-related barriers, and difficult home lives (for example, lack o f transportation, many children, child care problems, domestic violence), which makes their employability a problem. To tackle long-term unemployment, a policy mix combining training opportunities with active counseling and information on job opportunities has proved quite effective in a number of European countries. However, the experience o f some transition countries shows that such measures are often insufficient. The long-term unemployed should resort to a combination o f temporary employment (public works or subsidized employment), on-the-job training, and regular job-placement assistance (see Egger 2003). To prevent inflow into long-term unemployment and to retain employability, especially young unemployed must be provided with effectivejob search assistance or active labor market measures within the first 6 months o f unemployment, while adult unemployed should be provided with the above services within the first 12 months o f unemployment. Early interventions include profiling o f job seekers to identify which one o f the individuals or groups o f unemployed are susceptible to long-term unemployment. The programs for relevant vulnerable groups typically include, among other things, the provision o f employability and training plans, job and career counseling services, various aptitude tests, and vocational assessmenttests. Addressing regional disparities. Like many other transition countries, in Armenia the disparities among regional labor markets caused by their uneven development potential are widening. There i s an increasingly important role for regional development strategies and the involvement o fkey actors inpolicy on employment and education. Many European countries have established regional economic development planning services, which are intended to help communities develop the local economy and generate new employment and investment by identifying and marketing regional resources, potential 117 areas o f development, and opportunities to potential investors. (See EC 2004a). Such services could help to ensure the links between training provision and the development plans o f the local economy through the development o f local partnerships. Local economic potential studies have been initiated, as kind o f local Action Plans for Employment, to identify labor market resources available, to promote and market local products and services, develop promotional materials on local economic potential and investment projects, and so on. These activities mightbe supported by special programs to support disadvantaged regions. In Latvia, the Regional Fund provides support and facilitates entrepreneurial activities in territories approved by the Cabinet o f Ministers. The largest part o f assistance granted consists o f interest rate payments or partial interest rate payments for businesses, and co- financing for business training projects and municipalities developing their business support infrastructure(including fundingfor setting up tourism and business support centers). It is important that vocationaltraining inmarzes is being adjusted to the needs ofthe regional and local labor market. The training programs in VET should be focused on the key sectors o f the region and implemented through the cooperation o f all actors concerned. Labor markets are shifting largely according to local demand, and based on specially designed regional development plans, the local VET institutions may have a better perspective on what direction the local economy i s movingin. Involving social partners. Social partners and other partnership groups representing organizations from different sectors can play an important role as a means o f identifying local solutions to the challenges o f economic and social transformations. It i s especially important for local offices o f the SES to develop partnerships with other public or private bodies in order to introduce localjob-creation strategies to meet the new needs. Involvement o f the social partners can be beneficial in order to achieve broad participation in preparing legislation, developing training content, and monitoring the activities of the SES and VET institutions to make them more responsive to new developments. Involvement o f employers' organizations would be an important instrument for minimizing skills mismatch in the labor market. First, occupational standards, or employment specifications, must be defined by employers following procedures agreed upon by all stakeholders. Second, training standards, or learning specifications, could be used to define curriculums in training institutions. The content must be based on occupational standards (Fretwell and others 2002). The list (nomenclature) o f vocational profiles should be reviewed and new occupational standards developed for them through consultation between economic actors, including the private sector, taking into consideration the current and future needs o f the economy. Untilrecently, social partners were dissociated from involvement inthe vocational education system because they rarely participated in the decision making process (for example, the definition o f skills needed by the enterprises, the vocational profiles to be developed, and the qualifications to be provided). In 2004, the Ministry o f Education adopted a VET Reform Strategy that identified the key problems and major targets o f the reform: educational continuity, skills provision according to personal capacity and to economic and labor market 118 needs, involvement o f social partners, and lifelong learning. Following the Strategy, in July 2005 a new law on vocational education and training was adopted, which improves the position of social partners. They are now recognized as full partners in the VET system. Nevertheless, at this stage no changes to the qualifications system are anticipated. Capacity buildingof decisionmakers and stakeholders inthe area of VET and labor market programs ingeneral canbeidentifiedas one ofthe policydirections inArmenia. 119 REFERENCES Adams, Richard (2003). InternationalMigrations, Remittances and the Brain Drain: A Study o f 24 Labor-ExportingCountries. World Bank Policy Research Working Paper N o 3069 (June). Washington, DC. Adams, Richard and John Page (2003). InternationalMigration, Remittances and Poverty in Developing Countries. World Bank Policy Research Working Paper 3179. Washington, DC. AST (Advanced Social Technologies 2005). Labor Migration from Armenia in2002-2005. Presented to OSCE Office inYerevan. Yerevan. Avag Solutions Ltd. (2005a). Analysis of the Armenian Labor Markets and Youth Employment Issues. Part A, B, and C. Mimeo. Yerevan. Avag Solutions Ltd. (2005b). Rural Labor Markets. Mimeo. Yerevan. Betcherman, Gordon, Amit Dar, Amy N.Luinstra and Makoto Ogawa (2000). Active Labor Market Programs: Policy Issues for East Asia. Social ProtectionDiscussion Paper no. 0005, January. World Bank. Washington, DC. Betcherman, G., A. Luinstra, and M.Ogawa (2001). "Labor Market Regulation: International Experience inPromoting Employment and Social Protection." Unpublishedpaper. Social ProtectionUnit, World Bank. Washington, DC. Betcherman, G., Olivas, K., and A. Dar (2004). Impacts o fActive Labor Market Programs: New Evidence from Evaluationswith Particular Attention to Developing and Transition Counties. Social ProtectionDiscussionPaper Series, N o 0402. Washington, DC. Cazes, Sandrine and Alena Nesporova (2003). Labour markets intransition: balancing flexibility and security inCentral and Eastern Europe. ILO. Geneva. CIS STAT (1994). Rynok truda v stranakh sodruzhestva v tsifrakh idiagrammakh (241 (66)). (InRussian). Moscow. CIS STAT (2005). Commonwealtho f Independent States in2004. Statistical Yearbook. Moscow. Dar, Amit and Indermit S. Gill (1998). Evaluating RetrainingPrograms inOECD Countries: Lessons Learned. World Bank Research Observer 13, No. 1. Dar, Amit and P. Zafiris Tzannatos (1999). Active Labor Market Programs: A Review o f the Evidence from Evaluations. Social ProtectionDiscussionPaper No. 9901, January. World Bank, Washington, DC. 120 DharamGhai (2003). Decent work: Concept and indicators. International Labour Review, Vol. 142 (2003), No. 2. Dijk,M.P. (1995). The Internationalizationofthe Labour Market. In:GlobalEmployment. An InternationalInvestigationinto the FutureofWork. Vol. 1. Ed.byM.Simai with the assistance o f V. Moghadam and A. Kuddo. UNU/WIDER.Zed Books Ltd. London, New Yersey and Tokyo. EBRD(European Bank for Reconstruction and Development2000). Transition Report UpdateMay 2000. London. EBRD-World Bank (2005). Business Environment and Enterprise Performance Survey. Washington, DC. EC (European Commission2003a). Employment inEurope2003. Recent Trends and Prospects. Employment & European Social Fund. Brussels. EC (European Commission2003b). Jobs, Jobs, Jobs: Creatingmore employment inEurope. Report o f the Employment taskforce chairedby Wim Kok. Brussels, November. EC (European Commission2004a). Practical Handbook on DevelopingLocal Employment Strategies inthe New Member States and Candidate Countries o f the European Union. Employment & European Social Fund.Brussels. EC (European Commission2004b). Undeclared work inan enlarged Union: An Analysis o f Undeclared Work: An In-DepthStudy of Specific Items. Brussels. EC (European Commission2005a). Employment inEurope 2005. Employment & European Social Fund.Brussels. EC (European Commission2005b). The evolution o f labour law (1992-2003). Volume 1-11. Employment & Social Affairs. Brussels. EC (European Commission2006). Employment inEurope2006. Employment & European Social Fund. Brussels. Egger, Philippe (2003). Decent work and competitiveness: Labour dimensions o f accession to the EuropeanUnion. International Labour Review. Vol. 142, No. 1. Freinkman, Lev, Gohar Gyulumyan, and Artak Kyurumyan (2003). Quasi Fiscal Activities, HiddenGovernment Subsidies, and FiscalAdjustment inArmenia. World Bank. Washington, DC. 121 Fretwell, DavidH., Jacob Benus, and Chstopher J. O'Leary (1999). Evaluatingthe Impact o fActive Labor Market Programs: Results o f Cross Country Studies inEurope and Central Asia. Social ProtectionDiscussionPaper No. 9915. World Bank,Washington, DC. Fretwell, David and Sandra Wilson (1999) Public Service Employment: A Review o f Programs inSelected OECD Countries and Transition Economies, World Bank Discussion Paper, (June), No. 9913. Washington, DC. Fretwell, D.,J. Benus, and C. J. O'Leary (1999). Evaluatingthe Impact o f Active Labor Programs: Results o f Cross Country Studies inEurope and Central Asia. Social Protection DiscussionPaper No. 9915. The World Bank. Washington, DC. Fretwell, D., and A. Wheeler. (2000). Secondary education intransition countries: Rethinking the framework. World Bank, Washington, DC. Fretwell, D., Lewis, M.and A. Deij. (2001). A Framework for Defining and Assessing Occupational and Training Standards inDeveloping Countries. Information Series No. 386. World Bank. Washington DC; Ohio State University, Columbus, Ohio; European Training Foundation, Turin. Fretwell, David (2004). Mitigatingthe Social Impact o f Privatizationand Enterprise restructuring. World Bank, SP DiscussionPaper, No. 0405. Washington, DC. Godfiey, Martin (2003). Youth Employment Policy inDeveloping and Transition Countries -PreventionaswellasCure.SPDiscussionPaperNo.0320.WorldBank, Washington, DC. GOSKOMSTAT SSSR (1991a). Itogi Vsesojuznoi perepisinaseleniya 1989 goda, Vol. VIII, Moscow. GOSKOMSTAT SSSR (1991b). Itogivsesojuznoi perepisi naselenija 1989 goda, Vol. XI, Moscow. GOSKOMSTATSSSR (1991~).Natsionalnyi sostav naselenija SSSR. Moscow Haltiwanger, John, Scarpetta, Stefan0 and Milan Vodopivec (2003). How institutions affect labor market outcomes: evidence from transition countries. Mimeo. World Bank, Washington, DC. Heron, R., and C. Vandenabeele (1999). Labour Dispute Resolution: An Introductory Guide. ILO Regional Office for Asia and the Pacific. I L O East Asia Multidisciplinary Advisory Team, Bangkok. IFC-World Bank (2003). Armenia: Administrative Barriers to Investment -Update. Foreign Investment Advisory Service. Washington, DC. 122 ILO (2003). Summary of InternationalPublic Employment Services (PES) Administrative Statistics (updated September 2003). Geneva. ILO (2004a). Global Employment Trends. Geneva. ILO (2004b). Global EmploymentTrends for Youth. August. Geneva. ILO (2004~).World EmploymentReport 2004-2005. Geneva. IMF (1999). Armenia: Recent EconomicDevelopments and Selected Issues. ReportNo. 99/128. November. Washington, DC. IMF(2005). FirstReview Underthe Three-Year Arrangement Underthe PovertyReduction and Growth Facility. IMF Country Report No. 05/422. Washington, DC. I O M (International Organization for Migration 1998).Migration Potential inCentral and Eastern Europe. Geneva. Iradian, Garbis (2003). Armenia: The Roadto Sustained Rapid Growth, Cross-Country Evidence. IMF Working paper, WP/03/1003. IMF, Washington, DC. Javorcik, Beata, and Mariana Spatareanu (2004). D o Foreign Investors Care About Labor Market Regulations? Policy Research Working Paper No. 3275. World Bank, April. Washington, DC. Kuddo, Arvo (1995). EmergingLabour Markets. Labour Market Developmentsand Transitional Unemployment inCentral and Eastern European Countries. World Development Studies 2. UNU/WIDER. Helsinki. Kuddo, Arvo (2004). Public Works inEurope and Central Asia. In:ECA Spectrum. World Bank. Washington, DC. Martin, John P. (2000). What Works among Active Labour Market Policies: Evidence from OECD Countries' Experiences," OECD Economic Studies, No. 30. Paris. Martin, John P., and David Grubb (2001). What Works and for Whom: A Review o f OECD Countries' Experiences with Active Labour Market Policies. Office of Labour Market Policy Evaluation, Working Paper 2001:14. Paris. MED(MinistryofEconomic DevelopmentofGeorgia2005). LabourMarket inGeorgia 2002-2004. Tbilisi. Minasyan, Anna and Blanka Hancilova (2005). Labor Migration from Armenia in2002- 2005: A Sociological Survey o f Households. OSCE and Advanced Social Technologies NGO.Yerevan. 123 MS(MinistryofStatistics ofArmenia 1997). LaborMarket 1990-1 996, inFigures and Charts. State Register and Analysis o f the Republic ofArmenia. Yerevan. NSS (National Statistical Service of the Republic of Armenia 2001). Statistical Yearbook of Armenia 2001. Yerevan. NSS (National Statistical Service o f the Republic o f Armenia 2002). Labour Force inthe Urban Settlements of RA. Labour Force Sample Survey. August. Yerevan. NSS (National Statistical Service o f the Republic o fArmenia 2003).The Results o f the 2001 Census o f the Republic o f Armenia (Figures of the Republic of Armenia). Yerevan. NSS (National Statistical Service o f the Republic o f Armenia 2005a). Armenian Time Use Pilot Survey.July 1-312004. Yerevan. NSS (National Statistical Service o f the Republic o f Armenia 2005b). Labour force and child labor inArmenia. Yerevan. NSS (National Statistical Service o f the Republic of Armenia 2005~).Statistical Yearbook o f Armenia 2005. Yerevan. NSS (National Statistical Service of the Republic o f Armenia 2005d). The Armenian Migration Survey 2001/2002. Mimeo. Yerevan. N S S (National Statistical Service of the Republic o fArmenia 2005e). Women and Men in Armenia. Statistical Booklet. Yerevan. NSS (National Statistical Service o f the Republic o f Armenia 2006). Statistical Yearbook of Armenia 2006. Yerevan. OECD (1998). Employment Outlook 1998. Paris. OECD (1999). Implementing the OECD Jobs Strategy: Assessing Performance and Policy. Paris. OECD (2001). Employment Outlook 2001. Paris. OECD (2002). Employment Outlook 2002. Paris. OECD (2003). Combating Child Labour. A Review o f Policies. Paris. OECD (2004a). Education at a Glance: OECD Indicators. Paris. OECD (2004b). Employment Outlook 2004. Paris. OECD (2005). Employment Outlook 2005. Paris. 124 O'Higgins, Niall (2003). Trends inthe Youth Labour Market inDeveloping and Transition Countries. World Bank, SP DiscussionPaper No. 0321, Washington, DC. O'Leary, Christopher, Alena Nesporovaand Alexander Samorodov (2001). Manual on Evaluationo f Labour Market Policies inTransition Economies. InternationalLabour Office, Geneva. Oxenstierna, S. (1990). From Labour Shortage to Unemployment? The Soviet Labour Market inthe 1980s. Swedish Institutefor Social Research. Gotab. Stockholm. Rychly, Ludek, and Rainer Pritzer (2003). Social Dialogue at National Level inthe EU Accession Countries. Working Paper No. 12. ILO, Geneva. SC (Statistical Committee o f Azerbaijan Republic 2001). The population census o fthe Azerbaijan Republic 1999. Baku. Standing, G. (1988). Unemployment and Labour Market Flexibility: Sweden. ILO, Geneva. UNDP (2002). Education, Poverty, and Economic Activity inArmenia. Yerevan. UNICEF (1995). Poverty, Children and Policy: Responses for a Brighter Future. Regional Monitoring Report No. 3. Florence. USAID/Armenia (2004). Remittances inArmenia: Size, Impacts, and Measures to Enhance their Contribution to Development. Yerevan. Vodopivec, Milan and Raju, Dhushyanth(2001). Income Support Systems For The Unemployed: IssuesAnd Options. Social ProtectionDepartment, Human Development Network (June). World Bank. Washington, DC. Vroman, Wayne (2002). Unemployment Insurance and Unemployment Assistance: A Comparison. Social ProtectionDiscussionPaper, No. 0203. World Bank. Washington, DC. Wilson, Sandra and Arvil V. Adams (1994). Promotiono f Self-Employment for the Unemployed: Experience inOECD and Transitional Economies. World Bank, Washington, DC. Wilson, Sandra and David Fretwell (1999). Public Service Employment: A Review o f Programs inSelected OECD Countries and Transition Economies, World Bank Discussion Paper, No 9913. June. World Bank (I 995a). Armenia: The Challenge o f Reforminthe Agricultural Sector. Washington, DC. World Bank (1995b). World Tables 1995. Washington, DC. 125 World Bank (1999). Improving Social Assistance inArmenia. Report No. 19385. AM. Washington, DC. World Bank (2002a). Armenia: PovertyUpdate. December. Washington DC. World Bank (2002b). Growth challenges and government policies inArmenia, Volume 1. February. Washington, DC. World Bank (2002~).Impact Evaluation: Techniques for EvaluatingActive Labor Market Programs. Employment Policy Primer SeriesNo. 2. Social ProtectionUnit.Washington, DC. World Bank (2002d). Labor Market inRussia: Movingfrom Crisis to Recovery.. Washington, DC. World Bank (2002e). 10Years o f Transition. Washington, DC. World Bank (2003a). Armenia: PovertyAssessment. Washington, DC. World Bank (2003b). 2003 World Development Indicators. Washington, DC. World Bank (2004a). Armenia 1990-2003: Growing Poorer or Richer?ECA Region & PREM-EP. December 23. Washington DC. World Bank (2004b). Rural InfrastructureinArmenia: Addressing Gaps inService Delivery. Washington, DC. World Bank (2004b). World Development Indicators. Washington ,DC. World Bank (2005a). DoingBusiness 2006. Washington, DC. World Bank (2005b). EnhancingJob Opportunities: Eastern Europe and the Former Soviet Union/ Jan J. Rutkowski, Stefan0 Scarpetta. Washington, DC. World Bank (2005~).EU-8 Quarterly Economic Report. April 2005. Part 11. Labor Taxes and Employment inthe EU8. Washington, DC. World Bank (2005d). Ukraine Jobs Study: FosteringProductivity and Job Creation (InTwo Volumes). Washington, DC. World Bank (2005e). The Roar o f the Caucasian Tiger: Policies to Sustain Growth in Armenia. Report No. 32770-AM. June 27. Washington, DC. World Bank Investment Climate Survey (2005e). In: http://rru.worldbank.org/InvestmentClimate/. 126 World Bank andNSS (2006), Armenia: Growth, Poverty and Labor Markets 1998-2004 (forthcoming). Washington, DC. World Bank (2006). DoingBusiness 2007. Washington, DC. Young, David (2003). Employment ProtectionLegislation: Its Economic Impact andthe Case for Reform. EuropeanCommission, Economic PapersNo. 186. Brussels, July. Yeganyan, Ruben, andNelson Shahnazaryan(2004). Labor Migration from Armenia (An Overview of Literature). OSCE and IOM. Yerevan. 127 ANNEXES Figure 1: Share of Main Sectors inTotal Employment, Percent 50 0 , 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Industry -Agriculture andforestry +Construction - - - . Services and other sectors Source:National Statistical Service and authors' calculations. Figure 2: Employment Dynamics by Sector in 1994-2005 (1990 = 100) 80 250 70 60 200 50 150 40 30 100 20 50 10 0 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Non-agricuilure(left scale) Industry (left scale) - Construction(left scale) Transport and comnications (left scale) Total (right scale) -Agricuiture (right scale) 1Trade (right scale) wee: National Statistical Service and authors' calculations. 128 Figure 3: Jobs inthe State and Private Sectors of the Non-agricultural Productive Sphere in 1998-2005 500 450 - 400 - 350 - - - ~ 300 -345.0 - __ - __ - ~ ~ 1998 2000 2001 2002 2003 2004 2005 Jobs in productive state sector Jobs in productive private sector )urce: National Statistical Service and authors' calculations Figure 4: Labor Indicators inIndustry (1990 = 100) - 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 r jValueaddedinindustry Jobs in industry h Labor productivity Unitlabor cost iource: National Statistical Service and authors' calculations. 129 Figure 5: Labor Indicators in Construction (1990 = 100) 900 800 eA 4 700 # 600 - .* 500 * # 200 100 0 - 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 czzzi Value addedinconstruction Job inconstruction I Laborproductivity Unitlabor cost ~ ~~ Source: National Statistical Service and authors' calculations. Figure 6: Labor Indicators in Trade and Public Catering (1990 = 100) 150 125 100 75 50 25 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Value added in trade Jobs in trade = L Labor productivity Unit labor cost Source: National Statistical Service and authors' calculations. 130 Figure 7: Total Fertility Rate inArmenia, 1985-200475 Source: NSS (2005e). Figure 8: Labor ForceParticipation, Employmentand Unemployment Rates by Age, Percent, 2004 LFS Data Yource:NES (2005b). 75The average number o f children that would be born alive to a woman during her lifetime if she were to pass through her childbearing years (ages 15-49) conforming to age-specific fertility rates o f given years. 131 Figure 9: UnitLabor Cost Dynamicsby Main Branches in 1990-2003 0.60 0.50 0.40 0.30 0.20 0.10 0.00 1990 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Construction Trade Transport and cornnuncations - - Total -Agriculture Source: National Statistical Service and authors' calculations. Figure 10: Ratio of Annual Wages-to-GDP Per Capita in 1990-2003 I 200 _- 140 1990 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Industry (left scale) Construction(left scale) Trade (left scale) Transport and comnications (left scale) = Total (right scale) -Agriculture (right scale) *INon-agriculture(right scale) Source: National Statistical Service and authors' calculations 132 Figure 11: PublicEmployment Service Staff: InternationalComparisons (2002,2003) 337 ~ nia and I-tezegovina 815 I 721 700 500 300 100 100 300 500 700 900 Note: for Armenia the number o f PES staff dealing directly with job-seekers and employers i s calculated as a di erence between total number of PES staff and sum of the-numbers o f staff at the State Employment Service central office, and assistant(secretary) and support staff positions at the Regional Employment Centers. Source: State Employment Service; ILO Summary o f International Public Employment Services (PES) Administrative Statistics (Updated September 2003), ILO on-line database: Figure 12: Annual PublicExpenditureon Labor Market Programs Per Officially Registered Unemployed: InternationalComparisons (2002,2003; U S dollars) France Netherlands Portugal Austria Ireland Finland Denmark Spain I 417 Germany I Slovakia 624 Sweden 26,905 0 200 400 600 800 0 5,000 10,000 15,000 20,000 25,000 30,000 Source: State Employment Service; ILO Summary o f International Public Employment Services (PES) Administrative Statistics (Updated September 2003), ILO on-line database: www.ilo.org. 133 3 9 3 N A x d 3 3 ? > n 3 3 3 d m > > > N 3 3 2 x > 5 Table 2: MainLabor and Employment Indicators inIndustry in 1990-2003 (Thousand Persons) 2003 2002 2001 2000 1999 1998 1997 1996 1995 1990 Total labor 138.8 143.1 169.6 179.7 195.2 209.4 228.9 255.0 302.9 498.0 Industrialpersonnel 83.5 94.7 109.1 124.1 143.8 160 173.3 202.6 230.6 398.0 Personnel on 5.0 15.6 36.6 40.9 64.4 81.6 72.7 101.2 66.0 administrative leave Actual employment in 78.5 79.1 72.5 83.2 79.4 78.4 100.6 101.4 164.6 reporting enterprises Job balance, total* -4.3 -26.5 -10.1 -15.5 -14.2 -19.5 -26.1 -47.9 -52.3 Job balance inreporting -11.2 -14.4 -15.0 -19.7 -16.2 -13.3 -29.3 -28 -33.1 enterprises* ofwhich: `tforma1"jobs -10.6 -21.0 -4.30 -23.5 -17.2 8.9 -28.5 35.20 6.20 Actual employment -0.60 6.60 -10.7 3.80 1.00 -22.2 -0.80 -63.2 -39.30 change to the previous year Job balance innon- 6.9 -12.1 4.9 4.2 2.0 -6.2 3.2 -19.9 -19.2 reporting enterprises* Labor inmining and 121.2 124.0 151.6 161.9 176.4 189.1 210.8 242.2 292.5 490.4 manufacturing Industrial personnel in 68.60 78.80 91.20 106.3 125.0 139.7 155.2 189.8 220.2 390.4 mining and manufacturing Personnel in 7.64 18.71 36.54 40.90 64.40 81.60 72.70 101.2 66.00 administrative leave Actual employment 60.96 60.10 54.66 65.40 60.60 58.10 82.50 88.60 154.2 0 Job balance, total* -2.86 -27.6 -10.2 -14.5 -12.7 -21.7 -31.4 -50.3 -54.4 Job balance inreporting -10.2 -12.4 -15.1 -18.7 -14.7 -15.5 -34.6 -30.4 -35.20 enterprises* ofwhich: `fforma1"jobs -11.0 -17.8 -4.36 -23.5 -17.2 8.9 -28.5 35.2 83.35 Actual employment 0.87 5.44 -10.7 4.80 2.50 -24.4 -6.10 -65.6 -118.5 change to the previous year Job balance innon- 7.34 -15.2 4.88 4.20 2.00 -6.20 3.20 -19.9 -19.20 reporting enterprises* *Sign "-"indicatesjob cuts, and sign "+" -job increases compared to the previous year. Source: National Statistical Service. 135 Table3: MainLaborandEmploymentIndicatorsinConstructionin1990-2003 (Thousand Persons) 2003 2002 2001 2000 1999 1998 1997 1996 1995 1990 Total labor 37.2 36.1 41.2 46.5 53.6 56.7 59.7 68 76 184.1 Personnel 17.7 18.4 19.2 22.7 26.3 26 34.2 36.8 46.1 137 Personnel in administrative 0.8 1.7 3.2 4.8 5.1 5.7 leave Actual employment in 16.9 16.7 16 17.9 21.2 20.3 reporting enterprises Job balance, total* 1.1 -5.1 -5.3 -7.1 -3.1 -3 -8.3 -8 -20.8 Job balance inreporting -0.7 -0.8 -3.5 -3.6 0.3 -8.2 -2.6 -9.3 -12 enterprises* of which `Iformal" jobs -0.9 -1.5 -1.6 -0.3 -0.6 Actual employment change 0.2 0.7 -1.9 -3.3 0.9 to the previous year Job balance innon-reporting 1.8 -4.3 -1.8 -3.5 -3.4 5.2 -5.7 1.3 -8.8 enterprises* *Sign "-"indicates job cuts, and sign "+" -job increases compared to the previous year Source:National Statistical Service. 136 Table 4: Number and Size of Enterprises inIndustryin 1990-2003 2003 2002 2001 2000 1999 1998 1997 1996 1995 1990 Total number o f registered enterprises 9354 9077 8677 8234 7887 7394 6504 5216 2744 operating inindustry (annual average) % of total number of registered 19.44 19.46 19.24 18.98 18.8 18.74 19.27 20.43 20.84 enterprises Average size o f registered enterprises, 14.84 15.77 19.55 21.82 24.75 28.32 35.20 48.89 110.4 industrial personnelper enterprise Number o f enterprises reporting to the 2053 1815 1721 1234 1177 1050 786 786 773 767 N SS Average size ofreporting enterprises 40.77 52.18 63.39 100.5 122.1 152.3 220.4 257.7 298.3 518.9 (by industrialpersonnel) Average size o f reporting enterprises 38.26 43.58 42.13 67.42 67.46 74.67 127.9 129.0 212.9 (byactual employment) "Formal"jobs (administrative leave) in 2.51 8.60 21.27 33.14 54.72 77.71 92.49 128.7 85.38 reporting enterprises, thousand % to industrial personnel 6.15 16.47 33.55 32.96 44.78 51.00 41.95 49.95 28.62 Average output per reporting 134.6 131.7 121.2 160.6 158.2 168.5 229.9 227.6 227.5 487.6 enterprise, million 1995 constant drams Number ofnon-reporting enterprises 7301 7262 6956 7000 6710 6344 5718 4430 1971 Average size o f non-reporting 5.90 5.34 6.99 6.77 6.54 6.71 8.60 10.14 27.46 enterprises, employee per enterprise Memorandum item. Value added in 19.8 18.8 20.1 21.8 21.1 19.9 22.4 23.4 24.2 30.2 industry, % of GDP Source: National Statistical Service and authors' calculations. Table 5: Armenian Agricultural Sector: a Comparative View Value Employment, Labor Employed, added, % o fthe world productivity, %to labor force YOo fthe total 1000' US dollars world total World (2000) 100 100 0.98 42.47 Europe (2000) 11 1.o 10.81 4.18 North America (2000) 16 0.4 37.58 3.24 South America (2000) 12 4 2.95 22.35 Countries intransition 7 2 3.44 10.71 (2000) Middle East and North 5 3 1.64 37.26 Africa (2000) Africa (2000) 6 12 0.49 51.16 Asia (2000) 21 38 0.54 54.07 China (2000) 22 40 0.54 66.40 Armenia (2003) 0.048 0.04 I . I 7 41.37 Source: World Bank (2003b); ILO (2004b). 137 Table 6: Labor Productivity and Fixed Capital per Personnel inIndustryby Sector in 1990- 2003 (million 1995 constant Drams)* 2003 2002 2001 2000 1999 1998 1997 1996 1995 1990 Labor productivity per personnel Industry, total 3.312 2.526 1.913 1.597 1.296 1.106 1.043 0.883 0.746 0.977 Energy 1.731 1.597 1.657 1.798 1.659 1.624 1.812 2.360 2.674 7.956 Miningand manufacturing 3.655 2.713 1.963 1.564 1.241 1.030 0.953 0.784 0.655 0.639 Metallurgy 7.154 6.798 6.852 5.336 3.633 2.771 1.562 1.821 1.950 4.262 Machinery and 0.796 0.436 0.347 0.214 0.144 0.171 0.208 0.250 0.246 0.330 metalworlung Chemical industry 1.302 1.143 1.169 1.164 0.812 0.826 0.727 0.668 0.565 0.861 Timber industry 1.700 1.225 0.738 0.657 0.245 0.379 0.766 0.271 0.276 0.355 Building materials 2.901 2.276 1.484 1.161 1.413 1.502 1.264 0.948 0.747 1.230 Light industry 0,457 0.303 0.264 0.181 0.161 0.148 0.123 0.089 0.209 0.346 Foodprocessing 7.427 6.954 5.679 4.556 4.037 3.372 3.305 2.216 1.787 4.233 Other sectors 6.793 5.465 3.595 4.385 5.540 4.242 5.019 4.391 2.688 0.713 Fixed capital per personnel Industry(excl. energy), total 4.202 3.079 2.294 1.668 1.251 1.019 0.738 0.508 0.397 Metallurgy 11.15 6.964 7.537 7.338 7.880 6.824 4.366 4.149 3.856 2 Manufacturing 3.419 2.721 1.960 1.383 0.994 0.794 0.588 0.376 0.283 Machinery and 2.169 1.755 1.227 0.937 0.747 0.653 0.509 0.349 0.271 metalworking Chemical industry 5.658 3.310 2.605 1.102 0.541 0.610 0.362 0.276 0.223 Timber industry 2.363 2.159 1.339 1.492 0.741 1.169 0.293 0.177 0.137 Buildingmaterials 2.505 2.086 1.955 1.655 1.472 1.066 0.923 0.585 0.528 Light industry 1.049 0.730 0.611 0.594 0.354 0.266 0.241 0.141 0.096 Foodprocessing 6.656 6.473 5.135 3.461 2.608 1.847 1.536 0.921 0.689 Other sectors 14.49 16.15 10.70 18.10 17.26 16.56 12.03 6.680 4.220 8 4 4 1 6 5 9 * Labor productivity is calculated by dividing output in constant 1995 prices by the number of personnel in the sectors of industry.Fixed capital is presented at book value. Source: National Statistical Service. 138 Table 7: Unemployment Rate by Age, 1998/99 and 2004 (Percent), ILCS Age groups 1998199 2004 Total (16+) 27.0 19.3 16-24 45.9 43.0 25-54 26.5 17.2 55-64 20.9 13.7 65+ 6.5 5.0 Source: World Bank and NSS (2006). Table 8: Participation, Employment and Unemployment Rates by Regions, 1998/99-2004 (population 16 +), ILCS Total Urban Including: Yerevan Other urban Rural Participation rate 1998199 63.0 56.3 55.0 57.6 72.7 2004 60.3 53.8 53.8 53.8 72.0 Employment rate 1998199 46.0 30.4 28.4 32.3 68.5 2004 48.7 38.4 38.0 38.8 67.2 Unemployment rate 1998199 27.0 46.0 48.3 43.9 5.7 2004 19.3 28.7 29.4 27.9 6.7 Source: World Bank and NSS (2006). Table 9: Population Forecast of Armenia, 2001-2025 Scenario I 2001 2005 2010 2015 2020 2025 2030 Total population 3212.4 3243.4 3282.9 3313.1 3306.9 3253.9 3157.6 Including: 0-14 780.2 644.0 535.0 497.3 487.4 439.0 372.2 15-24 581.4 630.4 601.7 484.8 364.6 325.6 337.5 7-17 684.0 593.1 453.4 374.0 365.4 364.4 324.2 16-retirement 1934.8 2107.2 2314.3 2347.7 2277.5 2164.2 2099.1 15-64 2121.1 2237.1 2407.5 2451.2 2393 2292.7 2204 Retirement+ 431.6 423.3 385.9 430.4 510.1 616.5 653.4 65+ 311.1 361.2 340.4 364.7 426.3 522.2 581.4 Scenario I1 Total population 3212.4 3256.2 3313.0 3390.4 3434.9 3450.5 3424.0 Including: 0-14 780.2 657.8 565.1 565.7 583.2 570.2 519.9 15-24 581.4 630.4 601.7 484.9 377.4 355.6 393.3 7-17 684.0 604.6 459.6 397.0 413.6 437.8 420.5 16-retirement 1934.8 2107.4 2314.3 2350.9 2293.0 2199.1 2169.4 15-64 2121.1 2237.1 2407.5 2454.9 2413.3 2333.5 2284.9 Retirement+ 431.6 423.3 385.9 436.1 523.6 643.4 693.2 65+ 311.1 361.2 340.4 369.8 438.5 546.8 619.1 Source: authors' calculations. 139 Table 10: DistributionofHouseholdsbyMainSourceof Incomes,Percent,2004 LFS M a i n source of income Second source of income Salary (wage) 46.6 0 Income from entrepreneurial 3.1 0.8 activity Income from self-employment 19.1 8.5 Pension, other types o f social aid 17.9 30.7 Income from private peasant farm 6.9 31.8 Income from real estate 0 0.9 Help from relatives 5.9 26.5 Income from sale o f products 0.4 0.7 Other 0.1 0 Total 100 (933 households) 100 (449 households) Source: NSS (2005b). Table 11: DistributionofEmployedWho Want to Changethe Work Situation,ByReasonsand Status of Employment,Percent,2004 LFS Total Employees Self-employed Prospective dismissal o fpeople 0.4 0.8 0 Expiration o f the labour 0.3 0.7 0 contract To earn a higher income 80.6 78.9 80.9 Current work does not 5.5 5.1 6.3 correspond to speciality To work more hours 2.7 3.2 2.9 To work less hours 3.O 3.9 2.5 Improving o f worlung 5.3 4.7 5.2 conditions Current work place is far from 0.9 1.4 0.6 home Healthreasons 0.8 0.7 1.3 Other 0.5 0.7 0. Total 100 100 100 Source: NSS (2005b). 140 Table 12: Average Wages, MinimumWages and Pensions in CIS States In2003 and 2004 Average Minimum Ratio of Average Average Ratio of wage, wage (as minimum monthly pension the 2004 of end- wage to nominalwage (as of average year), average converted into end-year, pension 2004 wage, 2004 U S dollars 2003) to (US$) average wages, 2003* 1995 2004 Azerbaijan, 483400 100000 0.207 14,l 98,4 119111 26.8 manats Armenia, drams 41976 13000 0.310 17,4 78,7 7452 19.4 Belarus, 350200 128390 0.367 65,5 161,8 113943 40.6 Belarusian rubles Georgia, lari 126 ... ... 10,5 ... 16.5 13.0 (2003) Kazakhstan, tenge 28200 6600 0.234 78,5 207,3 6617 24.9 Kyrgyzstan, soms 2203 100 0.045 34,O 51,7 553 27.5 Moldova, lei 1104 100 0.091 31,8 893 211 20.9 Russia, rubles 6828 600 0.088 103,7 237,O 1747 29.2 Tajilustan, somoni 64 7 0.109 8,6 21,5 13.07 25.7 Turkmenistan, 1750000 ... ... ... ... ... ... manats Uzbekistan, sums ... 6530 ... 35,8 ... ... ... Ukraine, grivnas 590 237 0.402 49,8 110,8 182 34.5 *- Note: average pension in 2003 is calculated as an average for end-year 2002 and end-year 2003 pensions. Depending on the timing for pension increase during the calendar year, the actual averagepension for 2003 may differ. Source: CIS STAT (2005). Table 13: Average Time Spent on Activities, Proportion Undertaking An Activity and Average Time Spent by Those UndertakingActivities During July 2004 (hours, minutes and percentage respectively; Armenian population 15-80 years; all days) Average Proportion undertaking Average by those the activity undertaking activities Males Females Males Females Males females Gainful 5:18 1:44 65 34 8:12 5:04 employment Housework 1:06 5:46 62 96 1:45 5:59 Personal 11:44 11:48 100 100 11:44 11:48 needs Study 0:04 0:2 1 3 8 2:17 4:30 Free time 5:47 4:20 98 97 5:53 4:29 Other, 0:Ol 0:Ol 2 0 1:20 250 unspecified Total 24:OO 24:OO 100 100 Source: NSS (2005a). 141 Table 14: Distribution of RespondentRural Households by Size Of Agricultural Land, Percent Strata Landles House- House- House- Household Househol Household Househo Total s house- holds with holds with holds s with 2-3 ds with s with 5-10 Idswith holds, up to 0.5 0.5-1 ha with 1-2 ha 3-5 ha ha agricul- 10 and YO ha agricul- agricul- ha agricul- agricul- tural land, more ha tural land, tural land, agricul- tural land, tural % agricul- % Y O tural YO land, YO tural land, % land, % Strata 1.8 42.1 34.7 18.2 2.9 0.0 0.0 0.3 100.0 1 Strata 6.5 37.5 27.8 19.5 4.3 3.2 1.2 0.0 100.0 2 Strata 0.8 28.7 17.7 26.3 14.3 9.8 1.8 0.6 100.0 Strata 1.8 22.0 15.9 24.1 12.3 16.6 5.7 1.6 100.0 4 Strata 1.9 31.2 8.7 9.4 10.6 13.1 15.6 9.5 100.0 5 Total 2.4 31.5 21.5 21.7 9.6 8.6 3.4 1.3 100.0 Source: Rural Employment Survey 2005. Table 15: Investments inFixed Assets (inbillion Drams), 1999-2004 1999 2000 2001 2002 2003 2004 Total investments (excluding housing) 52.8 79.9 81.7 101.4 124.7 147.6 Agriculture 10.5 6.6 5.5 6.9 6.2 7.3 in % to total 20.0 8.2 6.7 6.8 4.9 4.9 Memorandum item: Share of agriculture inGDP, % 27.0 23.2 25.5 23.4 21.5 22.6 Source: National Statistical Service and authors' calculations. Table 16: Stratification and Distribution of the Sample by Strata: RuralEmployment Survey 2005 Strata Agricultural Share of Number of Share of Number of Share of Vulnerability population surveyed surveyed surveyed surveyed Index (AVI)* intotal households households, population population, for population sample, YO % Y O Strata 1 Upto 15 17.0 340 17.0 1,670 17.2 Strata 2 15-30 20.2 400 20.0 2,002 20.7 Strata 3 30-60 33.5 660 33.0 3,147 32.5 Strata 4 60-70 22.3 440 22.0 2,182 22.5 Strata 5 70 and more 7.9 160 8.0 692 7.1 *Total 100.0 2,000 100.0 9,693 100.0 0- least vulnerable, 100-most vulnerable. Source: Rural Employment Survey 2005. 142 Table 17: Economic Activity and Unemployment: Rural Employment Survey 2005 Strata Employ- Non-emp- Unem- Economi- Economi- Unem- Propor- Adjusted ment-to- loyed, % ployed, % cally cally ployment tion of unemp- populatio of (share of active inactive rate, YO under- loyment nratio, populatio unemp- popu- populatio employed rate, % YO(share naged loyedPO- lation n, YO willing to of emp- 15+ pulation (employed work full loyed aged 15+ +unemp- time (UR), populatio intotal) loyed), YO Yoof eco- naged nomically 15+ in active po- total) pulation Strata 1 77.4 22.6 0.9 78.3 21.7 1.1 33.6 9.5 Strata 2 74.8 25.2 1.7 76.4 23.6 2.2 31.4 12.7 Strata3 79.3 20.7 0.7 80.0 20.0 0.9 32.8 14.6 Strata 4 81.0 19.0 0.5 81.5 18.5 0.6 36.1 17.1 Strata5 85.6 14.4 0.4 86.0 14.0 0.5 31.0 17.3 Total 78.8 21.2 0.9 79.7 20.3 1.1 33.2 14.1 Source: Ruralemployment survey 2005. Table 18: Average Monthly Income from Crop Productionper Household, Drams Average monthly of which: Average monthly Income from income from crop income from estimate of income from crop production production per production own production per sold/bartered as household soldhartered, consumption, hectare of land, share of total (including own dram dram dram estimated income, consumption in Y O household/farm), dram Strata 1 20,175 12,100 8,075 33,074 60.0 Strata 2 21,695 16,729 4,965 31,904 77.1 Strata 3 21,252 8,293 12,960 16,475 39.0 Strata4 17,904 5,182 12,722 11,404 28.9 Strata 5 35,474 8,521 26,953 13,644 24.0 Total 21,538 9,961 11,577 17,654 46.2 Source: RuralEmploymentSurvey 2005. 143 Table 19: Average Monthly Income from Livestock per Household, Drams Share of Average monthly of which: Incomefrom households incomefrom incomefrom estimateof production engagedin animal livestock production own sold/bartered as husbandry, YO ProductionPer sold/bartered, consum- share of total household dram ption, dram estimatedincome, (including own YO consumptionin household/farm), Strata 1 40.4 6,399 2,427 3,971 37.9 Strata 2 41.9 8,386 2,890 5,496 34.5 Strata 3 56.1 14,168 3,825 10,343 27.0 Strata 4 60.9 22,558 8,630 13,927 38.3 Strata 5 61.2 36,880 14,014 22,866 38.0 Total 64.9 15,321 5,254 10,067 34.3 Source: Rural employment survey 2005. 144 td c 8 td fi 8 8 td fi 8 8 8 8 8 8 8 8 8 8 8 8 8 d fi E E 2 E 0 0 0 0 0 W 0 0 00 10 m N -zs 0 3 c;' c;' 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 m 0 0 m z d m 0 0 0o 0o 0 0 0 0 W d ~ W m W m E= d 0 z z mm0m 00 NoWm0 N0m 00 00 m m m m m w *0 m m0 m0 0 0 0 0 0 0 m m 0 0 m W W m m 00 W m m r- W W d d c 8 8 8 8 l- v d 0 N 0 m m w w rn 8 8 8 8 8 iWT0 t? W m m 00m 0 mC.l co m W m N m m W 00 m 00 r - m 2 d (0 m \o r-00 m td c c td 8 d fi o c W U o c N 9:C c n c s : N m m m W r- 13 13 m m cn2 Table 21: Distribution of RegionalEmployment Centers (REC) by Number of Job-seekers and Unemployed per Staff Member, December 2004 Number ofjob-seekers Number of unemployed Number ofjob-seekershnemployedper Number of % of total Number of % of total REC staff centers number of centers number of centers centers <50 0 0.0% 3 6.3% 50-100 6 11.8% 13 27.1% 101-200 17 33.3% 17 35.4% 201-300 14 27.5% 11 22.9% 301-400 6 11.8% 2 4.2% >400 8 15.7% 5 10.4% Memorandum items The largest value for regional centers, per REC 56 23 staff The smallest value for regional centers, per 1,132 911 REC staff Average for the countryby all regional centers, 344 265 per REC staff Source: Survey of RegionalEmploymentCenters 2004. 146 Table 22: Public Expenditure on Labor Market Programs: International Comparisons Unemployment rate Public expenditure ratio of GDP % % of GDP expenditure to unemploymentrate OECD, average (1985) 8.2 2.29 0.28 OECD, average (1998) 7.7 1.96 0.26 OECD, average (2001) 6.4 1.64 0.26 OECD, average (2002) 6.9 1.77 0.26 Czech Republic (1998) 6.5 0.36 0.06 Czech Republic (2001) 8.2 0.45 0.05 Czech Republic (2002) 7.3 0.45 0.06 Slovak Republic (1998) 12.6 1.25 0.10 Slovak Republic (2001) 19.3 0.91 0.05 Slovak Republic (2002) 18.6 0.94 0.05 Slovenia (1998) 7.7 1.21 0.16 Slovenia (2001) 5.9 1.04 0.18 Slovenia (2002) 5.9 0.88 0.15 Baltic States Estonia (1998) 9.9 0.16 0.02 Estonia (2001) 12.6 0.22 0.02 Latvia (1998) 14.1 0.72 0.05 Latvia (2001) 13.1 0.64 0.05 Lithuania (1998) 13.3 0.25 0.02 Lithuania (2001) 0.27 0.02 Armenia (1998) ...17.0 (9.4) 0.10 ... (0.01) Armenia (2001) 38.1 (10.4) 0.07 0.002 (0.01) Armenia (2002) 35.3 (10.8) 0.08 0.002 (0.01) Armenia (2003) 31.2 (10.1) 0.07 0.002 (0.01) Note: Data for OECD are calculated as unweighted averages for those countries for which data for corresponding years are available. Particularly, for 1985, 1998,2001 and 2002 unweighted averages are calculatedbased on data for 19, 25, 23 and 19 countries, respectively. Unemployment rates are based on labor force survey statistics. For Armenia, in parenthesis officially registered unemployment rates and correspondingratios of GDP spending to unemployment rates are presented. Source: National Statistical Service, State Employment Service, OECD 2004b, Employment Service o f Slovenia, Statistical Office o f Slovenia, OECD data base, TransMONEE 2003 Database, UNICEF IRC, Florence, "Labour Flexibility and Migration in the EUEastward EnlargementContext: The Case o f the Baltic States", Ezoneplus Working Paper No. 11,2002. 147 IBRD 33364 N°14 N°04 N°93 E° 47 E° 47 voir Mingechevir Reser sarA FüzilioT ARMENIA To Qubadli To Qubadli ' AZERBAIJAN Goris Kapan Vorotan n g e R a Megri GEORGIA e SYUNIK'SYUNIK z u r g E° a n 46 Angekhakot Z OrdubadoT Kura GäncäoT ' e g tsvashen To ardenisV ' Jer Ar Gäncä naR muk aikV NaxçivanoT GäncäoT UNIK' tuni Mar mir Lake Sevan YOTS'AVYOTS DZOR egegnadzorY Kar aprA AZERBAIJAN Ijevan VUSHAT sinedraV GEGHARK'UNIK'GEGHARK Kilometers ount a i n s Miles Gavar Aras 50 30 E° T To 45 Dilijan Naxçivan 04 Sevan Hrazdan M K O TAY K 'AN 0 02 ARARA Ararat 03 T'bilisioT Alaverdi aucasus nadzarH IRAN 01 YEREV AN tashat REPUBLIC Ar Aras 021 anadzorV YEREV OF 0 0 C LORRI m) T'bilisioT ashirT E° Lesser ISLAMIC Stepanavan Aragats (4090 TSOTN Ashtarak tik Ar E° 44 VIR 44 Gyumri ARAGA mavir Bank. Ar Bank any SHIRAK ARMA World or information The World BorjomioT of other ALS To TOWNS Kars aprA The territory, Unit any of any and part of CAPIT BOUNDARIES AL BOUNDARIES Design AND the GEORGIA on status boundaries. Map such the legal (MARZ) CAPIT (MARZ) of CITIES imply, TIONAL by denominations the not on ROADS Aras do colors, TIONAL acceptance produced map judgment or SELECTED PROVINCE NA RIVERS MAIN RAILROADS PROVINCE INTERNA TURKEY was this on any ARMENIA map boundaries, E° This The shown Group, endorsement 43 E° 43 °N °N 41 40 SEPTEMBER 2004