Report No. 48510-MK FYR Macedonia Poverty, Jobs and Firms: An Assessment for 2002­2006 November 19,2009 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank CURRENCY AND EQUIVALENT UNITS (Exchange Rate Effective May 25, 2005) Currency Unit = Denar (MKD) US$1.00 = 47.65 MKD Fiscal Year January 1 to December 31 Acronyms and Abbreviations CEM Country Economic Memorandum CIS Commonwealth of Independent States EBRD European Bank of Reconstruction and Development ECA Europe and Central Asia EU European Union FYR Former Yugoslav Republic GDP Gross Domestic Product GNI Gross National Income GOM Government of FYR Macedonia HBS Household Budget Survey HIV Human Immunodeficiency Virus LAC Latin American and Caribbean LFS Labor Force Survey MDG Millennium Development Goals MIC Middle income countries MKD Macedonian Denar MPI Migration Policy Institute OECD Organisation of Economic Co-Operation and Development PA Poverty Assessment PPP Purchasing Power Parity SSO State Statistical Office of Macedonia UN United Nations UNDP United Nations Development Programme UNICEF United Nations Children's Fund WB World Bank WDI World Development Indicators Vice President: Philippe H. Le Houerou Country Director: Jane Armitage Sector Director: Luca Barbone Sector Manager: Benu Bidani Team Leader: Sarosh Sattar ii Table of Content KEY RESULTS FOR 2002-2006 AND RECOMMENDATIONS .................................................... IX EXECUTIVE SUMMARY .................................................................................................................... XI Poverty and Inequality ............................................................................................................. xiii Income Opportunities and Poverty ............................................................................................xv Job Dynamics in FYR Macedonia ......................................................................................... xviii The Way Forward ......................................................................................................................xx CHAPTER 1. LIVING STANDARDS IN FYR MACEDONIA ..............................................................1 Introduction ..................................................................................................................................1 Has higher growth helped FYR Macedonia reduce poverty? ......................................................2 Living Standards In FYR Macedonia ..........................................................................................4 CHAPTER 2. SELECTIVE GROWTH AND THE POVERTY TRAP: POVERTY TRENDS, 2002-2006 .9 Introduction ..................................................................................................................................9 Estimating Poverty In FYR Macedonia .....................................................................................10 Poverty trends 2002-2006 ..........................................................................................................11 Growth and redistribution of consumption ................................................................................13 Spatial Poverty divergence ........................................................................................................14 CHAPTER 3. THE OLD AND NEW POOR: A SNAPSHOT OF LIVING STANDARDS IN 2006 ..........19 Where do they live? The spatial aspect......................................................................................19 Who are they? Characteristics of households and household members ....................................20 How do the poor live? ................................................................................................................26 Subjective opinions on welfare and welfare changes ................................................................28 CHAPTER 4. GROWTHLESS JOBS AND JOBLESS GROWTH: INCOME OPPORTUNITIES AND POVERTY ..........................................................................................................................31 Falling income ...........................................................................................................................31 Growth patterns - no trickle down .............................................................................................32 Growthless jobs and jobless growth ..........................................................................................33 Social assistance reaches less than half of the poor ...................................................................37 CHAPTER 5. JOB DYNAMICS BEYOND SUMMING UP AND NETTING OUT...............................41 Labor markets lost in transition?................................................................................................41 Job dynamics in FYR Macedonia ..............................................................................................43 Job dynamics at a disaggregate level ­ who is creating jobs and where?..................................45 CHAPTER 6. CONCLUSIONS: SUSTAINABLE JOBS AND SAFETY NETS .....................................51 The Way Forward ......................................................................................................................52 REFERENCES ..........................................................................................................................55 INTRODUCTION ...................................................................................................................................67 BACKGROUND .....................................................................................................................................69 Brief Review of the Labor Market in Macedonia ......................................................................69 The Issue of Non-participation in the Labor Market .................................................................70 Data Collection Process .............................................................................................................73 Structure of the Sample ..............................................................................................................74 Data Collection Methods ...........................................................................................................75 iii RESULTS ......................................................................................................................................77 History of (Un)employment .......................................................................................................77 Motivation for Re-qualification and Skills Upgrading ..............................................................78 Reasons for Not Working / Not Seeking Employment ..............................................................79 Cultural Perspective on the Issue of Female Employment ........................................................85 Management of Income and Productive Assets .........................................................................87 Access to Infrastructure .............................................................................................................90 Motivation for Employment ......................................................................................................90 Feelings of Discrimination .........................................................................................................91 Recommendations for Increasing Participation in Low-Skilled Minority Women ...................93 General Recommendations for Increasing Participation ...........................................................95 Recommendations for Adjusting the LFS..................................................................................96 List of Figures Figure 1.1: Poverty Reductions in the ECA Region ....................................................................... 2 Figure 1.2: Job-Less Growth, Growth-Less Jobs............................................................................ 3 Figure 1.3: Three Data Sources: One Message............................................................................... 4 Figure 1.4: Output Growth and Job Creation Remain Low by Regional Standards....................... 4 Figure 1.5: Macedonia has Made Progress on Several MDGs ....................................................... 8 Figure 2.1: The Poverty Headcount Index Stagnated between 2002 and 2006 ............................ 12 Figure 2.2: Poverty is Becoming Deeper and More Severe as Well ............................................ 12 Figure 2.3: The Poorest have Seen Shrinking Real Consumption Levels Since 2002 ................. 13 Figure 2.4: Low and Volatile Growth together with Increasing Inequality has Stalled Poverty Reduction......................................................................................................................... 14 Figure 2.5: Consumption Growth has been Anti-Poor ................................................................ 14 Figure 2.6: Skopje is Diverging from the Rest; The Eastern Regions Account For the Overall Increase in Poverty in FYR Macedonia .......................................................................... 15 Figure 2.7: Regional Divergence .................................................................................................. 15 Figure 2.8: Eastern Regions have Seen Big Falls in Consumption Levels for the Poor and Vulnerable ....................................................................................................................... 17 Figure 3.1: The Extremely Poor Live are Concentrated in the North ........................................... 20 Figure 3.2: Large Households with Many Children are the Poorest Even when Economies of Scale and Age Profiles are Taken into Account ........................................................................... 21 Figure 3.3: Poverty Rates have Increased for Small Children, their Parents, and for the Elderly 21 Figure 3.4: Poverty Rates have Increased among those Living with Elderly Household Heads.. 22 Figure 3.5: A Vast Majority of the Poor Are Low Educated; they have Seen Highest Poverty Increases .......................................................................................................................... 22 Figure 3.6: Household Head's Lack of Education is A Strong Correlate of Poverty... ............... 23 Figure 3.7: The Poor are More Likely to Lack Health Insurance ................................................. 24 Figure 3.8: Housing Conditions have Improved for the Poor ....................................................... 27 Figure 3.9: Food Expenditures Constrain the Non-Food Expenditures of the Poor ..................... 28 Figure 3.10: Subjective Poverty Measures Indicate an Improvement in Wellbeing .................... 29 Figure 4.1: Private Consumption and Private Services are the Two Drivers of Growth from Expenditure and Activity Sides ....................................................................................... 32 Figure 4.2: Net Migration has Increased and so Have Remittances ............................................. 33 iv Figure 4.3: More Jobs, but Not Better Earning Opportunities ...................................................... 35 Figure 4.4: Poverty Rates among the Working are Increasing ..................................................... 35 Figure 4.5: Access to Labor Markets is Losing Importance for Poverty Reduction .................... 36 Figure 4.6: Average Household Size Has Increased Among the Very Poorest ............................ 36 Figure 4.7: Job Creation is Not Resulting in Poverty Reduction at a Regional Level ................. 37 Figure 4.8: Social Protection performance ................................................................................... 38 Figure 4.9: Social Assistance Coverage has Fallen, Pensions Coverage Improved ..................... 39 Figure 4.10: Pensions Income is Critical for Poor Receiving Households, .................................. 39 Figure 4.11: Remittances are Reaching a Higher ­ but Still Limited ­ Share of Poor Households ....................................................................................................................................................... 40 Figure 5.1: Job Turnover, Different Indicators ............................................................................. 43 Figure 5.2: Job Creation and Job Reallocation Increased between 2002 and 2006 1/................... 44 Figure 5.3: Firm Survival Improved over Time............................................................................ 44 Figure 5.4: FYR Macedonia Has Comparatively High Firm Turn-Over...................................... 45 Figure 5.5: Highest Efficiency in Job Creation in Services .......................................................... 46 Figure 5.6: A Shift Towards Higher Job Creation in the Manufacturing and Services Sectors ­ Except in the Textiles and Transports Sectors, which Account for Nearly Half of Employment 47 Figure 5.7: Job Creation is Highest among Smaller Firms, and Only the Private Sector is Creating Jobs on a Net Basis ........................................................................................... 48 Figure 5.8: Significant Net Job Creation Only in Low Wage Firms ............................................ 49 Figure 5.9: Firms with Low (Negative) Productivity Growth are Creating Jobs ......................... 50 Figure 6.1: Very Fast Job Growth would be Needed to Reach the Target Employment Rate ..... 51 List of Tables Table 1.1: FYR Macedonia Performs Quite Well on Selected Well Being Indicators................... 6 Table 1.2: International Poverty Comparisons Suggest Low but Resistant Poverty /a .................. 7 Table 1.3: Significant Spatial Inequality in FYR Macedonia, 2004 /a ........................................... 7 Table 2.1: Extreme and Total Poverty Lines 2002-2006 .............................................................. 10 Table 2.2: Inconsistent Poverty Comparisons .............................................................................. 11 Table 3.1: Poverty is an Urban Phenomenon................................................................................ 19 Table 3.2: Access to Good Jobs, not only Jobs, is Crucial for Poverty Reduction....................... 24 Table 3.3: Housing Conditions Gap is Most Significant in Urban Areas ..................................... 27 Table 4.1: Stagnant Income Growth and Falling Transfers for the Poorest ................................. 32 Table 4.2: Key Labor Market Indicators Have Improved............................................................. 34 Table 5.1: Key Numbers from the Firm Registry ......................................................................... 43 Table 5.2: Job Dynamics in Skopje vs. Other Areas .................................................................... 45 List of Boxes Box 1.1: What do we mean by poverty? ........................................................................................ 5 Box 2.1: Do Poor People Move ­ And do they Bring Poverty with Them?................................. 16 Box 3.1: Are Women in Macedonia Facing an Under Participation Trap? .................................. 25 Box 4.1: Labor Market Indicators ­ Sources and Definitions ...................................................... 34 Box 5.1: Key Concept and Data ................................................................................................... 42 v ACKNOWLEDGEMENTS This report was produced by the World Bank with cooperation from the State Statistical Office of the Republic of Macedonia. It is the culmination of a process which began in 2002 when the two institutions agreed to work collaboratively in order to understand the poverty situation based upon the availability of better data that would allow drawing robust conclusions on living standards. The task on the World Bank side was carried out under the leadership of Sarosh Sattar. The team members were Sara Johansson de Silva (consultant, poverty analysis and labor and firm analysis), Yeva Gyulnazaryan (research assistance, poverty monitoring evaluation), Diego Angel-Urdinola (labor markets and gender), Victor Macias (consultant, labor markets), Helena Makarenko (administrative assistance), and Evgenij Najdov (economist). The report was prepared under the guidance of Asad Alam (then Sector Manager) and Luca Barbone (Sector Director, ECSPE). Useful guidance and comments were provided by Markus Repnik (Country Manager), Bernard Funck (Sector Manager, ECSPE), Kathy Lindert (Sector Manager, ECSHD), Sanjay Kathuria and Caterina Laderchi (peer reviewers), and Bojana Naceva (education specialist, ECSHD). The Annex: "Constraints to Labor Force Participation in Macedonia: A qualitative approach" was prepared by the Centre for Research and Policy Making (CRPM) in close collaboration with a team comprising Diego F. Angel-Urdinola (Economist ECSHD, Task Team Leader), Arvo Kuddo (Senior Economist, ESCHD), and Bojana Naceva (Education Economist, ECSHD). Helpful guidance, comments, and discussions with the team were also provided by Gordon Betcherman (Sector Manager and Lead Economist, ECSHD), who provided mentoring and guidance and Victor Macias (Consultant, ECSHD). The Annex was prepared under the leadership of Tamar Manuelyan Atinc (Sector Director, ECSHD). vii KEY RESULTS FOR 2002-2006 AND RECOMMENDATIONS Non-income dimensions of poverty improved while income poverty deteriorated despite economic growth. - FYR Macedonia's social indicators improved, especially indicators of health, hunger, and access to basic amenities. - Total poverty1 remained stagnant at 19 percent in 2002 and 2006, while extreme poverty increased from 4.5 percent to 7.4 percent of the population. - Spatial inequality increased. Skopje saw a decline in poverty while the Northern and Eastern areas saw increasing total poverty, due to higher extreme poverty. - Though GDP per capita was rising on average by slightly less than 4 percent 2002-2006, the average (real) consumption per capita of the poor fell by some 7 percent. Few net jobs were created mostly in low productivity sectors. - Jobs were not created in the growth sectors. Many new jobs were unpaid family work. - The working poor increased, especially in secondary urban centers. They held low-wage low- productivity jobs. - Despite low net employment growth, the enterprise sector had high job creation and destruction rates, characteristic of an economy in an earlier stage of transition. - Mostly firms with negative productivity growth created jobs. The social assistance program helps the poor but not enough. - The social assistance program exhibits good performance by international standards as measured by the proportion of its total transfers received by the poorest 20 percent of the population. - Nevertheless, in 2006 less than 40 percent of households living in extreme poverty received social assistance compared to over 50 percent in 2002. - An estimated 60 percent of total transfers of the social assistance programs were received by the non- poor in 2006. - About one-third of the poor lived in households that received pension income. - Social assistance income contributed 32 percent to total consumption of recipient households in extreme poverty. Facing the challenge: Protecting the poor from the impact of the global economic crisis - Sustainable economic growth is essential to poverty alleviation. Thus, weathering the global crisis and undertaking necessary structural reforms could lay the foundation for future growth and poverty reduction. 1 In 2006, the extreme and total poverty lines were respectively 57.3 and 102.0 denars per capita per day. Those living in extreme poverty were unable to meet basic food needs, while those living below the total poverty line were unable to meet basic food and non-food needs. ix - The poor's welfare is linked to improvements in competitiveness. Reforms to improve the business environment, retrain workers, and develop infrastructure are important. - Upgrading the education system by improving school quality and expanding adult education will increase the productivity of both new job entrants as well as the stock of existing labor. - Avenues to improve the social welfare system include establishing a single national registry of social benefit programs and beneficiaries which will increase transparency and improve targeting, standardizing application of eligibility criteria for different types of social benefits, and introducing a comprehensive definition of total countable income for social assistance purposes. x EXECUTIVE SUMMARY 1. The prevailing global crisis has undermined economic progress in many countries and FYR Macedonia is no exception. After robust growth for much of this decade rising to near 6 percent in 2007 and much of 2008, FYR Macedonia's economy abruptly has come to a virtual standstill as measured by output growth. Though economic growth stalled largely due to adverse external conditions, the impact will potentially be felt locally through the deterioration of the population's living standards and an expansion of the number of poor. Thus, among all the other economic and social challenges facing the government at this difficult juncture in the country's history, ensuring that those least able to help themselves are not overlooked should remain among its top priorities. 2. This poverty assessment report is based upon pre-crisis data for the period 2002-2006. Though some may have reservations that this information base is not well suited to informing public policy choices of today, this view would be short-sighted. Economic development is not an overnight occurrence. It is also not something that can be reversed through an economic shock ­ no matter how virulent. The analysis of data from the first half of this decade reveals small though perceptible shifts in the transmission channels for poverty reduction, that is, labor markets and enterprises. Thus, the reform process which began earlier this decade indeed saw some positive changes emerging in the real sector. These are captured in this report and provide an understanding of how the complex issue of poverty can be addressed in the future. 3. Nevertheless, FYR Macedonia is confronted with significant policy challenges on multiple fronts. By 2006, though non-income dimensions of poverty had improved modestly, poverty had not fallen in FYR Macedonia despite economic growth during 2002-2006. This conclusion has been reached based upon the analysis of multiple data sources that reveal important facets of the growth-employment-poverty nexus (see Figure 1). Three main findings are as follows: · Economic growth in 2002-2006 did not lead to declining poverty. The average consumption per capita of the poor fell, inequality rose, and the number of working poor grew. In addition, safety nets became less effective in helping the poor. · Job creation in the analyzed period had not been a channel for poverty alleviation as new job creation was primarily in unpaid or low wage employment. Many of these jobs were not in the sectors with high value added growth. · Net formal job creation was by firms with negative productivity growth. Thus, these jobs were not a means out of poverty as they were low wage and unlikely to be long-lasting. Furthermore, job destruction rates were high in FYR Macedonia indicating frequent turnover in formal sector jobs. 4. Though the overall situation may appear discouraging, this is not the case for several reasons. First, we saw the emergence of so-called "green shoots" in the latter part of 2002-2006. Two of the most important developments were (a) an increase in the share of the population holding jobs and (b) an increase in net job creation by formal sector enterprises. Second, the experience of FYR Macedonia in terms of moderate growth and little or no change in poverty was seen in other economies in the Western Balkans such as Albania, Serbia and Kosovo. Moreover, poverty data for these countries reveal that after a long incubation period, there was a reduction in poverty in the post 2005-2006 period. Third, the Government has responded aggressively to improve the targeting of the safety net and adopt pro-growth policies. Given these developments (the recent crisis not withstanding), there is hope that the basis for xi sustained poverty reduction may exist, especially if a range of structural reforms are adopted which accelerate growth. Figure 1 The Growth-Employment-Poverty Nexus Household Surveys Increasing inequality, regressive consumption growth, spatial inequality, more working poor and under-performing safety nets No Poverty Reduction in FYR Macedonia Labor Force Surveys and Firm data macro data Firms with negative labor Job creation but unpaid jobs, productivity growth create primary sector, disconnected jobs from output growth. 5. Thus, given the myriad challenges facing the Government, not least the short-term fall-out of the global economic crisis, what policy areas that if focused on would yield benefits in terms of poverty reduction in the short- and medium-terms? The list of five interventions is as follows: · Accelerate economic growth and job creation through structural reforms. Reforms that would promote competition, strengthen property and creditor rights and reduce the administrative burden on businesses are a key requirement for stronger economic growth. · Improve competitiveness and focus on those reforms that have immediate welfare benefits ­ streamlined bankruptcy procedures, worker retraining, further reduction in the labor tax wedge for low wage earners and infrastructure development. · Upgrade the education system by improving quality of teaching and learning as well as school management. Implement the new law on adult education. In the short term, launch additional training and retraining courses for young unemployed on skills required by the labor market. · Improve the social welfare system by completing the single registry of social benefit programs and beneficiaries, standardizing the application of eligibility criteria, consolidating cash benefits, introducing a comprehensive definition of total countable income for social assistance purposes and improving targeting and coverage of the poor. · Improve the poverty monitoring system through the adoption of an absolute poverty line and modifications in the household survey. 6. The list above overlaps with areas that the Government is already committed to reforming and in some cases is even working on with the World Bank's support. These reforms will have a direct impact on poverty reduction Moreover, as seen in the specific reforms highlighted in each policy category, there are specific actions which even in the short-term can improve the poor's welfare. 7. The remainder of this executive summary provides an overview of the main findings of the analysis. We cover the following main areas discussed in the poverty assessment: (i) a description of the extent of poverty, the characteristics of the poor, trends in non-income and income dimensions of poverty xii over 2002-2006, and inequality; (ii) the role of economic growth, employment changes, and social transfers in affecting poverty rates; (iii) job dynamics and the characteristics of firm birth and destruction and its linkages to job opportunities, labor productivity, and long-term income growth; and (iv) policy reforms that would lead to poverty reduction. Table 1 Changes in Poverty, 2002-2006 (in percent) Trends in Total and Extreme Poverty Total Poverty Rates by Rural/Urban 25 30 25 20 25 21 21 15 19 19 20 20 18 10 7 5 15 5 12 10 - 2002 2003 2004 2005 2006 5 Total poverty - Skopje Other urban Rural Extreme poverty GDP growth per capita 2002 2006 Source: World Bank staff estimates based on the Household Budget Surveys (SSO, 2002-2006). POVERTY AND INEQUALITY 8. FYR Macedonia is a middle income country with GNI per capita of $4,629 in 2008. The economy has grown at a rate of 3.7 percent per annum during 2002-2006, the period covered by this poverty assessment. Also, in December 2005 FYR Macedonia achieved candidate status to the European Union. Though, selected measures of the process towards developing a market economy shows progress since the 1990s, FYR Macedonia's economy was not catching up with the rest of the region: its growth and employment rates were lower than other comparable economies though in 2007 and the first half of 2008 there has been a turnaround. 9. Non-income dimensions of poverty as measured by human development indicators improved in several areas but remain relatively low by regional standards. Health and education outcomes have improved since the 1990s, together with access to basic infrastructure. Specifically, life expectancy, infant mortality rates, and nutrition rates improved while youth literacy rates ­ which were already high ­ stabilized. Although FYR Macedonia performs well compared to other middle income countries, in 2006 (the latest available data), it was ranked the fourth lowest (above Albania and Bosnia and Herzegovina, and Moldova) on the UNDP's human development index of all countries in Eastern and Central Europe (0.808, compared to 0.823 for the Western Balkans and 0.859 for the EU7 countries).2 2 The average for the Western Balkans includes Albania, Bosnia and Herzegovina, Croatia, Montenegro, and Serbia. The EU7 countries are Bulgaria, Hungary, Latvia, Lithuania, Poland, Romania, and Slovak Republic. xiii 10. The concept of poverty utilized throughout this report is that of "absolute poverty" and relies on a basic needs approach to measure deprivation in a consistent manner over time.3 The previous poverty assessment report, FYR Macedonia: Poverty Assessment for 2002-2003, provides a detailed explanation of the methodology used to compute the poverty lines while Box 1.1 of this report provides a summary. There are two levels of absolute poverty used in this document ­ that of total and extreme poverty.4 Those persons with consumption below the total poverty line are unable to meet their food and nonfood needs. Those living in extreme poverty have consumption levels which fall below what is needed to meet basic caloric (food) needs of about 2100 calories per person per day, and hence are in the most dire straits. The greatest value of this concept of absolute poverty is that it allows a government to monitor whether its actions and any reforms it undertakes are lifting up the people at the bottom of the income distribution to achieve a minimal standard of living. It can also be valuable in helping the government target its social assistance programs to those who are in the greatest needs. 11. The share of the population living in poverty stagnated while those in extreme poverty grew to capture a larger share of the population over 2002-2006.5 Developments in poverty were not favorable for FYR Macedonia in the first half of this decade. The proportion of the population living below the total poverty line remained at about one-fifth of the population during 2002-2006 though with some modest variation during the interim years. The level of extreme poverty grew significantly, especially during 2005-2006, increasing to 7 percent of the population. Moreover, there were other aspects of poverty that indicate an even less favorable situation. Specifically, poverty became deeper and more severe; that is, the gap between the average consumption of the poor from the poverty line increased and more so for the extreme poor. In fact, the poorer fifteen percent of the population became worse off since 2002. 12. Deepening poverty is linked to low and volatile private consumption growth together with a rise in inequality. Thus, the Gini coefficient of consumption increased from 0.38 to 0.40 for FYR Macedonia during 2002-2006. Inequality increased in both urban and rural areas, but particularly in the latter.6 Because growth patterns were not pro-poor, the poorest three deciles saw their share of consumption shrink below ten percent of total national household consumption ­ with the biggest drop for the poorest ten percent. Moreover, the poorest quintile of the population consistently experienced a decline in its average consumption per capita during 2002-2006. During 2004-2006 the middle class (the second and third quintiles, 20-60 percentiles of the population) also experienced a significant decline in average consumption per capita, leaving only the top 25 percent of the population seeing a gain in 2005- 2006. 13. Spatial inequality has grown. The overall stagnation also masks regional differences. Skopje pulled away from the rest of the country, and the northeastern part fell behind other regions. More specifically, moderate poverty (the poor who are not among the extreme poor) fell in Skopje, contributing to a lower overall poverty rate; at the same time, extreme poverty doubled in secondary cities, exceeding extreme poverty rates in rural areas. Between 2002 and 2006, secondary urban areas saw an increase in the number of poor by 28,000 persons while Skopje saw a fall in a comparable amount. Similarly, two of the poorest regions, notably Northeast and Vardar, together with the Eastern region, saw sharp increases 3 Following EU practice, FYR Macedonia's official poverty line is a relative poverty line. However, absolute poverty continues to exist in FYR Macedonia. Since relative poverty lines measure distribution only, it cannot evaluate the change in living standards over time. 4 In 2006, the extreme and total poverty lines were respectively 57.3 and 102.0 denars per capita per day. 5 Throughout this report, when "poverty" is used by itself, unless specified otherwise, it refers to the "total poverty" using the basic needs approach. 6 Large changes in the Gini coefficient are unusual and further work needs to be carried out to determine what has led to this result. xiv in extreme poverty between 2004 and 2006 (regional data is only available for this period). This increase in extreme poverty shifted total poverty upwards along the eastern part of FYR Macedonia. 14. Who are the poor in FYR Macedonia? The analysis of the Household Budget Survey for 2006 yields the following results: Poverty is an urban phenomenon, with a majority of the poor, especially the extreme poor, in secondary cities. It is also a Northern phenomenon. The Northeast, Polog and Greater Skopje region are home to more than one in two of the moderately poor and two in three extreme poor. Poverty affects larger households, even when controlling for economies of scale and age consumption profiles. Families with children are particularly affected by poverty, due to an increase in extreme poverty among households with small children. By the same token, persons in their prime working age (24-39 years) are more likely to be poor than other adults. Poverty rates for households with an elderly household head and many household members have increased. Low or no formal education remains a significant correlate of poverty. Poor children are also more absent from school, thus leading to a cementation of poverty status across generations. A job is not always an earning opportunity. Poverty rates are highest for households where the main breadwinner is out of work, but are also surprisingly high where household heads work in low income occupations such as crafts and trade services. Poverty retains an ethnic dimension in FYR Macedonia. 15. Who are the nonpoor? Persons with higher education degrees have a low probability of being poor. The likelihood of poverty decreases for this group if the persons are employed, live in more affluent parts of the country, and have few if any young dependents. INCOME OPPORTUNITIES AND POVERTY 16. Are falling consumption levels due to lower labor income or lower transfers? The changes in the sources and levels of income are key variables for understanding the linkage among economic growth, labor markets, and poverty developments. In the case of FYR Macedonia, the poor's consumption per capita has fallen consistently since 2002 which is a reflection of either lower labor income or lower transfers.7 Lower labor income could be the result of unemployment, decline in hours worked, or negative wage growth. Lower transfers could result from a decline in private transfers (either from relatives/NGOs at home or abroad) or a decline in safety net payments or coverage.8 17. During the 2002-2006 period, the poor have seen shrinking public transfers and a shift towards higher dependence on agricultural and informal income. The poor experienced a more significant transformation of their sources of income than the non-poor. The period in 2002-2004 saw rapid income growth from informal work (agricultural and non-agricultural) while public transfers fell. The period in 2004-2006 saw negative income growth for the poor. Income from formal work fell, and 7 The poor are unlikely to have savings to increase their consumption. Moreover, the information on household savings and other assets is limited, especially in terms of changes over time which is the relevant information for income generation. 8 Private transfers includes remittance income. xv public social security fell significantly. Private transfers increased, suggesting that they may fill some insurance function. Figure 2 Changes in Income and Employment Annual growth in the poor's income per capita by Job creation, by sector (in thousands) source, 2002-2004 and 2004-2006 (in %) 100 30 80 2002-2004 2004-2006 60 20 40 20 10 0 -20 0 -40 Agriculture Industry Services Pensions Transfers Transfers Formal Non- Other Agricultural Informal Non-Ag -10 Private Public Ag 2004-2005 2005-2006 Source: World Bank estimates based on Household Budget Surveys and Labor Force Surveys. 18. Economic growth appeared delinked from job creation. Construction and the manufacturing industry, together with trade and transport, accounted for some 54 percent of total economic growth between 2002 and 2006. The average contribution of agriculture had been modest (12 percent, or about 0.4 percentage points) and in fact negative in 2006. In terms of absolute numbers, the job situation improved in FYR Macedonia, but sectors with value added growth were not creating employment. The agricultural sector created the most jobs on a net basis, although its contribution to value added growth was minimal or negative. Thus, job creation was been connected with low or negative productivity growth and with low productivity sectors. 19. The new jobs created were mostly unpaid jobs for low or unskilled youth, and may have represented hidden unemployment and underemployment. Three fifths of all new jobs in 2004-2006 were taken by workers with primary education or less, and over one third were unpaid jobs in the primary sector and one fifth represented self-employment in the primary sector.9 These categories generally have the least job security and social protection, and the highest underemployment rates (that is, workers are underutilized). Thus, job creation did not result in poverty alleviation. In fact, the Southwest, Southeast, and Pelagonija regions, saw poverty numbers stagnate or increase. If anything, the relationship between job creation and poverty increase was positive in this period because the jobs the poor held were low wage or unpaid. 20. Consequently, the group of working poor increased. Poverty rates were higher among the unemployed or inactive than among the employed, and households with a higher share of employed were less poor than households where few or none were working. However, there was a shift towards higher poverty rates among the employed as well, where extreme poverty rates approached those of the inactive/unemployed. Between 2004 and 2006, the poor occupied one third of all new jobs. However, since many of these jobs were unpaid and informal (and hence without social insurance benefits), they did 9 Primary sector refers to agriculture and mining. xvi not contribute to poverty alleviation. Similarly, poverty rates for households with higher share of employed persons increased. Figure 3: Pension income was critical for poor receiving households, but social assistance was insufficient to make ends meet For receiving households: transfer % in consumption /a Remittances (% of GDP) 90 4.5 4.2 80 4.0 3.9 80 75 Extreme 4.0 3.8 Moderate 3.5 70 3.5 Non poor 60 56 3.0 2.8 53 50 40 2.5 2.3 2.1 40 32 33 32 2.0 30 27 21 1.5 20 10 8 1.0 10 0 0.5 2002 2006 2002 2006 0.0 Pensions Soc. Assist. 2000 2001 2002 2003 2004 2005 2006 2007 a/ This graph shows that share of reported transfers in total consumption expenditures of recipients. Thus, in 2006, the sum of all pensions received by the extreme poor divided by the sum of all consumption expenditures of extreme poor living in households receiving pension income of the extreme poor was 80 percent. Source: World Bank estimates based on HBS. 21. The social protection system has a wide reach and its social welfare components could benefit from improved targeting to the poor. The social protection system in FYR Macedonia is comprised of contributory social insurance programs and non-contributory social benefits which include income support benefits, child allowances and disability benefits. Nearly 90 percent of total social protection benefits, which include pensions, are allocated to households above the poverty line. The social insurance programs compensate those who paid into the social insurance system as they held jobs in the formal sector. The non-contributory social benefits programs are designed to help those who are in need of state aid and comprise instruments to reach disadvantaged groups such as able body individuals who are unable to provide financial security for themselves, invalids, students in need of financial aid, and families with young children. 22. Though social benefits programs performed well compared to international averages, key public social transfer programs continued to see significant leakage of benefits to the non-poor. Social programs exhibited some pro-poor targeting; yet, there was substantial leakage to the more affluent households. Specifically, about 40 percent of total transfers reached the poor, implying that between half and two-thirds of the total funds allocated to this program did not reach the intended beneficiary group which is partly due also to the fact that the some programs are rights-based (categorical) rather than means (or income)-tested. This being said, FYR Macedonia's performance is among the top four out of 26 in the Central and Eastern Europe and Central Asia region. 23. Pensions provided better coverage and a higher proportion of income for the poor, while the targeting of social assistance and other welfare benefits deteriorated since 2002. The share of poor households receiving social assistance benefits declined. In 2002, half of the extremely poor households and one in three moderately poor households received social assistance. By 2006, only one in three extremely poor households and one in five moderately poor households received social assistance benefits. In contrast, though the pension system is not an anti-poverty program, its coverage of the poor was much higher. Pension coverage of the extreme poor increased from 17 percent to about 34 percent (to reach the average for FYR Macedonia) while the coverage of the moderate poor remained constant at xvii about 30 percent. Furthermore, the extreme and moderate poor who receive pensions increased in 2002- 2006 by 5 and 3 percentage points respectively to reach 80 percent and 56 percent of consumption per capita. Figure 4: Job creation and destruction Jobs created and destroyed through Net job creation and efficiency expansion, contraction, entry and exit. by sector in 2002-2006 Thousands (average per year). Source: World Bank estimates based on Firm Registry Data. 24. Private transfers compensated for some of the fall in public transfers. Remittance income took on a more important role for the poor over time. In 2002, less than one percent of the extremely poor population lived in a household which received remittances from abroad. By 2006, this number had increased to over 8 percent. For the moderately poor, the coverage increased from 1 to 5 percent. Since the share of non-poor remained stagnant, a higher share of the poor than the non-poor was covered by some form of remittance income. Notwithstanding this increase, less than one in ten of the poor could rely on remittance income (at least as recorded in the HBS) to support their consumption. JOB DYNAMICS IN FYR MACEDONIA 25. In this report, we use firm level data from the government's Enterprise Registry to examine shifts in formal job creation ­ and opportunities for the population.10 The analysis of firm enterprise data provides a window into changes occurring in the economy, reflective of the impact of public policies as well as international developments. Specifically, changes over time in job creation and job destruction rates and patterns can be particularly informative. Limitations do exist as it is not possible to identify 10 There is no standard definition of a "formal worker". However, one reasonable definition is an employee with a formal work contract. A reasonable proxy in FYR Macedonia for the presence of a contract is whether the worker receives "net pay" indicating that taxes are deducted from his or her paycheck. We use this proxy since the Labor Force Survey does not explicitly ask the worker whether he or she has a contract, but does ask about net pay. According to the Labor Force Survey (2006) 62 percent of all employed are employees receiving net pay. This proportion represents the upper bound of the proportion of formal sector workers as it excludes unpaid labor, self- employed, employers, and those employees who either did not or "do not know" if they received net wages. xviii whether the jobs being created or destroyed are those held by the working poor. However, the analysis of enterprise data does provide us with the third part of the story of why poverty remained stagnant despite the increase in employment in FYR Macedonia. 26. Low net formal job creation numbers mask high rates of job creation and job destruction at the firm level. Between 2002 and 2004, net job creation was negative, and although job creation resumed between 2004 and 2006, employment did not recover to 2002 levels. Between 2002 and 2004, about 27 percent of all formal jobs were destroyed each year, but only 20 percent of jobs recreated ­ and hence, formal employment appeared to be shrinking on a net basis. But while job destruction remained just as high in 2004-2006, job creation increased to 31 percent. In other words, in any one year, 3 in 5 jobs were "reallocated" (i.e., destroyed or created) across the economy. For each job gained in net terms, more than thirteen jobs had to be reallocated across the economy. Thus, one important issue in future labor market work is to measure the vulnerability to unemployment or informal sector employment for persons holding formal sector jobs. 27. Firm survival rates improved. Jobs can be created and destroyed through the creation or liquidation of firms, or through the expansion or contraction of existing firms who do not go out of business. In both 2002 and 2004, more jobs were created through start-ups than through firm expansion. But in the latter period of 2004 to 2006, firm survival appears to have improved, as the number of jobs destroyed through the exit of firms fell. More jobs were in fact destroyed in continuing firms than in liquidated firms. Little is known either about severance arrangements or the impact on households of this high rate of job turnover ­ or whether those who lost their jobs were the ones who acquired the new ones. 28. Reallocation needs remained very important in FYR Macedonia. The international experience is that as the restructuring process matures, entry and exit rates slow down. In 2004-2006, about 24 percent of firms were created and about 14 percent liquidated (or exited) as a proportion of all firms ­ implying a reallocation rate of 38 percent. FYR Macedonia's level of churning is thus high by comparison both with other transition countries and OECD countries. For example, about 15 percent of all firms either enter or exist in any given year in the United States, 22 percent for Romania, and 26 percent for Latvia. This may be an indication that FYR Macedonia is in an earlier stage of transition than some of the other Eastern European countries. 29. Net job creation was only positive in the services sector. The inter-sectoral shifts in FYR Macedonia's job structure are consistent with an economic transition path. Industry, which appears to be overrepresented in the administrative enterprise data, compared to the Labor Force Survey, saw high job destruction, with massive reallocation above what would be needed to just transfer jobs from one sector to the other, but job creation increased over time. The services sector experienced positive and increasing net job creation and was the sector with the most efficient job creation, in the sense that the excess reallocation is the lowest. 30. In spite of a general shift towards higher job creation, important sectors saw little or negative job growth. In the manufacturing sector, which accounts for some 37 percent of employment in the sample, there is a shift towards higher job creation and lower job destruction in 2004-2006 compared to the early 2000s. However, an exception to this is the textiles industry, which alone accounts for 14 percent of jobs. The textiles industry created virtually no formal jobs in the second period, while gross reallocation remained the same. Within the services sector, sub-sectors have similarly shifted towards higher job creation and lower job destruction. The one exception was transports and communications sector, where both job creation and job destruction increased. Again, this is one of the sub-sectors with the highest shares of employment (around 10 percent). Figure 5: A shift towards higher job creation­ except in major employing sectors xix Manufacturing and construction: Job creation and job Services sector: Job creation and job destruction rates, destruction rates, 2002-2004 and 2004-2006 2002-2004 and 2004-2006 0.60 0.50 2002-2004 0.45 2004-2006 0.50 0.40 Basic metals Gross Job Destruction Gross Job Destruction 0.35 0.40 0.30 Wholesale Food Construction 0.30 Textiles 0.25 Retail Trsp&Com Food Textiles 0.20 Retail Construction Trsp&Com Wholesale 0.20 0.15 Finance Basic metals Finance 0.10 0.10 0.05 0.00 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.00 0.10 0.20 0.30 0.40 0.50 Gross Job Creation Gross Job Creation Note: The legend is the same for both charts. Source: World Bank estimates based on Firm Registry Data. 31. Only firms with negative productivity growth were creating jobs towards the end of the period under analysis. Positive net job creation between 2004 and 2006 was driven entirely by low wage jobs. From a poverty perspective, more employment in firms with low productivity levels is likely to mean more job opportunities for the poor. At the same time, without a concomitant boost in output, there was no basis for real wage growth. But in fact, only firms with negative labor productivity growth (especially the second quintile ­ firms between the 20 and 40 percentiles of the productivity distribution) were creating jobs on a net basis. 32. In spite of some dynamics, there did not seem to be a shift towards better paying or more skilled jobs. The job creation process became more efficient over the period 2002-2006. However, since new jobs did not appear to be connected with labor productivity growth, their sustainability would be threatened in a competitive environment, and they could not provide income growth for their workers. The story emerging from the firm data is thus consistent with the poverty stagnation, the increase in working poor, and the disconnect between poverty and job creation. THE WAY FORWARD 33. The global economic crisis poses huge challenges for Macedonia, including to effectively protect the poor. The Government of FYR Macedonia has the opportunity to transform the dynamics of economic growth and poverty reduction from one where the poor did not share in growth to one that utilizes the poor's labor assets. The current global crisis may delay measurable benefits of adopting policies that would have yielded returns in the near term. Nevertheless, these policies remain important to pursue because the cost of inaction could potentially keep a significant share of the population mired in poverty and delay FYR Macedonia's overall social and economic development. Moreover, the greater danger is that it would potentially slow down convergence with the rest of the EU in the long-term and even the New Member States of the EU in the medium-term. 34. The years 2002-2006 represents a difficult period in FYR Macedonia's development. It was characterized by moderate economic growth and low-quality job creation. The economy grew at rates below those experienced in regional comparator countries. Many of the jobs that were created in FYR Macedonia implied underemployment, precarious work, and low and even falling pay ­ rather than economic dynamism. Moreover, social welfare programs (as distinct from social insurance) were not adequately targeting the poorest in society who were unable to meet their basic needs. xx 35. The government is pursuing a broad agenda and must do so in the face of both short-term and long-term challenges. However, from the perspective of reducing poverty, there are certain priority areas, that if tackled, could help to better alleviate poverty. 36. First, given the strong relationship between economic growth and poverty reduction globally, policies which accelerate growth to above 5 percent per annum have the best chance of reducing poverty. Having policies which enhance shared growth is the most effective ­ and most cost efficient ­ instrument for raising the greatest number of people out of the poverty. The recent World Bank report, FYR Macedonia ­ Moving to Faster and More Inclusive Growth: A Country Economic Memorandum, discusses the benefits of increasing annual average growth by over two percentage points by implementing a series of structural reforms. Continuation of reforms in the judiciary, contract enforcement and land titling is necessary to further strengthen creditor and property rights. Reforms that would promote and strengthen competition (i.e. further trade and market liberalization, stronger regulatory bodies etc.) are also a key requirement for stronger economic growth. Public administration reform are needed to reduced the administrative costs of the public sector and promote investments. Moreover, World Bank analytical work on labor markets pointed out that the responsiveness of employment to growth was low in part because of the high cost of hiring labor due to the excessive social contributions.11 At the same time, the mix of macroeconomic policies should support macroeconomic stability and sustainability of fiscal and external balances. 37. Second, the government should continue its focus on enhancing competitiveness and improving the business environment despite the challenges posed by the current global crisis. Areas of importance for private sector development are easing regulations for closing a business, especially given the economic contraction and permits and licensing; infrastructure development which not only yields medium- to long-term benefits but can also generate employment in the context of the economic crisis; and ensure that regulatory procedures are simplified. A second important angle to pursue is that greater exploitation of the enterprise data could provide greater insights into which firm characteristics are most strongly associated with high net job creation rates. Through better understanding to what extent job creation and destruction in firms is driven by wages, productivity levels, and industry or service sector, the government could identify more targeted and specific policies to accelerate job growth. Thus, it would be useful to complement the policy advice from the CEM with a better understanding of the micro determinants of growth. 38. Third, policies which improve the coverage and quality of the education system and enhance the population's marketable skills, can reduce poverty in the medium- to long-term. As discussed in this report, jobs alone are insufficient to decrease poverty, since there must be productivity growth to sustainably raise wages over time, especially at the lower end of the income distribution. While access appears to be decent, quality and affordability remain a concern. The government has already taken significant steps to improve education, e.g., making secondary education compulsory and introducing a Conditional Cash Transfer Program (supported by the World Bank) to encourage school attendance among recipient households. Implementation of a Professional and Carrier Development for Teachers should produce better trained and more motivated teachers. In addition to increasing instructional time at early levels of education, wider use of pre-primary education needs to be promoted. To maximize the benefits of the recently purchased IT equipment that authorities needs to devise content and ensure quality teacher training. The government must remain vigilant to set national policies (such as minimum service standards) and ensure these are complied with at the subnational levels given the decentralization that is occurring in the education system. The assessment and quality assurance systems at all levels of education 11 As of 2009, the Government has started to reform the social contributions system by reducing the tax wedge and the administrative burden on enterprises. xxi need to be well-established. Though education reforms both in school and colleges or universities are being introduced, these will help new labor force entrants. Given the large stock of working aged adults, a significant share who have little or no education and are also poor, the government may wish to accelerate establishment of the adult education mechanisms, anticipated in the law for adult education. At the same time, support to vulnerable groups and students with special needs (provision of text books, expanded use of CCT etc.) should ensure full enrollment and avoid drop-outs. Similarly, budget financing of core adult education programs (based on regular labor market needs assessments) can help adult poor without skills. 39. Fourth, improving the social welfare system, especially the social assistance program, can help to reduce poverty by expanding coverage of the poor while reducing leakage to the affluent ­ all within the same budget envelope. The Government of Macedonia has already been bold in initiating reform of the safety net. Though instruments exist for transferring assistance to the poor, the system needs to be upgraded to standardize the application of the eligibility criteria, improve targeting, and reduce leakage to the non-poor. The government is tackling these issues and it is important to ensure that the system is regularly evaluated and monitored to measure improvements and remaining areas of weakness. The importance of having an effective social welfare system is particularly high given the above average level of under- and unemployment in FYR Macedonia combined with slow job creation rates. The recently revised Law on Social Protection opens up possibilities for significant reforms. Still, adoption of a by-law that would define "countable income" to include all transfers is a critical element for its effective implementation. Further policy reforms may also be needed such as expansion of the use of CCT to other sectors (including labor market activation). The institutional capacities should be strengthened as well: the ongoing activities for the establishment of the Centralized Database on Programs and Beneficiaries are important for this but the capacity of implementing agencies needs to be upgraded also. The workload of the Centers for Social Works is rather big while the internal capacity of the Ministry of Labor and Social Policy to assess the effectiveness of the programs remains limited. Closely related to the topic of social safety net, the large informal economy resulted in relatively low coverage of the pension system. As a result, over the medium- to long-run, a growing number of citizens will not meet the minimum requirements for receiving a pension and would have to rely on the social safety nets to provide their subsistence and the system should initiate activities to accommodate this. 40. Fifth, in order to better monitor progress in poverty reduction, FYR Macedonia should adopt an absolute poverty line and strengthen its household survey instrument. The government currently uses a relative poverty line to monitor poverty; however, this instrument should be complemented by a measure of absolute poverty. Annual changes in relative poverty do not yield sufficient information for the government that could inform its policies. Moreover, there are some key issues that need to be addressed in the design of the household survey which are modest in scope but could yield large benefits in terms of quality and relevance. xxii CHAPTER 1. LIVING STANDARDS IN FYR MACEDONIA INTRODUCTION 1.1 The success of FYR Macedonia's economic reforms will not be judged by how efficient or productive the economy is but whether the citizens of FYR Macedonia achieve greater and sustained prosperity. Though economic reforms should indeed lead to higher welfare, they still need to be closely monitored. Sometimes the benefits of reform are insufficient and fail to reach a large share of the population or the poor. Much depends upon the pace of growth, appropriate policy choices, the quality of governance, and how well regulations and programs are implemented by public and private institutions. 1.2 In depth work on understanding FYR Macedonia's economic progress and impediments to growth has been carried out by many institutions including the Government, the World Bank, and the European Union. However, much of the published information is not current nor are important questions such as whether reforms have translated into improved living conditions and who among the population has benefited. The last stocktaking exercise on living standard developments was completed jointly by the World Bank and the State Statistical Office in 2005 based upon data from 2002 ­ 2003 and prior to that in the mid 1990s. Since then, there have been significant policy shifts as well as new developments, especially with FYR Macedonia's bid to join the European Union. 1.3 This document provides a perspective different from the previous two World Bank documents and the Government's own publications on poverty. Though this assessment includes information on the level and trends in poverty as well as the profile of the poor, it is more ambitious than previous work. This report aims to fill the information gap on changes in the welfare of FYR Macedonia's citizens. A key question is whether economic growth in FYR Macedonia has translated into higher incomes for the population and better living conditions. This report analyzes whether the changes in poverty are explained by and consistent with changes observed in labor markets and the enterprise sector; thus, providing a better understanding of the channels through which the economic welfare of low income groups is affected. This is possible, since unlike most poverty assessments which rely primarily on household data, this one exploits labor market surveys and administrative enterprise data as well. 1.4 The report is organized as follows. The remainder of this chapter looks at broad measures of development in FYR Macedonia in relation to other countries, including the progress towards the Millennium Development Goals. The second chapter focuses on the key poverty trends between 2002 and 2006 and stays within the realm of the household budget data to separate out the effect of growth and distribution on poverty developments. The third chapter provides a detailed poverty profile for 2006, highlighting any significant changes since 2002. The fourth chapter addresses poverty stagnation from the perspective of growth, labor markets and social protection policy. The fifth chapter looks at job dynamics and the characteristics of firm birth and death in FYR Macedonia, linking job opportunities, labor productivity, and the potential for long term income growth. The final chapter summarizes a few conclusions based on the analysis. HAS HIGHER GROWTH HELPED FYR MACEDONIA REDUCE POVERTY? 1.5 Poverty has fallen in Central and Eastern Europe and Central Asia since the early 2000s. New estimates from the World Bank suggest that the new millennium has brought significant reductions in poverty in ECA countries. Between 2002 and 2005, an estimated 23 million people escaped poverty, measured by expenditures at $2 per day,12 thus reducing the number of poor by over one-quarter in a few years only - an accelerating rate of poverty reduction compared to the previous period. Figure 1.1: Poverty Reductions in the ECA Region Million persons living on less than $2 /day /a 120 101 100 84 80 61 60 40 20 0 1999 2000 2003 a/ These indicators have been derived from the analysis of nationally representative household surveys. Source: World Bank POVCAL data. 1.6 Economic growth and job creation has increased in FYR Macedonia. The Macedonian economy has begun to show more dynamism since the last Poverty Assessment was prepared. Macroeconomic stability has been maintained, and economic reforms have advanced in several areas, with FYR Macedonia coming out as one of the top reformers in the World Bank's Doing Business Rankings, although according to other indicators, for example, the EBRD's transition rankings, puts FYR Macedonia behind other regional comparators. The economy, which in 2002 and 2003 was still reeling from the 2001 instabilities, recovered solidly afterwards and economic growth increased to about 4 percent. The labor market also appeared to improve: between 2004 and 2006 (the years for which there is comparable labor market data13) employment rates, i.e., the share of working age population that was employed, increased and unemployment rates fell. 1.7 However, the quality of jobs has deteriorated. Many of the jobs created did not really provide the basis for income growth, which would have helped reduced poverty, if they were filled by persons from low income households. In fact, primary sector14 and unpaid work increased, and most sectors with rapid job creation saw weak value added growth and vice versa, indicating that output and job creation at large remain decoupled (Figure 1.2). While the poor may have occupied these jobs, they are not likely to have resulted in much income growth. 12 In order to make international comparisons of poverty, the World Bank uses three poverty lines -- $1.25, $2.5, and $5 per capita per day. The lowest poverty line is the average of national poverty lines of the poorest 10-20 countries. The other two poverty lines are a multiple of the $1.25 per day. 13 From 2004, the LFS is administered continuously during the year and data are reported on quarterly basis. Data prior to 2004 was collected once a year, in April of each year. 14 The "primary sector" refers to the agriculture and mining sectors. 2 Figure 1.2: Job-Less Growth, Growth-Less Jobs GDP growth Value added growth vs. job growth 6.0 30 25 Tourism 4.0 20 Growthless jobs Mining Empl growth 2.0 15 Agr 10 Construction 0.0 5 Finance 2000 2001 2002 2003 2004 2005 2006 2007 Manuf Edu, Health 0 Utilities -2.0 Trade Transport Public Adm Other -5 Jobless growth -4.0 -10 0 5 10 15 20 25 30 -6.0 VA growth Source: Estimates based on World Bank data and LFS 2004-2006. 1.8 A poverty update is thus both warranted and timely. The continued disparity between economic growth and employment expansion suggests that absolute poverty remains a significant problem in FYR Macedonia. In spite of higher growth rates and a relatively widely distributed stock of human capital, a large share of the population remains locked out of real income opportunities. This poverty report analyses poverty trends over 2002-2006 and also accompanies and complements recent World Bank work on the Macedonian economy and its labor markets. 1.9 The report draws on three distinct sources of data to focus on the poverty/labor market/growth nexus. The remainder of the report centers on three different sources of data. The 2002- 2006 Household Budget Surveys (HBS), which are the source of the bulk of the analysis, provide information on poverty and living conditions more broadly. The 2004-2006 Labor Force Surveys gives detailed insights into the poverty and vulnerability as these phenomena relate to labor market opportunities. And finally, an enterprise database illustrates the links between job dynamics, productivity and income opportunities. 1.10 The analysis emerging from this report shows that during 2002-2006 the lack of improvement in unemployment and poverty are rooted in moderate growth, inequitable distribution of that growth, and inadequate safety nets as seen in Figure 1.3. The report shows that in spite of moderate growth and some improvements in human development indicators, consumption poverty, i.e., the share of people who do not consume enough to fill their basic needs, stagnated between 2002 and 2006, while the number of extreme poor increased. Consumption growth was moderate on average, but largely negative for the poorer households. Spatial inequality increased, with Eastern regions seeing increases in poverty while Skopje saw significant reductions. There was insufficient job creation in growth sectors, such as services, as the labor force shifted towards more jobs in the primary sector and unpaid jobs, with little or no income growth potential. This is consistent with the firm data which shows that only firms with negative labor productivity growth in fact created jobs. And safety nets have not stepped in to reverse these effects: even excluding the regressive distribution of pensions, the wealthiest two fifths of the population receive about 20 percent of total social assistance spending (Figure 1.3). 3 Figure 1.3: Three Data Sources: One Message Household Surveys: Increasing inequality, regressive consumption growth, spatial inequality, more working poor and inadequate safety nets No Poverty Reduction in FYR Macedonia Labor Force Surveys: Firm data: Job creation but unpaid jobs, Firms with negative primary sector, disconnected labor productivity from output growth. growth create jobs LIVING STANDARDS IN FYR MACEDONIA 1.11 While improving during 2002-2006, FYR Macedonia's economic performance did not catch up with other Eastern and Central European countries. Although growth increased, it was still lower than most comparator countries in Eastern and Central Europe (World Bank, 2008), meaning that FYR Macedonia was, in fact, diverging from other EU countries. Job markets are also more lethargic: FYR Macedonia had the lowest employment rates in Europe. Although labor market indicators improved as job creation picked up, the nominal increase in employment was coupled with poor productivity developments and tended to reflect different forms of underemployment, much of it in the form of unpaid work (Figure 1.4). Figure 1.4: Output Growth and Job Creation Remain Low by Regional Standards Growth rates Employment rates 8.0 80 7.0 2002-2004 70 6.0 2004-2006 60 5.0 50 4.0 40 3.0 30 2.0 20 1.0 10 0.0 0 EU10 Serbia Bosnia&Herz Albania Croatia FYR MK EU15 Hungary EU 15 Estonia Slovenia Latvia Czech Rep Lithuania Slovakia Romania Bulgaria Poland Albania Croatia Serbia FYR MK Source: Eurostat, World Bank 2007abc, World Bank, 2008. EU10 refers to the "new" EU member states (2004), namely Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic, and Slovenia. EU15 refers to countries which were EU members already in 2003. 4 Box 1.1: What do we mean by poverty? This report refers to several different measures of poverty. The World Bank uses either national "absolute" poverty lines or global international poverty lines ­ both which are based upon a basic needs approach. The State Statistical Office of FYR Macedonia uses relative and subjective poverty lines. When we use a particular poverty line, we are measuring well-being but the definition is not standardized. Below are definitions of the various definitions of poverty used: People living in extreme poverty are persons who do not have sufficient income or expenditures to acquire a minimum amount of calories (for FYR Macedonia the amount is 2100 calories) even if s/he only consumed food and nothing else (such as housing or clothes). The computation of extreme poverty uses a "basic needs" methodology as it establishes thresholds of material wellbeing. It is also an absolute amount in local currency which can be used to track poverty over time in a specific country. People living in total poverty (also referred to simply as poverty) are persons whose income or expenditures are below a minimal subsistence threshold ­ which includes food and non-food goods and services. Thus, people in poverty may be able to consume sufficient calories but do not have enough money to spend upon items considered necessities in their society which could include such items shelter, water, clothes, and school supplies. People living in total poverty include the extreme poor. Those persons who are poor but not living in extreme poverty are called the moderate poor. In order to measure poverty across countries, international poverty lines have been established since the same reference poverty line has to be applied, and expressed in a common unit across countries. For the purpose of global aggregation and cross-country comparison, the World Bank now uses reference lines set at $1.25 and $2 per day (2005 Purchasing Power Parity terms).1 The $1.25 poverty line was established by taking the average of national povety liens of 10-20 low income countries. Only by applying an international poverty line can we compare poverty across countries whether globally, regionally, or sub-regionally (e.g., in Central and Eastern Europe or the Western Balkans). The methodology for computing international poverty lines must correct for the fact that a dollar equivalent in FYR Macedonia does not always buy the same amount of goods and services in other countries such as Germany or Egypt. Thus, the poverty lines are corrected in order to obtain "purchasing parity" across countries. International poverty lines have their drawbacks since applying a threshold that may be relevant for one group of countries may not be relevant for tracking welfare in another group of countries.. Relative poverty lines are set arbitrarily usually as a proportion of some average income or expenditures. In FYR Macedonia, the relative poverty line is established at 70 percent of median household expenditures of the current year per equivalent adult (using the OECD scale to adjust for household size and composition). Thus, people with less than 70 percent of average expenditures per capita are considered to be living in poverty. The relative poverty line can be a good indicator of how high disparities are at a particular point in time by the standards of any particular country. But it has two main drawbacks. First, the relative poverty line tells us nothing about whether people can fulfill certain necessary material needs such as food. Second, it cannot be used to track improvements in welfare overtime. Thus, if relative poverty rates increase from one year to the next, this tells us nothing other than inequality may be growing. Furthermore, the choice of using a threshold of 70 percent is arbitrary. The EU uses 60 percent instead. Relative poverty lines are used when measuring inequality is important in a society where no one lacks for food or other material necessities. Subjective poverty is based upon peoples' views of what is a reasonable income level to escape poverty. This is estimated based upon people's subjective perceptions of what a minimum income is to reach an acceptable level of well-being in their particular context. A typical survey question to measure subjective poverty is "what income level do you consider to be absolutely necessary to make ends meet?" Thus, if making ends meet could include paying for expensive medicines for an older member of the family and private tutoring for children; or not, if the household in question does not have these additional obligations. 1/Previously the poverty lines were $1.08 and $2.15 as the cost-of-living was based on the best data available at that time. However, in 2005, the poverty lines were updated since the cost-of-living was higher than expected. Now the poverty lines are $1.25 and $2.5. Sometimes these two respective poverty lines are shortened to $1 and $2 per day. 5 1.12 FYR Macedonia's human development indicators are in line with its income levels, but below those of some neighboring countries. Overall, it is clear that FYR Macedonia, over time, has made strong achievements in several areas of human development. General health indicators have improved since the 1990s. Children and youth broadly have access to education, and literacy is all but universal among the young. Virtually all of the urban population have water supply in their house; mobile telephony is widely available. The UNDP's human development index ranking puts FYR Macedonia among countries with high human development, and higher, in fact, than would be expected given its income levels15. However, it remains near the bottom of the league of countries in Central and Eastern Europe (UNDP, 2008) though above Albania, Bosnia and Herzegovina, and Moldova. Table 1.1: FYR Macedonia Performs Quite Well on Selected Well Being Indicators mid 1990s 2006 FYR Macedonia FYR Macedonia West Balkan /a EU7 /b Middle Income /c Human Development Index n.a. 0.808 0.823 0.859 0.776 Years of life expectancy at birth 72 74 75 73 71 Infant mortality per 1,000 live births 23 15 10 9 26 Undernourished persons (% of population) 15 5 7 4 10 Physicians (per 1,000 people) 2 3 2 3 2 Literacy among youth (%) 99 99 100 98 97 Net primary school enrollment (%) 93 92 94 91 93 Persistence to last grade of primary 97 98 90 97 n.a. (% of cohort) Population covered by mobile telephony %) 70 99 98 99 n.a. a/ West Balkans refers to Albania, Bosnia and Herzegovina, Croatia, Montenegro and Serbia. b/ EU7 refers to Bulgaria, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic. c/ Data for Middle income countries is for 2005 Source: UNDP, World Bank. 1.13 Only about 3 percent of the population lived on less than PPP-corrected 2 dollars per day in 2003.16 On the basis of international poverty lines, which allow for comparisons across countries, FYR Macedonia has not displayed high levels of deep poverty in recent years. In 2003, only about 3 percent of the population lived with less than 2 PPP adjusted dollars per day, half the level of neighboring Albania. At higher poverty lines (5 dollars per day) poverty rates increase significantly, but have nonetheless fallen since 1998. However, it is also noteworthy that between 1998 and 2003, the share of very poor changed only marginally. Extreme poverty in FYR Macedonia thus appears to be "sticky" and unresponsive to economy- wide conditions ­ something which is borne out in the national poverty analysis provided in Chapter 2. 1.14 On several of the Millennium Development Goals, FYR Macedonia has made significant progress, also compared to its neighbors. Life expectancy continues to increase, malnutrition is disappearing and almost all children finish primary school. Women have school enrolment rates comparable with or higher than those of men, especially at higher levels of education, although women are still more vulnerable in the labor market. The limited evidence available on disease prevention and control suggests that tuberculosis and HIV have been relatively well-contained in FYR Macedonia. However, small children and infants lives are still at a higher risk than in other southern European countries, and unlike in neighboring countries, there is no notable improvement since the early 2000s (Figure 1.5). 15 Based on a regression of HDI on a constant and the log of GDP per capita in PPP terms. 16 PPP refers to "purchasing power parity". 6 Table 1.2: International Poverty Comparisons Suggest Low but Resistant Poverty /a 2.0 dollars per day 5 dollars per day Headcount Gap Severity Headcount Gap Severity 1998 FYR Macedonia 3.5 1.3 1.05 39.5 11.8 5.1 2003 FYR Macedonia 3.2 0.6 0.23 31.1 10.2 4.5 2004 Bosnia & Herz 0.7 0.2 0.10 17.2 4.3 1.6 2003 Bulgaria 2.4 0.8 0.68 36.1 9.9 3.9 2005 Romania 3.4 0.9 0.43 46.5 14.1 6.0 2005 Albania 7.8 1.4 0.44 59.2 21.0 9.8 a/ The international poverty lines of $2 and $5 are applied correcting for purchasing power. Source: World Bank POVCAL database. 1.15 Beyond averages, spatial inequalities in opportunities and living standards are significant in FYR Macedonia. Although averages suggest that FYR Macedonia is doing quite well compared to other countries, there are significant disparities at a sub-national/regional level to take into account. Female employment rates in Polog and North East are less than half of the national average. Output per capita in Skopje is four times as high as in the Northeast or Polog.17 A study of socio-economic indicators for different municipalities in FYR Macedonia as shown in Table 1.3 indicates that location has significant effects on a wide range of social indicators, including infant mortality, illiteracy, immunizations, and access to clean water among others (UNDP, 2004). Table 1.3: Significant Spatial Inequality in FYR Macedonia, 2004 /a (in percent) Indicator Best Municipality Worst Municipality Female Illiteracy Rates 1 34 Net Enrolment Rates (primary) 109 50 Infant Mortality Rates (per 1000) 2 97 Clean Water (share of population with access) 100 71 a/ Using latest data available at the time of drafting the report. Source: UNDP, 2006 and State Statistical Office. 17 Note that regional output does not take into account internal migration (which would be measured by GNI): thus, estimates of sub-national output does not correct for when there is a divergence between persons' residences and their place of work or firm registration. 7 Figure 1.5: Macedonia has Made Progress on Several MDGs Goal 1: Eradicate extreme poverty and hunger Goal 2: Achieve universal primary education 10 100 Primary completion rate (%) 9 Malnutrition prevalence (%) 90 8 80 7 70 6 60 5 50 4 40 3 30 2 20 1 10 0 0 2000 2006 2000 2006 2006 2006 2000 2006 2000 2006 2006 2006 FYR Macedonia SE EU7 Middle FYR Macedonia SE EU7 Middle Income Income Goal 3: Promote gender equality and empower women Goal 4: Reduce child mortality 30 Proportion of seats held by women in 35 parliament (%) 25 30 25 Under five mortality 20 rate (per 1,000) 20 15 15 10 10 5 5 0 0 2000 2006 2000 2006 2006 2006 2000 2006 2000 2006 2006 2006 FYR Macedonia SE EU7 Middle FYR Macedonia SE EU7 Middle Income Income Goal 5: Improve maternal health Goal 6: Combat HIV/AIDS, malaria, and other diseases 35 120 30 Adolescent fertility rate (%) 100 25 80 20 Incidence of Tuberculosis 60 (per 100,000) 15 40 10 5 20 0 0 2000 2006 2000 2006 2006 2006 2000 2006 2000 2006 2006 2006 FYR Macedonia SE EU7 Middle FYR Macedonia SE EU7 Middle Income Income Goal 7: Ensure environmental sustainability Goal 8: Develop a global partnership for development 1.0 140 Aid per capita CO2 emissions 0.9 (KG per PPP$GDP) 120 0.8 0.7 100 0.6 80 0.5 0.4 60 0.3 40 0.2 20 0.1 0.0 0 2000 2006 2000 2006 2006 2006 2000 2006 2000 2006 2006 2006 FYR Macedonia SE EU7 Middle FYR Macedonia SE EU7 Middle Income Income Source: MDG website. 8 CHAPTER 2. SELECTIVE GROWTH AND THE POVERTY TRAP: POVERTY TRENDS, 2002-2006 INTRODUCTION 2.1 The goal of economic growth is to raise people's welfare in a sustained manner. Though poverty levels may move only incrementally year to year, over the medium-term, we would hope to see a perceptible fall in poverty rates and improvements in the living conditions of the population. Sometimes, economic reforms are introduced which though they may result in short-term disruptions, are expected to lead to not only a turnaround in growth but also an improvement in the population's welfare, especially of the lower income groups. 2.2 However, the reality differs from country to country, especially when economic growth is moderate at less than 5 percent per annum.18 In the Central and Eastern European and Central Asian countries, there has been no robust correlation between moderate growth (less than 5 percent per annum) and poverty reduction. In some countries, moderate growth led to falling poverty rates, sometimes sharply ­ indicating that there was "shared growth". In other countries, moderate and continued growth led to no change in poverty and even increasing poverty levels. However, when growth rises to above 5 percent per annum, poverty appears to fall across the board. 2.3 This chapter discusses the changes in poverty and inequality during 2002-2006 which were years that FYR Macedonia achieved moderate growth. There are three main areas of discussion: (1) changes in poverty, inequality and economic growth at the national level; (ii) changes in regional poverty and inequality; and (iii) changes in consumption of different income groups. The main findings of this chapter are as follows: · Economic growth did not result in poverty reduction during 2002-2006. · Overall poverty remained stagnant at 19 percent of the population between 2002 and 2006. · The share of the population living in extreme poverty rose in these years from 4.5 percent to 7.4 percent. · Inequality is increasing among the population, with residents of Skopje pulling ahead of other regions. 2.4 The analysis of this section is based upon data for 2002-2006 and does not include the results of 2007, though it is available at the State Statistical Office of Republic of Macedonia. At the time of drafting this report, the latest available year for which both the labor force survey results (all quarters) and the enterprise data was 2006. Consequently, in order to carry out the triangulation of data across three data sources, the latest survey results uses were from 2006. 18 See Growth, Poverty & Inequality in Eastern Europe and the Former Soviet Union, World Bank (2005). ESTIMATING POVERTY IN FYR MACEDONIA 2.5 The Government of FYR Macedonia uses a relative poverty line which does not measure absolute levels of welfare deprivation. Following EU practice, FYR Macedonia's official poverty line applied by the State Statistical Office (SSO) is a relative poverty line and poverty estimates are based upon the Household Budget Survey. The national poverty line is set at 70 percent of the median consumption level, and therefore measuring inequality in distribution only. However, FYR Macedonia's per capita GNI is at just above 10 percent of the GNI of the EU median (Greece) suggesting that not only distribution, but also absolute levels of welfare, are still likely to matter greatly. However, relative poverty lines do not permit a consistent tracking of poverty over time. The second poverty line used by the SSO is the subjective poverty line. The Household Budget Survey is also used to collect and present information on subjective poverty ­ how people feel about their income level and their capacity to meet necessary spending. However, this line is not useful for policy purposes especially in terms of tracking the number of poor over time and designing programs to help them. 2.6 FYR Macedonia still needs to monitor absolute welfare deprivations to ensure that policy interventions are appropriately targeted. A significant share of the population of FYR Macedonia still suffers from real deprivations at a basic level. As discussed in the 2005 poverty assessment and shown in the remainder of this report, FYR Macedonia remains a country where many people still do not reach a minimum threshold of wellbeing, and where the evolution of wellbeing therefore needs to be monitored over time to ensure timely, cost-efficient and successful poverty reducing interventions. Consequently, it is important for the Government of FYR Macedonia to adopt a poverty line which allows the tracking of the poor over time. 2.7 This report therefore relies on the cost of basic needs approach to establish such a threshold: the poverty line. The method to arrive at the poverty lines is presented in discussed in detail in the previous poverty assessment, FYR Macedonia: Poverty Assessment for 2002-2003, and an overview of some of the problems arising in the household budget data from the poverty monitoring perspective are discussed in Annex A. In what follows, we refer to two poverty lines: the extreme (or food) line and the total poverty line. The extreme poverty line represents the cost of reaching about 2100 calories a day per person; persons whose average daily consumption level (in Denars) falls under this level are thus not able to cover even their basic food needs. These are considered as living in extreme poverty, because they have to give up some of their food needs in order to cover other basic needs. The complete line represents the monetary value of food consumption to reach 2100 calories and the bare necessities of non-food consumption spending (shelter, clothes, health, etc.). People with consumption below this level are considered poor. The persons whose consumption lies between the food and total poverty line are considered moderately poor. 2.8 The daily poverty lines are presented below. The extreme poverty line makes up about half of the complete poverty line (56 percent). Table 2.1: Extreme and Total Poverty Lines 2002-2006 2002 2003 2004 2005 2006 Extreme (food) poverty line (Denars/day) 54.8 55.5 55.3 55.5 57.3 Total poverty line (Denars/day) 97.5 98.7 98.3 98.8 102.0 Memo: Deflators (CPI) 1.012 0.996 1.005 1.032 Source: Estimates based on HBS. 10 2.9 Poverty levels and trends differ significantly between poverty measures. Different approaches can lead to very different conclusions about both poverty levels and poverty trends (Table 2.2). The basic needs approach and subjective poverty result in the lowest level of poverty: in 2006, one in five persons in FYR Macedonia could not meet their basic needs, and one in five was not satisfied with her/his level of income (subjective poverty). One in four consumed less than 70 percent of the median (relative poverty). However, the basic needs per capita approach shows that absolute poverty has stagnated over time. The share of population below the relative poverty line has fallen, however, implying that inequality, using this specific measure, has been reduced. Likewise, the share of persons who are completely dissatisfied with their level of income has fallen, and quite dramatically. Has the situation, in fact, improved? Table 2.2: Inconsistent Poverty Comparisons 2002 2003 2004 2005 2006 Basic needs 19.1 19.2 18.5 20.4 19.0 Relative Poverty1 29.1 26.9 28.9 26.5 25.8 Subjective Poverty2 34.6 26.8 24.1 23.7 19.0 1. % of population below 60% of the median income. 2. Persons reporting that they were completely dissatisfied with their level of income. 2.10 Only the basic needs approach provides a consistent comparison of poverty developments over time. While the relative poverty line and the subjective poverty measures can be interesting complements to the basic needs approach for a specific point in time, they are unfit for poverty comparisons over time and also for policy purposes. If between two points in time, overall consumption growth is negative but slightly less negative for the bottom share of the distribution, relative poverty might fall (the distribution improves) but a person who previously had enough to eat and could afford health care no longer can (he has become poor). Conversely, a person's subjective impression of what is an appropriate level of income for his needs is given by his or her role models and the context in which s/he lives. 2.11 There are currently some unresolved quality issues in the household data for FYR Macedonia. As discussed in more detail in annex B, these weaknesses, including a short recall period, weighting issues and some abnormalities in consumption level data, affect all poverty measures and make poverty diagnosis and monitoring less reliable. POVERTY TRENDS 2002-2006 2.12 Poverty, as measured by the percentage of persons below the complete poverty line, has stagnated between 2002 and 2006. In 2006, poverty touched 19.0 percent of the population compared to 19.1 in 2002. As seen in Figure 2.1, the fall in poverty between 2005 and 2006 (from 20.4 to 19.0 percent of the population) thus represents more a return to status quo than a sustainable improvement in living conditions. 2.13 At the same time, the stagnation in total poverty is masking an increase in extreme poverty. At 7.4 percent in 2006, extreme poverty remains relatively low in FYR Macedonia ­ in other words, one in fifteen persons cannot meet their basic food needs. However, this situation is representing a worsening compared to 2002, when only one in twenty suffered from food poverty. And the fall in moderate poverty, from 14.6 percent in 2003 to 11.7 percent in 2006, is merely representing a shift of population from 11 moderately poor to extremely poor.19 The increase in food poverty is also more consistent and less volatile than moderate poverty, except in 2006, when it increased with a jump. As a result, in 2006, more than one in ten children lived in extreme poverty in FYR Macedonia. Figure 2.1: The Poverty Headcount Index Stagnated between 2002 and 2006 Percent of population below extreme and total poverty lines, 2002-2006 25.0 Food poverty Moderate poverty 20.4 20.0 19.1 19.2 19.0 18.5 15.0 15.3 11.7 14.6 15.0 13.4 10.0 5.0 7.4 4.5 4.3 5.1 5.1 0.0 2002 2003 2004 2005 2006 Source: Estimates based on HBS. 2.14 Poverty is also becoming deeper and more severe, underscoring a problem of more extreme poverty. The poverty gap index and poverty severity index have also increased over time, successively between 2002 and 2005, and with a discrete jump between 2005 and 2006 (Figure 2.2). The poverty gap index measures the distance of the average consumption expenditures of the poor from the poverty line while the poverty severity measures the same gap but gives greater weight to the poorest. In all, these numbers suggest that there has been a successive increase in poverty among the poorest in the population. In conclusion, on average, the poor are not better off since 2002, and some of them are actually worse off. Where are these changes coming from? Figure 2.2: Poverty is Becoming Deeper and More Severe as Well Poverty gap Severity 6.0 4.0 Food Moderate 5.2 Food Moderate 3.5 5.0 4.7 4.8 4.6 4.4 3.0 2.7 4.0 2.5 2.1 2.2 2.0 2.0 3.0 2.0 2.0 1.5 2.0 1.5 1.4 1.0 0.8 0.9 0.8 0.6 1.0 0.5 0.5 0.3 0.3 0.0 0.0 2002 2003 2004 2005 2006 2002 2003 2004 2005 2006 Source: Estimates based on HBS. 19 A fall in moderate poverty can represent either an improvement or a worsening in living conditions: the moderately poor shift out of poverty altogether, or they shift into extreme poverty. 12 2.15 Poorer households have seen shrinking real consumption levels. A closer look at the level of per capita consumption for the poorest 40 percent of the population confirms this story. The poorest 15 percent of the population have universally become poorer, as illustrated by a downwards shift in Figure 2.3 below. Vulnerability to poverty, measured by the amount of people whose consumption levels are just above the poverty line, has not increased, however. If the poverty line was set 50% higher (even above the international 5 dollar a day line), poverty would in fact have fallen since 2002, because of higher growth in the second consumption quintile. Poverty in FYR Macedonia remains an issue of core poverty. Figure 2.3: The Poorest have Seen Shrinking Real Consumption Levels Since 2002 2002 2004 2006 50% increase in pov line 50000 Annual consumption per capita Complete poverty line 25000 Food poverty line Poverty increases 0 0 10 20 30 40 Percent of population Source: Estimates based on HBS. GROWTH AND REDISTRIBUTION OF CONSUMPTION 2.16 What accounts for the lack of progress in poverty reduction in FYR Macedonia? A first step to understanding the poverty trends in FYR Macedonia is a more detailed analysis of household data. Have household consumption levels stagnated, in spite of moderate economic growth rates? Have the poor lost out compared to wealthier households? The decomposition of the sources of poverty changes points to a combination of low and volatile private consumption growth and an unfavorable distribution of that growth in the population. 2.17 Per capita consumption growth has been volatile and, according to the household surveys, negative in 2005 and in 2006. Per capita consumption growth, whether measured in national accounts or in the household budget survey mean, was negative in 2003 and bounced back in 2004 (at around six percent, compared to GDP growth of 4 percent).20 But in 2005 and 2006, there are significant differences between the two sources: the survey mean growth falls much short of that of national accounts which remains close to GDP growth. Indeed, in 2005 ­ the year in which poverty increases - household consumption in national accounts remains at a respectable 4 percent growth, while growth according to the survey was negative. The high volatility in the household survey consumption levels translates into a negligible average growth rate of around 0.14 percent for the period 2002-2006 (Figure 2.4). 20 There are several reason s why estimates of household consumption in national accounts differ from those of household surveys in both level and growth rates: measurement errors/noise, different coverage, and different deflators are among the chief explanations (see e.g., Ravallion, 2001). 13 2.18 Inequality has increased. Parallel to low, volatile growth, the distribution of consumption has worsened across the board, but mostly in the regions where poverty has increased and deepened. The Gini coefficient of consumption expenditures has increased from 0.38 in 2002 to 0.40 in 2006. By 2006, the distribution was more unequal in rural than in urban areas. Figure 2.4: Low and Volatile Growth together with Increasing Inequality has Stalled Poverty Reduction Growth: Mean per capita consumption GINI coefficient, 2002, 2004 and 2006.1/ (surveys, national accounts) and GDP growth. 8.0 0.42 PC NA PC HBS GDP NA 2002 2004 2006 0.41 0.408 6.0 0.403 0.400 0.40 0.395 4.0 0.389 GDP NA 0.39 0.380 PC NA 0.377 0.379 2.0 0.38 0.370 PC HBS 0.37 0.0 2002 2003 2004 2005 2006 2007 Av 2003- 0.36 2006 -2.0 0.35 All Macedonia Urban Rural -4.0 Source: Estimates based on HBS. 2.19 Growth has not benefitted the poor. The poor's share in total consumption expenditures also remains very limited and has in fact fallen since 2002, and the fall has been sharpest for the very poorest. The poorest ten percent of the population now account for only 2 percent of total consumption in FYR Macedonia, reflecting the fact that the richer segments of the population have seen much higher consumption growth than the poorer. The only exception to this rule is the year 2005 when consumption growth was negative across the board. Those whose consumption levels fall in the poorest quintile (roughly equivalent to those below the poverty line) have consistently seen their consumption levels fall since 2002. This adverse growth pattern may explain the increase in food poverty (Figure 2.5). Figure 2.5: Consumption Growth has been Anti-Poor Share of consumption of the three poorest deciles Growth in consumption per consumption quintile 12% 10 Poorest 1 2 3 8 2nd 10% 6 3rd 4.9% 5.0% 4 4th 8% 4.8% 4.7% 4.7% 2 Richest 6% 0 -2 3.8% 3.8% 4% 3.7% 3.6% 3.5% -4 -6 2% 2.4% 2.5% 2.1% 2.2% 2.0% -8 0% -10 2002 2003 2004 2005 2006 2002-03 2003-04 2004-05 2005-06 Source: Estimates based on HBS. SPATIAL POVERTY DIVERGENCE 2.20 The past few years have seen regional differences in wellbeing intensify, in two dimensions. First, the situation in Skopje has overall improved compared to the rest of the country. Second, the northeastern part of the country has fallen behind other regions. 14 2.21 Skopje is pulling away from the rest of the country. Total poverty fell in the city of Skopje, stagnated in rural areas, and increased in secondary cities (Figure 2.6). Thus, Skopje has seen moderate poverty fall nearly by half over time, from 14 to 8 percent of its population, with a major reduction between 2005 and 2006. In 2002, one in four poor Macedonian persons lived in Skopje; in 2006, less than one in six. At the same time, the share of extreme poor in the population did not change, reaching 4.2 in 2006 compared to 4.5 in 2002. In secondary cities, moderate poverty fell from 13.7 to 11.8 percent. But this reflected a shift of moderately poor into extreme poverty, which more than doubled, from 4.1 to 9.5 percent of the population. A similar though less extreme pattern occurred in rural areas: moderate poverty fell from 15.9 to 14.0 percent of the population while extreme poverty increased from 4.8 to 7.4 percent of the population. As a result of these changes, by 2006, food poverty affected a higher share of residents in secondary cities than in rural areas ­ food poverty was even more prevalent there than moderate poverty in Skopje. In net terms, 28,000 poor persons shifted from Skopje to other urban areas between 2002 and 2006. Figure 2.6: Skopje is Diverging from the Rest; The Eastern Regions Account For the Overall Increase in Poverty in FYR Macedonia Moderate and extreme poverty in rural areas, Skopje, Share of total poor and other urban areas 18 Skopje Other urban Rural Skopje Other Urban Rural 16 159,000 poor 161,000 poor 100% 14 Moderate poverty 90% 12 80% 41.1% 41.4% 10 70% 60% 8 50% 6 Food poverty 40% 36.0% 42.9% 4 30% 20% 28,000 poor 2 10% 22.9% 15.7% 0 0% 2002 2003 2004 2005 2006 2002 2006 Source: Estimates based on HBS. 2.22 And regional differences are striking. Figure 2.7: Regional Divergence Data disaggregated by administrative region is Food, Moderate and Total Poverty Index, 2004-2006 only available after 2004. However, these data 45 suggest that two of the poorest regions, notably Significant 40 increase in Food Moderate the Northeast and Vardar, as well as the Eastern 35 food poverty region, have seen sharp increases in food poverty 30 between 2004 and 2006 ( 25 Figure 2.7). The increase in food poverty has 20 shifted total poverty upwards along the eastern 15 part of FYR Macedonia (Eastern, Northeastern 10 and Southeastern regions). It is notable that only 5 the Skopje region has seen a significant reduction 0 in poverty.21 Since net migration between regions 2004 2006 2004 2006 2004 2006 2004 2006 2004 2006 2004 2006 2004 2006 2004 2006 is quite modest, intraregional changes in poverty Northeast East Vardar Southeast Pelagonija Polog Skopje Southwest ­ rather than the movement of poor between Source: Estimates based on HBS. regions ­ has determined the changes in overall poverty since 2002 (Box 2.1). 21 Note that the Skopje Region includes rural areas surrounding the greater city of Skopje and is therefore not synonymous with the city. The differences are small, however. 29 percent of FYR Macedonia's total population lives in the Skopje region, compared to 24.6 percent in the greater city. 15 Box 2.1: Do Poor People Move ­ And do they Bring Poverty with Them? Since 2004, poverty has increased quite significantly in the Northeastern, Eastern and Southeastern regions (Kumanovo, Strumica and Shtip) of FYR Macedonia: in two years only, the number of poor increased by a total of 44,000 - 50 percent - in the three regions taken together. But is poverty conditioned by location, or is it driven by population changes, including migration? Do poor people move and do they bring poverty with them? A first answer is given by the fact that net population changes between regions appear to be limited (on a gross basis there may be significant commuting and people moving between regions, but the HBS cannot track this). Even inflows into Skopje city represent an annual increase of only 1.4 percent per year between 2002 and 2006. There is, thus, a priori no reason to believe that migration of poor people is accounting for poverty changes. A formal way to approach this question is to separate out intra-regional poverty effects vs. the interregional or "between-regional" poverty effects. The former measures the increase in total poverty that had occurred in FYR Macedonia had nobody moved across regional borders. The interregional effect, in contrast, measures how much poverty would have changed had the population moved as it did while the poverty situation in the respective regions remained the same. These two effects can be seen as measuring the poverty of regions vs. the poverty of persons. This decomposition exercise - including an interaction effect - shows that intraregional effects account for the vast majority of the change in poverty headcount indices and that the interregional and interaction effects are marginal. Since 2002, poverty reduction in the city of Skopje has contributed to lowering the national headcount index significantly. In the period 2004-2006, however, this effect was more than counteracted by poverty increases in other urban areas (especially) and in rural areas. A similar decomposition by administrative region shows that the significant poverty reduction taking place in the Skopje region was not sufficient to counteract the increase in poverty in the three Eastern regions of FYR Macedonia. Figure: Contribution to Changes in the Poverty Headcount Index (Percentage Points) By location By region 1.5% 1.50% Interaction effect Intraregional, between and interaction effects Beween effect Total poverty increase Rural Other urban 1.00% 1.0% Skopje 0.5% 0.50% 0.54% 0.00% 0.0% Southwestern Between Skopje Reg Pelagonija Northeastern Southeastern Polog Eastern Interaction Vardar 2002-2004 2004-2006 -0.50% -0.5% -0.62% -1.00% -1.0% -1.50% -1.5% Source: Estimates based on HBS. 2.23 At a regional level, the same pattern is seen: increasing inequality and low or negative consumption growth for the poor. The national pattern translates to the regional level (Figure 2.8). Among FYR Macedonia's eight administrative regions, inequality has increased in six of them: only in Skopje and Vardar has the Gini coefficient fallen, and the North Eastern and Eastern regions have seen quite significant jumps between 2004 and 2006. This increase in inequality is consistent with an anti-poor growth pattern in the eastern regions, where the poorest quintile's consumption levels were reduced by an 16 astonishing 6 to 8 percent per year in real terms. In the North East, average consumption levels also fell quite dramatically, but in the Eastern and Southeast, the growth pattern was specifically anti-poor. Figure 2.8: Eastern Regions have Seen Big Falls in Consumption Levels for the Poor and Vulnerable Consumption growth 2004-2006: average vs. the two Gini coefficient, 2004 and 2006 poorest quintiles 0.50 6.0 2004 2006 Average 0.45 4.0 Poorest Q 0.40 Second poorest 2.0 0.35 0.30 0.0 0.25 -2.0 0.20 -4.0 0.15 -6.0 0.10 -8.0 0.05 0.00 -10.0 Southwest Polog Pelagonija Skopje Northeast East Southeast Vardar MK Urban Skopje Polog FYR MK Vardar S. West S. East N.East East Rural Pelagonija Source: Estimates from HBS. 2.24 Yet differences within regions are more significant than differences between regions. For example, the ratio of the average regional consumption level of the poorest quintile and the average regional consumption level varies between 24 percent (Eastern region) and 35 percent (Vardar). The ratio of average consumption in the poorest region (North Eastern region) to the richest region (South Western) is much closer, about 70 percent. 17 CHAPTER 3. THE OLD AND NEW POOR: A SNAPSHOT OF LIVING STANDARDS IN 2006 3.1 Who are the poor? The poorest have seen consumption levels shrink, and some of the poorer regions have diverged from others and experienced worsening poverty. An important conclusion from this chapter is that the situation and characteristics of the poor also, iteratively, confirm that the issue of poverty using the basic needs approach is not yet obsolete in FYR Macedonia. Many of those whose consumption levels are considered to be inadequate lack very basic amenities such as clean heating systems, have less access to school and health care, and little success on the job market. WHERE DO THEY LIVE? THE SPATIAL ASPECT 3.2 Poverty is urbanizing. Poverty is widespread in FYR Macedonia, and pockets of poor exist in both urban and rural areas. But extreme poverty is no longer a characteristic of remote rural areas only. In fact, by 2006, poverty was more a feature of peripheral urban areas than of agricultural areas: half of FYR Macedonia's extremely poor, and two fifths of the moderately poor, now live in secondary cities. Another 14 and 16 percent are living in Skopje, making poverty an overall urban phenomenon. Table 3.1: Poverty is an Urban Phenomenon Poverty headcount index Location's % of all poor (% of location's population) Extreme Moderate Total Extreme Moderate Total Skopje 4.3 7.9 12.1 14.2 16.6 15.7 Other urban 9.5 11.8 21.3 49.3 38.9 42.9 Rural 7.2 14.0 21.3 36.5 44.5 41.4 National 7.4 11.7 19.0 100.0 100.0 100.0 Source: Estimates based on HBS, 2006. 3.3 Poverty is also most intense in the northern belt of FYR Macedonia, with the exception of the greater Skopje region (Figure 3.1). The North East region stands out among regions: out of five persons living in the region, one in extremely poor, and another one moderately poor. Polog also has significantly higher levels of poverty than the national average. The poverty incidence is comparatively low in Skopje, but since 30 percent of FYR Macedonia's population lives there, about twenty percent of the nation's extremely poor live in the greater Skopje region. 19 Figure 3.1: The Extremely Poor Live are Concentrated in the North Share of extremely and moderately poor, by administrative region, 2006 30 24.5 25 23.2 Food Moderate 21.4 20.7 20 19.1 15 12.4 9.9 10.3 10.5 10 8.1 8.7 8.8 7.9 8.0 5.2 5 1.1 0 Vardar SE E NE Pelagonija Polog Skopje SW Note: SW=southwest, SE-southeast, E-east, and NW-northeast. Source: Estimates based on HBS. WHO ARE THEY? CHARACTERISTICS OF HOUSEHOLDS AND HOUSEHOLD MEMBERS 3.4 As is typically the case, larger households, with a large number of dependents, are more at risk of poverty than others. The extreme poverty incidence of people living in households with more than 6 persons is several times higher than for smaller households. Similarly, the extreme poverty incidence in households with three or more children below the age of 14 is almost five times higher than those with two children or less; one in four is extremely poor, and nearly half of them are poor (that is fall below the total poverty line). 3.5 This ranking still holds when poverty lines are adjusted for economies of scale in household consumption and age consumption profiles. Are these high poverty levels a consequence of the way that the per capita consumption line is constructed? As discussed in the 2005 poverty assessment, the per capita consumption poverty measure (where total household consumption is divided by the number of household members) does not take into account the fact that the age composition and size of household may influence consumption needs. For example, caloric needs are not the same over the life cycle. Similarly, larger households may need to spend less on semi-public goods within the family (appliances, shelter, etc.). However, using the adult equivalent scales that do adjust for economies of scale and age consumption profiles (panel b in Figure 3.2 below), the poverty rates of larger households with children is significantly lower, especially for extreme poverty - but still much higher than for other households. One in eight persons living in households with three or more children is extremely poor. 20 Figure 3.2: Large Households with Many Children are the Poorest Even when Economies of Scale and Age Profiles are Taken into Account Per capita approach Adult equivalent 50 50 Moderate Moderate 45 45 Food poverty Food poverty 40 40 35 19.7 35 30 17.4 30 25 25 20 20 14.1 16.8 15 15 25.4 10 10.9 11.9 20.3 10 8.5 13.0 2.8 5 5 7.6 10.0 5.5 3.2 2.1 4.5 2.1 1.2 0 0 0.1 2 children or 3 or more 3 or less 4 to 5 6 or more 2 children or 3 or more 3 or less 4 to 5 6 or more less less Dependent Children Persons in household Dependent Children Persons in household Source: Estimates based on HBS. 3.6 Children make up a disproportionately high share of the poor. For the same reason, age is a significant correlate of poverty. In 2006, over one hundred thousand children under age 15 - that is more than one in four - lived in a poor household and more than one in ten in an extremely poor household. The share of children in extreme poverty increased five percentage points, equivalent to some 16,000 children, between 2002 and 2006. The key to these developments is an increase in extreme poverty among households with small children. Figure 3.3: Poverty Rates have Increased for Small Children, their Parents, and for the Elderly 3.7 And parents are getting poorer. Poverty rates by age Poverty rates have also increased for persons aged 25-39 years. These are people who are in 35% 2002 2004 2006 the age for family formation: the average age 30% of women for having the first child in FYR Macedonia is 25.3 years, and only 7.2 percent 25% of children are born to mothers under age 20 20% (Unicef, 2008). The high poverty rate for this group has repercussions for child poverty: 15% parents are getting poorer. But in addition, 10% the higher poverty rates among this group of young to middle aged workers are 5% suggesting that people in their prime 0% working age are becoming poorer (Figure <7 7 - 14 15 - 23 24 - 39 40 - 64 65+ 3.3). Source: Estimates based on HBS. 3.8 The age of household heads is also a significant factor for poverty. It follows that households with young household heads (below 40 years of age) are more likely to be poor than others: 25 percent of the population living in such households were poor in 2006, compared to 17 percent for household heads aged above 40 (Figure 3.4). Importantly, the situation for those living with elderly household heads (about 17 percent of the population) worsened significantly since 2002, mostly due to an increase in food poverty among this group. As a result, the ranking changed so that persons with elderly household heads and those with middle-aged household heads were equally poor, and the food poverty among the former become more important. 21 3.9 Poverty rates for elderly household heads with dependents have increased. There does not seem to be a higher risk of poverty for "poor, lonely widows", i.e., elderly persons who live alone without providers - poverty rates for single-person households have remained very low. Instead, poverty rates among households with an elderly household head that does not live alone have increased significantly and now touch nearly one in five persons. The interpretation of this is not clear. It is possible, however, that households with an elderly household head are larger households including several generations. This type of household could also be de facto female headed households, grouping single mothers with their children and elderly parents, or representing families where the main breadwinner is a migrant. Figure 3.4: Poverty Rates have Increased among those Living with Elderly Household Heads Poverty rates of population according to age of household head Poverty rates of population where household head is over 65, single-person and not single person 30% 20% 18% 2002 2006 25% 16% 2002 2006 14% 20% 12% 15% 10% 8% 10% 6% 4% 5% 2% 0% 0% Under 39 40 - 64 65 and over above 65, not single person above 65, single households Source: Estimates based on HBS. 3.10 Lack of education is a determinant, characteristic and result of poverty. Education levels are a strong determinant for job access and success and a strong correlate of poverty in FYR Macedonia as elsewhere (Figure 3.5). Poverty rates adults with primary education or less reach 24 percent, compared to 13 percent for those with secondary education, and 4 percent for those with tertiary education. Low skill adults make up 62 percent of the poor and have seen higher increases in food poverty than others. Figure 3.5: A Vast Majority of the Poor Are Low Educated; they have Seen Highest Poverty Increases Share of total poor, by education level Poverty headcount index 2002-2006, by education level Without school Higher 30 education 3% 7% Food Moderate Uncompleted 25 primary education 20 12% 14.9 Secondary 15 16.6 35% 10 8.3 10.7 5 9.2 5.6 3.8 2.0 4.3 1.9 0.9 2.5 0 Primary Primary or Secondary Higher Primary or Secondary Higher 43% less less 2002 2006 Source: Estimates based on HBS. 22 3.11 Education levels are even more crucial Figure 3.6: Household Head's Lack of in the case of main breadwinners. Some thirty Education is A Strong Correlate of Poverty... percent of those who live in households where the head has completed at most primary education are Household head level of education and poverty rates poor; three out of four of the extremely poor and 35.0 two out of three of the moderately poor live in such households. While secondary education does shield 30.0 household members from extreme poverty, one in 25.0 ten living with a household head with secondary 20.0 16.2 Food Moderate education is moderately poor. Less than five 15.0 percent of the population living in a household where the head has higher education is poor 10.0 12.5 9.3 (Figure 3.6). 5.0 3.6 3.8 0.0 1.2 3.12 And conversely, poor children are less Primary or less Secondary Higher likely to attend school, leaving them less well .. Poor children are more absent from school... equipped for escaping poverty. The education - Children aged 7-15 out of school 1 poverty link is also self perpetuating: poor children 8.0% access schooling less. Among extremely poor 7.0% Including absent bc. of age children aged 7-15, seven percent, about 1,700 6.0% Excluding absent bc. of age children are not in school; among the moderate 5.0% poor children, four percent are not in school. This 4.0% compares with 2 percent of the non-poor children. The number of children out of school also 3.0% increased by 13 percent between 2005 and 2006. 2.0% 1.0% 3.13 The very poorest may associate 0.0% Extreme poor Moderately Poor Non poor schooling with very high costs because of expenses on clothes, material and food, or ... for reasons of cost, quality and health opportunity costs. The reasons for absenteeism Reasons for not being in school (% of category total) 2 among poor school age children appear also to be 100% Health different from those of the non-poor. Excessive Health Health 90% Poor learning costs are stated as reasons for half of the extremely conditions 80% Poor learning conditions Poor learning poor families and for one third of the moderately 70% conditions Ambition poor, compared to one in twenty of the non-poor. 60% 50% Ambition Instead, the non-poor children are reported to not Ambition 40% Distance be in school largely because of lack of ambition: 30% Cost they are not interested or achieved the desired 20% Cost degree. The differences in health are also quite Distance 10% important, as one in six poor children is absent Cost 0% because of ill health. Admittedly, the sample size Extreme poor Moderately Poor Non poor becomes very small at this level of disaggregation Note: 1/ Some 38 percent stated age as the reason for not and the results need to be interpreted with care. being in school. 2. Excluding those who stated age as a However, it is an issue which deserves further reason. exploration. Source: Estimates based on HBS. 3.14 The poor are less likely to afford quality health services and less able to skip a day of work, or leave their children in somebody else's care, in order to seek medical assistance. Overall, the HBS indicates that in the month prior to the interview, the extreme and moderately poor had been less 23 likely than the non-poor to seek some form of medical assistance. As expected, a higher share of the poor did not have access to health insurance, either (Figure 3.7). Figure 3.7: The Poor are More Likely to Lack Health Insurance 25% Sought medical assistance 20% Did not have health insurance 15% 10% 5% 0% Extreme poor Moderately Poor Non poor Source: Estimates based on HBS. 3.15 Job access and job success. Most of the poor are poor because their families lack access to good job opportunities, rather than jobs per se. Remunerated work in the private formal sector is key to poverty reduction. Poverty rates are higher among households where the household head is not employed (22 percent) vs. employed (17 percent), but the differences are not dramatic. The share of persons contributing to household income relative to the dependents is a more significant indicator of poverty. But poverty rates are also surprisingly high among households where the household head is working, but in low skill occupations such as crafts and trades workers. In fact, poverty rates are higher for such categories than for skilled workers in the primary sector, and at par with poverty rates among unemployed household heads (Table 3.2). Moreover, there is a strong gender dimension to job access which is discussed in Box 3.1. Table 3.2: Access to Good Jobs, not only Jobs, is Crucial for Poverty Reduction Poverty headcount index 2006 (% of population within category) % of tot population Food Moderate Complete HH head employment status Employed 5.4 11.5 16.9 57 Not employed 9.9 11.9 21.9 44 HH head occupation Legislators, senior officials and managers 0.0 1.0 1.0 3 Clerks 1.1 1.9 3.0 5 Professionals 1.3 3.2 4.4 6 Technicians and associate professionals 1.4 6.8 8.2 5 Plant and machine operators, assemblers 4.3 14.9 19.2 8 Service, shop and market sales workers 9.6 10.9 20.4 8 Skilled agricultural and fishery workers 8.9 13.0 21.9 16 Income recipients 10.6 11.5 22.1 10 Dependent persons 8.4 14.3 22.7 6 Craft and related trades workers 7.2 17.1 24.3 32 Elementary occupation 10.7 17.5 28.2 2 Source: Estimates based on HBS 2006. 24 Box 3.1: Are Women in Macedonia Facing an Under Participation Trap? Unskilled women in Macedonia may be facing an "under participation trap". The female labor force participation rate in FYR Macedonia is 42 percent of all women above the age of 15 years compared to the average for Europe and Central Asia (52 percent). This differential is driven by the low participation of young unskilled women in rural areas. The existence of imperfectly competitive markets may generate an "under participation trap". In such conditions, some workers (mainly women) are paid below their marginal product of labor (i.e., cost of domestic production), which causes the supply of labor to be below that in a competitive setting. Indeed, Macedonia's labor market has some features that suggest that its labor market is not competitive. Low labor mobility, especially among low skilled women, strengthens the sense of "local" labor markets whereby differences in employment outcomes across regions are quite large. Nevertheless, workers ­ especially women ­ seem not to move from worse to better performing regions in order to seek better job opportunities. These combined factors are generally prevalent in labor markets where firms have monopolistic power. Furthermore, an interesting feature of the labor market in FYR Macedonia is the existence of a large wage gap between unskilled men and women despite little labor segmentation (i.e., unskilled women and men work roughly in similar professions). This suggests that firms can afford paying a wage premium to workers based on the gender. Low female participation rates are economic as well as social in nature. Some of the main issues are as follows: (i) women with lower education typically do not want to work or do not believe they would find a job with their qualifications. (ii) An important discouraging factor that prevents women to seek employment is related to high transaction costs and low chances to get good jobs given their education level. (iii) Unskilled women display high reservation wages and unfavorable job opportunities. (iv) There are still cultural constraints affecting mainly women in the non-ethnic Macedonian communities, especially resulting from the traditional roles in a patriarchal society. Some recommendations for escaping the under participation trap: · Investments in secondary education: Since educational attainment is a significant determinant for the inclusion of women into the labor market, it is important to encourage girls to actually complete secondary school and empower them to seek employment. In addition to this, there should be special programs which would provide adult education (primary and secondary) ­ ideally free of charge or with symbolic contributions - to individuals who have not completed their formal schooling. · A more flexible of the labor market: Low-skilled women (even those from traditional communities) do not actually object to working, but rather are uncomfortable working formally outside of their house and their community. Consequently, ensuring that labor regulations are sufficiently flexible enough to allow for the establishment of small businesses, especially in house (e.g., in-home day-care centers), could enhance the participation of women in the labor market. · Targeted programs to promote female employment: Programs such as subsidizing childcare and providing wage subsidies to women may also be considered as well as other policies aimed at strengthen market competition and at improving women's job condition. 3.16 Ethnic background remains a key correlate of poverty in FYR Macedonia. The ethnic divide is not clear-cut as households headed by ethnic Macedonians can include non-ethnic Macedonians and vice versa, due to inter-marriages. Yet, extreme and moderate poverty rates are higher for persons living in households headed by a non-ethnic Macedonian. Moreover, the increase in extreme poverty is more important among non-ethnic Macedonian households than ethnic Macedonian ones). Partly these 25 differences are due to lower schooling, regional belonging, larger households, and the fact that non ethnic Macedonians, especially women, are much less likely to be active in the labor market. However, the ethnic factor remains significant including in a multivariate analysis where other characteristics are controlled for.22 HOW DO THE POOR LIVE? 3.17 Location, rather than consumption per capita, remains a key determinant for infrastructure access. Broader measures of well being beyond consumption levels are generally correlated with consumption and income poverty. Thus, in terms of living conditions, those below the poverty line live in more crowded conditions and are more likely to lack key amenities. However, location remains generally a more significant correlate of housing conditions than poverty. Urban density both requires and makes possible more coordinated infrastructure services, while lack of infrastructure is a more general phenomenon in rural areas, and lacking access in turn may have less significant impact on well being. For example, not having sewage disposal is a serious problem in crowded urban neighborhoods, but may not be a measure of deprivation in a rural dispersed setting with access to land, however small. As an illustration, forty percent of non-poor rural citizens have no sewerage disposal, compared to 0.2 and 12 percent of the poor in Skopje and other urban areas respectively; only 8 percent of the rural non-poor have central heating. 3.18 The difference in access to services between poor and non-poor is therefore more pronounced in urban areas. The poor live in more crowded conditions, with less sophisticated heating systems, and less access to broader infrastructure services. Poor citizens in secondary cities are worst off, most likely because poverty is deeper there than elsewhere. The poor live on two-thirds of the area of the non-poor, three in four of Skopje's poor do not have access to central heating, compared to one in two non-poor persons, and about thirty percent of urban poor citizens do not have a telephone line, compared to less than ten percent of the non-poor. Telephony is important as it facilitates job search as well as access to other services (for example, making appointments with doctors) (see Table 3.3). 3.19 Consumer durables such as television and radio, refrigerators, vehicles and other possessions are fairly widespread although some gaps remain to the better-off. One in five of the poor live in households with no washing machine, compared to one in ten non-poor persons. One in six of the poor live in households without either mobile or fixed phone line; and only one in ten have access to a personal computer. Only three percent of the poor live without a refrigerator, however. The lack of vehicles and phones potentially makes access to jobs more difficult. The lack of such appliances as fridges and washing machines potentially adversely impacts women's labor force participation as someone is needed at home to fulfill time consuming tasks of frequent shopping, cooking, and hand washing clothes. 22 The Government introduced in 2005 a strategy to improve the integration of the Roma in the society. 26 Table 3.3: Housing Conditions Gap is Most Significant in Urban Areas Skopje Other urban Rural Non poor Poor Non poor Poor Non poor Poor Two-room or less 45.7 43.8 33.3 39.3 23.0 25.2 Not owning 14.4 25.1 13.2 11.9 9.6 8.3 Dwelling area per capita (m2) 20.8 14.7 25.5 17.2 24.8 18.8 Without water supply 0 0 0.2 4.9 9.3 7.8 Without sewage disposal 1.6 0.2 2.7 11.9 39.7 50.0 Without electricity 0 0.8 0.5 1.6 0.9 0.8 Without central heating 50.6 72.5 89.9 91.9 94.8 98.7 Without fixed or mobile phone 5.9 26.7 9.1 29.6 15.3 30.0 Without bathroom 1.0 2.4 3.0 8.7 12.9 17.1 Heating by stove - on electricity 23.5 20.4 16.6 5.7 1.3 3.7 Heating by stove - on solid fuels 29.5 50.4 75.0 90.1 94.5 95.1 Source: HBS. 3.20 In contrast to consumption poverty, the housing situation is improving slightly and the differential access rates between poor and non-poor is shrinking. Generally, housing conditions have improved slightly for the poor since 2002: their living areas have increased, and they are less likely to lack water, sewage and other amenities (Error! Not a valid bookmark self-reference.). These small improvements are noticeable across the different types of amenities and across space. The improvement is not easy to interpret, since in the same time span consumption measures suggest poverty has deepened in areas outside Skopje and the share of poor has stagnated (Skopje, rural areas) or increased (other urban areas). Figure 3.8: Housing Conditions have Improved for the Poor % living with deprivations, in 2002 and 2006 100 90 2002 2006 80 70 60 50 40 30 20 10 0 Two rooms No water No sewage No No central No washing No phone No PC or less electricity heating machine Source: Estimates based on HBS. 27 3.21 The poor spend much more on food, in relative terms, than the non-poor. Food expenditures account for 77 percent of all expenditures of the extremely poor, and 70 percent of all expenditures of the moderately poor (as seen on Figure 3.9). The non-poor, in contrast, allocate only 40 percent of their expenditures to food. As a result, the non-poor spend about twenty times as much as the extremely poor on non-food expenditures. The increase in food prices in 2007 consequently must have hit the poor hard.23 Figure 3.9: Food Expenditures Constrain the Non-Food Expenditures of the Poor 100 90 Other 80 Transp & Comm 70 Health 60 Shelter 50 40 77 Clothes 70 30 Smoking/ 40 Drinking 20 39 Food 10 0 All MK Extreme Poor Moderate Poor Non-Poor Source: Estimates based on HBS. SUBJECTIVE OPINIONS ON WELFARE AND WELFARE CHANGES 3.22 Two in three Macedonians are not satisfied with their household's level of income. The household survey includes subjective questions about the respondent's own perception of the household's financial situation. For 2006, these measures are not inconsistent with the overall poverty picture from the household survey (Figure 3.10). About 20 percent of the population is not satisfied at all with their income level in relation to their spending needs, and another 46 percent are more or less dissatisfied. The perception of what is an acceptable minimum of income/spending tends to increase with income/spending: only 40 percent of non-poor households are satisfied with their current income levels. As would be expected, a vast majority - about 90 and 83 percent ­ of extremely poor and moderately poor households are somewhat or completely dissatisfied with their income levels given their spending needs. 3.23 Subjective poverty is not consistent with consumption poverty. Ten percent of the extremely poor, and almost twenty percent of the moderately poor, are not satisfied with their income levels in 2006. Moreover, the share of Macedonians who to varying degrees are dissatisfied with their living standards has fallen since 2002, from 73 to 65 percent. This decline is as discernible among the poor as among the non-poor. It is consistent with the improvement in housing conditions and some of the social indicators discussed in Chapter 1, but is inconsistent with the increase in consumption poverty at the same time. 23 The Government is supporting a pilot project where food can be purchased at below market prices by beneficiaries of selected social assistance programs. 28 Figure 3.10: Subjective Poverty Measures Indicate an Improvement in Wellbeing Question: Considering the household's total income, does the Percentage reporting to be more or less dissatisfied or totally household make ends meet? dissatisfied with their household's ability to make ends meet. Satisfied More or less satisfied More or less unsatisfied Not satisfied at all 95% 100 90 19 16 90% 34 31 80 85% Poor 70 80% 60 45 46 75% 50 52 70% 40 55 30 65% Total MK 20 34 30 60% Non-poor 10 17 9 0 5 2 0.1 6 55% Extreme poor Moderately Poor Non poor 50% Total Poverty Status 2002 2003 2004 2005 2006 Source: Estimates based on HBS. 29 CHAPTER 4. GROWTHLESS JOBS AND JOBLESS GROWTH: INCOME OPPORTUNITIES AND POVERTY 4.1 What is the economic context for the increase in inequality and the falling consumption patterns of the poor? FYR Macedonia's overall poverty situation has worsened since 2002 due to a regressive growth pattern where consumption expenditure growth of the poorest has been low or even negative in the past years. The analysis that follows shows that the poor depend increasingly on informal sources of income, while social assistance to the poorest has declined. FALLING INCOME 4.2 In general, falling consumption levels can be traced to falling real income, whether from jobs, transfers from family and friends, or public safety nets. Falling consumption levels of the poor are likely to reflect falling real income ­ living near the subsistence minimum, they may not be able to save. Poor households hold few capital assets, relatively few of them depend on farming, and they are as a consequence limited to three potential sources of income (though sources of consumption may include subsistence farming and other means of in nature goods). First, they can derive income from household members' work. Second, they can derive income from family members not living in the household but transferring money (remittances, whether international or domestic). And third, they can receive income support in various forms from the state. 4.3 The average household has seen some income growth, driven essentially by labor income. Table 4.1 presents income structure in 2006, and growth of different income sources since 2002. Bearing in mind that income estimates from surveys are highly imprecise24, overall, FYR Macedonia's households appear to depend mostly on formal labor income and, for the eligible, pensions. The average household has seen relatively high income growth since 2002, with non agricultural income and private transfers increasing quite significantly. Public social security has fallen, however, especially public transfers. Agricultural income has been volatile but its impact has been less significant since it accounts for a small share of income. 4.4 In recent years, the poor have seen shrinking public transfers and a shift towards higher dependence on agricultural and informal income. Those in the poorest consumption quintile have seen much more significant shifts. First of all, the income structure is different: higher dependence on informal non-agricultural income, and a higher dependence on public transfers. Second, the period 2002-2004 saw quite high income growth, essentially driven by agricultural and informal labor income. Public transfers fell significantly in the same period. Third, the period 2004-2006 saw much less positive income growth developments. Agricultural income fell, non-agricultural income stagnated (due to declining formal income), and public social security fell significantly. Private transfers, including remittances though these cannot be singled out, increased at the same time, suggesting that they may fill some insurance function. 24 Income estimates are generally considered less reliable than consumption estimates because of a significant risk of underreporting though this risk is probably increasing with income level. Moreover, income is recorded for a recall period of 3 months. For example, the average value of the richest quintile's consumption is twice as high as their reported average income. 31 Table 4.1: Stagnant Income Growth and Falling Transfers for the Poorest Total FYR Macedonia Poorest 20% of the population Annual Growth Share Annual Growth Share 2002-2004 2004-2006 2006 2002-2004 2004-2006 2006 Total annual Income Per Capita 11 6 100 20 -5 100 1) Agricultural Income 26 -10 6 81 -24 6 2) Nonagricultural Income 10 12 61 26 1 59 a) Formal 10 11 55 22 -2 45 b) Informal 12 24 7 45 15 15 3) Public social security 10 -6 19 21 -16 21 a) Pensions 14 -2 17 56 -18 16 b) Public Transfers -7 -29 2 -27 -11 6 4) Private Transfers 10 25 5 -33 62 5 5) Other 15 5 9 -6 -23 8 Source: Estimates based on HBS. GROWTH PATTERNS - NO TRICKLE DOWN 4.5 Growth in FYR Macedonia has been relatively low compared to neighboring countries during 2002-2006, although it has been accelerated during 2007-2008. Moreover, the distributional consequences of growth, driven by its composition, also matter. For example, growth sectors ­ in the case of FYR Macedonia, private services sectors ­ will only have a positive impact on income distribution if it affects the employment and wage opportunities of the poor. Job-less growth, or job growth which centers on the high skilled only (who are relatively few), is likely to have a much less significant impact. 4.6 Private consumption makes the largest contribution to economic growth, but its role has been shrinking since 2003. Private consumption has been the main engine of economic growth in FYR Macedonia, adding between 3 and 5 percentage points to economic growth rates since 2003 (Figure 4.1). But while overall economic growth has increased, the role of private consumption has been slightly reduced, in favor of an increasing role for investment in 2006 and 2007. However, the high growth rate in 2005 is not consistent with the HBS data, as shown in Figure 2.4 above. Figure 4.1: Private Consumption and Private Services are the Two Drivers of Growth from Expenditure and Activity Sides Expenditure side: Contribution to growth (percentage points): Activity side: Contribution to growth (percentage points). Net exports 6.0 Public services 10.0 Change in stocks Gross domestic investment Private services Public consumption 5.0 Industry 8.0 Private consumption Agriculture 6.0 4.0 4.0 3.0 2.0 2.0 0.0 1.0 -2.0 -4.0 0.0 -6.0 -1.0 1997-2002 2002-2003 2003-2004 2004-2005 2005-2006 2006-2007 1997-2002 2002-2003 2003-2004 2004-2005 2005-2006 2006-2007 Source: Estimates based on World Development Indicators (World Bank). 4.7 In terms of economic activities, private services and to a lesser degree industry sectors have been driving economic growth. Construction and the manufacturing industry, together with trade and 32 transports, accounted for some 54 percent of total growth between 2002 and 2006. The average contribution of agriculture has been modest (12 percent, or about 0.4 percentage points) and in fact negative in 2006. 4.8 Official remittances have been increasing since the early 1990s but unofficial transfers are likely to be much higher. Migration data for FYR Macedonia are scant and unreliable. Estimates of the Diaspora range from between 350,000 to 2 million, and while the UN Population estimates suggest that net migration ­ immigration into the country less emigration ­ doubled between 2000 and 2005; Eurostat estimates a fall in net migration since 2002. Substantial transit migration from CIS countries, seasonal migration especially to Bulgaria and Greece, and illegal migration mean that official figures are not reliable. In 2005, 1282 Macedonians emigrated abroad according to official statistics, but in the same year, over 2000 Macedonians were readmitted for having attempted to enter a European country illegally (MPI, 2007). Official workers remittances have increased in importance from 2000 and onwards, and have surpassed total private capital flows since 2002. Again, however, it is difficult to gauge the underlying trend in total remittances: unofficial transfers are expected to reach several times these numbers (World Bank, 2008)). Figure 4.2: Net Migration has Increased and so Have Remittances Emigration and Net Emigration (thsds) Remittances (% of GDP) 30 4.5 4.2 4.0 3.9 4.0 3.8 25 3.5 3.5 20 3.0 2.8 Official Emigration 2.5 2.3 Net Emigration (Eurostat) 2.1 15 Net Emigration (UN) 2.0 10 1.5 1.0 5 0.5 0 0.0 2000 2001 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2005 2006 2007 Source: World Development Indicators (2008), MPI ( 2007). GROWTHLESS JOBS AND JOBLESS GROWTH 4.9 In terms of mere numbers, the job situation is improving in FYR Macedonia. Data from the labor force surveys from 2004-2006 show that employment rates, i.e., the share of the working age population that is actually employed, have increased since 2004, from 36.9 to 39.6 percent, representing nearly 50,000 additional jobs in the economy. These jobs covered both a growth in the working age population, and an increase in the share of active persons. Thus, labor force participation rates increased from 59.5 to 62.2 percent of the population, equal to 30,000 new job searchers on the market. But in spite of these improvements, by 2006, three in five persons of working age still did not have a job. 4.10 Job creation did not take place in sectors with economic growth, nor in high quality job forms. On a net basis, most new jobs, some 60 percent, were created in the primary sector, with an additional 30 percent in services. Hence most employment creation took place in agricultural regions, notably Pelagonija and the Southeast, where agricultural employment accounts for 30 and 45 percent of total employment. Polog also saw rapid employment growth due to the services sector which accounts for two thirds of employment there. In contrast, Skopje lost 20,000 jobs, most of them in the services sector, in spite of continued inflow of population and workers into the region. 33 Table 4.2: Key Labor Market Indicators Have Improved Changes in 2004 2005 2006 2004-2006 Labor market indicators, in percentages Employment rate 36.9 38.0 39.6 Unemployment rate 38.0 38.2 36.3 Labor force participation rate 59.5 61.5 62.2 Labor market indicators, in thousand people Working age population 1399 1412 1420 22 Working 516 536 563 47 Not working 882 876 858 -25 Unemployed 316 332 321 5 Inactive 566 544 536 -30 Source: LFS 2004-2006. Box 4.1: Labor Market Indicators ­ Sources and Definitions To investigate labor market data, this report uses two sources of information HBS data for 2002-2006 allows us to link poverty to employment outcomes directly for the entire period. Figure: Labor market indicators in HBS and LFS However, the HBS is less detailed and less reliable on (2004-2006) labor market data more generally, since employment status is not the main focus of the survey. 50% 45% Labor Force Survey data for 2004-2006 provides a 40% more accurate and richer view of labor markets. 35% However, the LFS does not measure poverty. In 30% addition, because of changes in the questionnaires, data 25% are comparable between 2004 and 2006 only. 20% U, HBS E, HBS 15% U, LFS E, LFS Because survey questions are not identical across LFS 10% and HBS, the labor market indicators tend to differ 5% between the two sources of data. Importantly, however, 0% differences are mostly in level: LFS shows lower 2004 2005 2006 employment rates and higher unemployment rates than the HBS. However, as shown in the accompanying figure, the trends over time are similar. 4.11 These jobs were mostly unpaid jobs for low or unskilled youth, probably largely representing hidden unemployment and underemployment. Out of the 47,000 net jobs created in 2004-2006, 60 percent were taken by low skilled (people with primary education and less), and a vast majority in the private sector. Job creation for the low skilled in the private sector would encompass the poor. But job creation can impact poverty only if it offers income opportunities. However, 36 percent of all new jobs ­ more than one in three ­ were unpaid jobs in the primary sector, with another 19 percent being self-employment in the primary sector. While unpaid jobs can have a positive impact on labor market access and earning capacities over the longer term by building experience for first time job seekers, the high share of unpaid jobs in the primary sector, and the low skill nature of the jobs, preclude such long term positive effects. Unpaid jobs or self employment are the ones with the highest share of 34 informality, lack of job security and social protection, and with the highest shares of persons wanting to work more time than the job offers (Figure 4.3). Figure 4.3: More Jobs, but Not Better Earning Opportunities Net job creation (thousands of jobs), Characteristics of new jobs (% of all new jobs by 2004-2006, by region and sector. category) 50 70% 40 13 AGR IND SERV 60% 30 7 20 50% 27 10 Unpaid and selfemployed Primary sector, Bitola and 40% 0 in primary sector Low skill Strumika -10 30% Self employed Unpaid work -20 -30 20% Southwest Skopje Eastern Total Vardar Northeast Southeast Pelagonija Polog 10% 0% Source: LFS 2002-2006. 4.12 Consequently, the group of employed poor is increasing. Poverty rates are higher among the unemployed or inactive than among the employed, as is logical ­ the employed contribute earnings to their household, and by consequence their household is less likely to be poor. However, there is a shift towards higher poverty rates among the employed as well, where extreme poverty rates now approach those of the inactive/unemployed. Between 2004 and 2006, the poor possessed one third of all new jobs. Since many of these jobs are likely to have been unpaid and informal, they were not enough to lift persons out of poverty (Figure 4.4). Figure 4.4: Poverty Rates among the Working are Increasing Poverty headcount, working population aged 15-64 Poverty headcount, inactive or unemployed population aged 15-64 25% 25% Moderate Poverty Food Poverty Moderate Poverty Food Poverty 20% 20% 15% 15% 10% 10% 5% 5% 0% 0% 2002 2003 2004 2005 2006 2002 2003 2004 2005 2006 Source: Estimates based on HBS. 4.13 And the trend towards more employed poor is visible at the household level as well. Households with lower dependency ratios (i.e., higher share of employed) are less poor than households where few or none are working, and households where the household head is employed are less poor than when the household head is not employed (Figure 4.5). But the categories are approaching one another, raising poverty rates for households with higher share of employed persons. Clearly, access to labor 35 market is not a sufficient criterion for poverty reduction in FYR Macedonia. Though upon excluding Skopje, no correlation exists. Figure 4.5: Access to Labor Markets is Losing Importance for Poverty Reduction Poverty rates for household dependency ratios Poverty rates by employment status of household head 40% 25% Decrease Decrease in poverty 2002 2004 2006 2002 2004 2006 35% in poverty 20% Increase in poverty 30% 25% increase in poverty 15% 20% 15% 10% 10% 5% 5% 0% 0% Ratio =1(all Ratio >1 and Ratio >2 and Ratio >4 No employed Employed Not employed employed) <=2 <=4 Source: Estimates based on HBS. 4.14 The working poor are now supporting larger households. Those among the extreme poor that are working are supporting larger households in 2006 than in 2002. This increase in household size can explain why higher employment rates are not, in fact, reducing poverty: income is spread out among more participants. This reinforces the combination of unfavorable income developments and shifts to income sources with less lower income potential (as shown above). Estimates of hourly earnings show no clear pattern of pro or anti-poor impact. Some of the important low wage sectors (trade, manufacturing) saw quite limited increases in wages while others, e.g., tourism, saw high increases.25 Figure 4.6: Average Household Size Has Increased Among the Very Poorest Average household size by poverty status Hourly earnings: level vs. growth 6.0 50% 2002 2006 Private HHds Construction 5.0 40% Tourism Wage growth 2004-2006 4.0 30% Mining 3.0 20% Utilities 2.0 10% Trade Public adm 0% Manufacturing 1.0 Transport Education and Other services health Finance 0.0 -10% Extreme moderate nonpoor 40 60 80 100 120 140 160 Wage level 2004 (index, average wage=100) Source: Estimates based on HBS and LFS. The lack of positive dynamics between job creation and poverty reduction is evidenced at a regional level. With a virtuous circle of production, productivity and job growth, poverty would have fallen in job creating regions. This has not been the case; in fact, the regions with Southwest, Southeast, Polog and Pelagonija, have seen poverty numbers stagnate or increase (Figure 4.7). If anything, the relationship between job creation (thousands of jobs) and poverty increase (thousands of poor) is positive. 25 Note that these numbers need to be taken with care as (i) earnings reported in surveys is generally underestimated (ii) some of the sample cells become very small at disaggregated level, reducing the representativeness (iii) hourly earnings are only compared for full time employees and thus does not take into account underemployment. 36 Figure 4.7: Job Creation is Not Resulting in Poverty Reduction at a Regional Level 30 Change in number of poor, 2004-2006 R2 = 0.1987 Northeastern 20 Eastern 10 Southeastern Southwestern 0 Pelagonija Polog -30 -20 -10 0 Vardar 10 20 30 -10 -20 Skopje -30 Change in number of jobs, 2004-06 Source: Estimates based on HBS and LFS. SOCIAL ASSISTANCE REACHES LESS THAN HALF OF THE POOR 4.15 The social protection system has a wide reach and its social welfare components could benefit from improved targeting to the poor. The social protection system in FYR Macedonia is comprised of contributory social insurance programs and non-contributory social benefits which include income support benefits, child allowances and disability benefits. Nearly 90 percent of total social protection benefits, which include pensions, are allocated to households above the poverty line. The social insurance programs compensate those who paid into the social insurance system as they held jobs in the formal sector (though the need for budgetary transfers to the pension fund suggests that the Government is subsidizing the system). Of the total amount of reported pension income reported in the HBS, about 92 percent is received by non-poor households; less than three percent is received by extremely poor households. The non-contributory social benefits programs are designed to help those who are in need of state aid and comprise instruments to reach disadvantaged groups such as able body individuals who are unable to provide financial security for themselves, invalids, students in need of financial aid, and families with young children. (See Annex B for an overview of the main cash transfer programs in FYR Macedonia.) 4.16 Though less than half of total social assistance transfers benefit the poor, this represents above average performance by ECA standards.26 An estimated 1 percent of GDP is allocated towards social welfare programs compared to the 2.5 percent average for OECD countries. In the ECA region, preliminary analysis of the targeting accuracy of social assistance programs using a slightly different welfare aggregate indicates that on average (unweighted) about 40 percent of all social welfare benefits 26 The methodology used to determine the targeting accuracy of the social protection coverage of the poor is the "post transfer) welfare aggregate. Thus, we look at the consumption without subtracting the transfer amount received, which means that consumption of recipient households is higher than would be the case in the absence of the transfer. When we look at "pre-transfer) consumption (that is, we exclude the transfer amount from the consumption aggregate), we find that poverty increases by about 5 percentage points of the population. 37 are received by the poorest quintile while for FYR Macedonia the share is closer to 50 percent.27 Among FYR Macedonia's social welfare system, around 19 and 22 percent of the total spending on the social assistance program is distributed to extremely poor and moderately poor households respectively. But there is still substantial leakage, with the wealthiest two fifths of the population receiving about 20 percent of total social assistance spending and over 60 percent of total public subsidies spending (food, etc.). This being said, FYR Macedonia's performance is among the top four out of 26 in the Central and Eastern Europe and Central Asia region. Figure 4.8: Social Protection performance Distribution of public transfers to the population by poverty status Extreme poor Moderate poor Non-poor 100 90 80 70 60 50 40 30 20 10 0 Poverty Invalidity Child Scolarship Pensions Subsidies Total Soc Health Ins. Assist. Status Soc. Prot Source: Estimates based on HBS. 4.17 Public social assistance is also covering fewer of the poor. The distribution for some items has improved slightly since 2002 ­ because although the allocations to the poor fell, they fell less than for the wealthier. The increase is minimal, however, and as shown above, the importance of social assistance from public sources fell quite significantly between 2002 and 2006, also for the poor. This decline goes hand in hand with a decline in coverage among the poor. The share of overall households receiving social assistance benefits declined from fourteen to eleven percent. The decline was more dramatic for the poor households: in 2002, half of the extremely poor households and one in three moderately poor households received social assistance. By 2006, only one in three extremely poor households and one in five moderately poor households received social assistance benefits (Figure 4.9). 4.18 However, pension coverage has increased. Over the medium and long term, the number of households receiving pensions will increase in FYR Macedonia as the population is aging. The share of households receiving some form of pensions income in FYR Macedonia increased from 34 to 38 percent between 2002 and 2006. Because of a significant widening in pensions coverage between 2002 and 2004, 37 percent of extremely poor households, and 34 percent of the moderately poor, now receive some form of pension income. The demographic composition of households will play a part in determining their eligibility for especially pensions and child benefits which makes comparisons across groups difficult. 27 These estimates are based upon work in progress by a team led by Kathy Lindert in the Human Development Unit in the Europe and Central Asia region of the World Bank. 38 However, the overall picture is consistent: better coverage and higher income from pensions, and less coverage from social assistance and other welfare benefits. Figure 4.9: Social Assistance Coverage has Fallen, Pensions Coverage Improved % of population in households which receive social % of population in households which receive pensions assistance. 60% 45% Extreme poor Moderate Poor Non poor All MK 40% 50% 35% 40% 30% 25% 30% 20% 20% Extreme poor Moderate Poor 15% Non poor All MK 10% 10% 5% 0% 0% 2002 2003 2004 2005 2006 2002 2003 2004 2005 2006 Source: Estimates based on HBS. 4.19 Pension income is also becoming more crucial in supporting consumption levels. For extremely poor households receiving pensions, pension income is equivalent to eighty percent of total consumption, and for moderately poor, more than half of consumption expenditures. The share of consumption has increased since 2002 ­ but during this time consumption levels of the poor have also fallen. In contrast, social assistance is covering considerably less consumption in 2006 compared to 2002 28 (Figure 4.10). Figure 4.10: Pensions Income is Critical for Poor Receiving Households, but Social Assistance is Not Helping to Make Ends Meet For receiving households: Share of transfer in total consumption 90 80 80 75 Extreme Moderate 70 Non poor 60 56 53 50 40 40 32 33 32 30 27 21 20 10 8 10 0 2002 2006 2002 2006 Pensions Soc. Assist. Source: Estimates based on HBS. 28 Again, comparisons between consumption and income are fraught by differences in recall period. 39 4.20 Remittances are compensating for some of the fall in public transfers. Remittances, like other transfers, are not directed more to the poorest in FYR Macedonia ­ this is quite consistent with international experience which suggests that migrants tend not to be from the poorest families, because of the transaction costs involved in migration. In 2006, non-poor households received 89 percent of total remittance income, leaving eleven percent for the poorest twenty percent of the population. However, remittance income has taken on a more important role for the poor over time. In 2002, less than one percent of the extremely poor population lived in a household which received remittances from abroad. By 2006, this number had increased to over 8 percent. For moderately poor, the coverage increased from 1 to 5 percent. Since the share of non-poor has remained stagnant, a higher share of the poor than the non- poor is now covered by some form of remittance income. Notwithstanding this increase, less than one in ten of the poor can rely on remittance income (at least as recorded in the HBS) to support their consumption (Figure 4.11 ). Figure 4.11: Remittances are Reaching a Higher ­ but Still Limited ­ Share of Poor Households % of population living in households receiving private transfers 9% Extreme poor 8% Moderaty Poor 7% Non poor 6% 5% 4% 3% 2% 1% 0% 2002 2003 2004 2005 2006 Source: Estimates based on HBS. 40 CHAPTER 5. JOB DYNAMICS BEYOND SUMMING UP AND NETTING OUT 5.1 Is FYR Macedonia's labor market stagnant ­ or are dynamics going on at a disaggregate level? This chapter looks at the dynamics of job creation in the non-agricultural formal sector.29 A conclusion is that the relatively limited net employment growth outside agricultural is masking much more diverse picture, with job creation and job destruction going on within economic sectors. There is also evidence of a shift towards more efficient job creation in the private services sector. However, the type of firms that are creating jobs on a net basis are not the ones who might offer good job opportunities ­ they are predominantly very small firms that experience negative labor productivity growth. LABOR MARKETS LOST IN TRANSITION? 5.2 The transition process involves a reallocation of productive activities and employment, from the state sector to the private sector; generally also from industry to services and from unproductive to productive firms. In former communist countries where the initial distribution of activities was particularly skewed to the public sector or where the reform process was slow, the private sector was initially slower to respond, total employment growth fell and unemployment soared. 5.3 Is FYR Macedonia lost in transition? Between 2004 and 2006, the primary sector (i.e., agriculture and mining) accounted for three in five new jobs in FYR Macedonia, a vast majority of which represented self-employment or unpaid family work. Agricultural underemployment and a relatively lethargic rate of job creation in the non-agricultural sector are reminiscent of job patterns in the early stages of most transition countries, especially in the CIS countries. However, behind insignificant aggregate net changes in employment in any economic sector, there may be substantial shifts as jobs are destroyed in one type of firms and created in others. Importantly, while labor force data can generally tell us about inter-sectoral shifts in employment, firm data can be used to understand many more diverse types of reallocation within sectors. In a country like FYR Macedonia, where employment creation continues to be the key policy challenge, the underlying dynamics can help us understand the job market in several dimensions. 5.4 This chapter uses the analytical tools of Davis and Haltiwanger to investigate employment dynamics and the efficiency of job creation at firm level, using firm registry data from FYR Macedonia for 2002, 2004 and 2006. Almost all firms are in the non-agricultural sector, although the industry sector is overrepresented compared to the employment structure depicted in the LFS. The firm data allow for a deeper analysis of shifts of job creation in the economy, for an understanding of 29 In 2006, an estimated 70 percent of all workers were in the formal sector (that is, persons who said that they were employed and registered with the Pension Fund) which includes persons employed in the agriculture and fishing sector (19 percent of all workers in 2006). If we look only at workers in the private sector (including agriculture), formal employment falls to 57 percent. However, the share of formal sector workers in total private non-agriculture sector increases to about [80] percent. (Estimates based upon work in progress by Victor Macias.) 41 reallocation and productivity changes beyond aggregate economic sector categories, and for a distinction between changes brought about due to dynamics among existing firms and those that are due to firm turn- over. The source of data and the key concepts are summarized in Box 5.1. Box 5.1: Key Concept and Data The firm data are from the Central Firm Registry to which Macedonian firms submit their annual statements, for 2002-2006. By 2006, the entire database comprised nearly 50,000 firms which accounted for some 260,000 jobs. Using the data from the LFS 2006 as a benchmark, the registry data cover about half of total employment in FYR Macedonia. The registry data covers about 90 percent of employment in the manufacturing sector, and 70 percent of the private services sector. Total value-added of the firms in the registry sums up to about half of total GDP. The analysis excludes all firms with zero-employment. This reduces the number of firms to about 38,000. Because of data deficiencies, we are using the data from 2002, 2004, and 2006. Unfortunately, this introduces a problem of our understanding of "real" entry and exit and of comparability, because normally, job turnover is measured on an annual rather than bi-annual basis. Specifically, firms that were created in 2003 and liquidated in 2004, or created in 2005 and liquidated in 2006, will not be registered at all in the data-base. Entry and exit rates could therefore be significantly higher than the numbers presented here suggest. The key concepts used in the note are as follows: · Gross job creation in one year (in a particular sector) is the sum of all employment gains that year in firms (in that sector) that start up or expand during the year. · Gross job destruction is the sum of all employment losses in firms that contract or shut down (in that sector) during the year. · Net employment growth is the difference between gross job creation and gross job destruction. · Gross job reallocation is the sum of gross job creation and gross job destruction, and characterizes the dynamics (extent of job creation and job destruction) of the labor market. Thus it is quite possible to have high gross job reallocation but low net employment growth (if high creation and destruction cancel out). Such dynamics, while not leading to more jobs, can lead to better/worse jobs, with higher/lower productivity and related earnings possibilities. · Excess job reallocation is the difference between the gross job reallocation and the absolute value of the net employment growth. This measure captures the amount of "churning" by firms, i.e. how much actual job reallocation exceeds what would be necessary to accommodate the net change in employment. These indicators are expressed as rates by dividing them by total employment numbers (in a particular sector). 5.5 Transition countries have seen rapid job destruction and varying degree of job creation; and a diminishing role for firm turnover over time. Because of the massive shifts in the economy, job creation and job destruction tended to exceed those of more developed countries. In general, transition countries experienced significant job losses, but differed greatly in the extent to which they managed to create jobs. Where job creation and job destruction has been unsynchronized in time, redundancies have increased and less people have been participating in labor markets (World Bank, 2005). As seen in Figure 5.1, transition countries (late 1990s data) generally experienced higher job creation and job reallocation rates than OECD and a comparable developing region (Latin American and the Caribbean). 42 Figure 5.1: Job Turnover, Different Indicators 0.35 OVERALL 0.30 OECD LAC 0.25 TRANSITION 0.20 0.15 0.10 0.05 0.00 Job creation Job destruction Net Job reallocation Excess job rate rate employment rate reallocation rate growth Note: LAC = Latin America and the Caribbean region. Source: Haltiwanger, Scarpetta, and Schweiger, 2008. JOB DYNAMICS IN FYR MACEDONIA 5.6 Jobs were destroyed on a net basis between 2002 and 2004, but recovered between 2004 and 2006, due to small firm growth. Between 2002 and 2004, net job creation was negative: some 15,000 jobs were destroyed. In the subsequent period, employment growth was positive, reaching 2.2 percent per year ­ however, by 2006, employment levels had not recovered to 2002 levels. In both periods, labor productivity (average value-added per worker) increased quite significantly. The weak recovery of employment since 2004 is consistent with LFS data (as seen in Chapter 4).30 Table 5.1: Key Numbers from the Firm Registry Year 2002 2004 2006 Number of firms (thousands) 29 33 38 Total employment (thousands) 266 250 261 Average employment 9.1 7.6 6.9 Total employment growth (annual %) -3.0 2.2 Average value added growth (per firm) -3.5 1.7 Average labor productivity growth 6.0 6.5 Source: Estimates based on Firm Registry Data. 5.7 Low net job creation numbers mask high rates of job creation and job destruction in at a firm level. Between 2002 and 2004, about 27 percent of all jobs were destroyed each year, but only 20 percent of jobs recreated ­ and hence, employment was shrinking on a net basis. But while job destruction remained just as high in 2004-2006, job creation increased to 31 percent. In other words, in any one year, 3 in 5 jobs were reallocated (destroyed or created) across the economy. The extra job reallocation needed 30 The share of total employment in the Firm Registry in total employment (as calculated using the LFS) is 46 percent. 43 to achieve the relatively weak employment growth remained at 53 percent. For each job gained in net terms, more than thirteen jobs had to be reallocated across the economy (Figure 5.2). Figure 5.2: Job Creation and Job Reallocation Increased between 2002 and 2006 1/ Employment creation and destruction (thousand jobs), Annual rates of gross job creation and destruction, net job 2002-2006. creation, and gross and excess job reallocation, 2002-2006. 50 Creation Destruction Net 40 30 20 39.5 26.4 10 5.5 0 -7.8 -10 -34.3 -34.0 -20 -30 -40 2002-2004 2004-2006 Source: Estimates based on Firm Registry Data. 5.8 Firm survival rates are improving. Jobs can be created and destroyed through the creation or liquidation of firms, or through the expansion or contraction of existing firms who do not go out of business. In both 2002 and 2004, more jobs were created through start-ups than through firm expansion. But in the latter period, firm survival appears to have improved, as the number of jobs destroyed through the exit of firms fell. More jobs were in fact destroyed in continuing firms than in liquidated firms. As is logical, contracting firms are generally the largest firms while the firms that exit remain very small, with an employment of 6 persons on average (Figure 5.3). Figure 5.3: Firm Survival Improved over Time Jobs created and destroyed through expansion, contraction, entry and exit. Average employment size Thousands (average per year), 2002-2004 and 2004- (in number of workers) 2006 25 35 Expanding Contracting Entry Exit 20 2002-2004 30 2004-2006 15 10.8 +4 25 10 6.8 20 5 + 3.6 9.0 6.4 0 15 -7.9 -0.8 -8.7 -5 10 -10 +1.0 5 -9.3 -8.3 -15 0 -20 no change in expanding contracting entry exit 2002-2004: Negative net job growth 2004-2006: Positive net job growth employment Source: Estimates based on Firm Registry Data. 44 5.9 There is a considerable level of firm turn-over in FYR Macedonia, suggesting that reallocation needs remain very important. In 2004-2006, about 24 percent of firms were created and about 14 percent liquidated. In transition countries, churning has generally increased after the economic restructuring process has taken off, as much in response to privatization reforms. As economies have matured, however, entry and exit rates have realigned themselves with levels comparable with more developed countries (World Bank, 2008). FYR Macedonia's level of churning is comparatively high by comparison both with other transition countries and OECD countries. This may be an indication that Macedonia is in an earlier stage of transition than some of the other Central and Eastern European countries, with continued significant reallocation needs (Figure 5.4). Figure 5.4: FYR Macedonia Has Comparatively High Firm Turn-Over Firm entry and exit as % of all firms 25 Entry Exit 20 15 10 5 0 S Finland France ussia stonia kraine R ania R ania lovenia Latvia M 02-04 M 04-06 ungary ortugal U om om R E U K K H P S Source: Estimates based on Firm Registry (FYR Macedonia), World Bank (2008). JOB DYNAMICS AT A DISAGGREGATE LEVEL ­ WHO IS CREATING JOBS AND WHERE? 5.10 Job destruction was increasing in Skopje. As this report shows, regional differences tend to be significant in FYR Macedonia and have also increased over time. Unfortunately, the data do not permit a distinction between different administrative regions; however, Skopje can be separated out from other areas. Consistent with the story emerging from the LFS data, net job creation is shifting significantly outside Skopje, from a high rate of net job losses to a high rate of positive job creation in the period 2004- 2006 (Table 1). In Skopje, job destruction increased, while in other areas, it fell in the same period, at the same time as job creation increased everywhere. Table 5.2: Job Dynamics in Skopje vs. Other Areas Job creation rate Job destruction rate Net job creation Gross turnover Excess turnover 2002-2004 Skopje 0.14 0.19 -0.04 0.33 0.29 Other 0.25 0.32 -0.07 0.57 0.50 2004-2006 Skopje 0.33 0.31 0.01 0.64 0.63 Other 0.30 0.23 0.06 0.53 0.47 Source: Estimates based on Firm Registry data. 45 5.11 Job creation has only been positive in the services sector. The inter-sectoral shifts in FYR Macedonia's job structure are consistent with an economic transition path. Industry, which admittedly is overrepresented compared to the LFS, has seen high job destruction, with massive reallocation above what would be needed to just transfer jobs from one sector to the other, but job creation has increased over time. The services sector has seen positive and increasing net job creation and is the sector with the most efficient job creation, in the sense that the excess reallocation is the lowest. Firms in the agricultural sector saw massive job destruction ­ but it must be remembered that they constitute a negligible share of all firms in the database (Figure 5.5). Figure 5.5: Highest Efficiency in Job Creation in Services Net job creation Excess Job Destruction in High job 120 destruction in Agricultural firms Industry but Higher net job 100 increasing job creation in 80 creation services, highest efficiency in job 60 creation 40 20 0 -20 -40 -60 02-04 04-06 02-04 04-06 02-04 04-06 Agriculture Industry Services Source: Estimates based on Firm Registry data. 5.12 In spite of a general shift towards higher job creation, important sectors saw little or negative job growth. In the manufacturing sector, which accounts for some 37 percent of employment in the sample, there is a shift towards higher job creation and lower job destruction in 2004-2006 compared to the early 2000s (manifested by a shift "south-east" in Figure 5.6 below). However, an exception to this is the textiles industry, which alone accounts for 14 percent of jobs. The textiles industry created virtually no (recorded) jobs in the second period, while gross reallocation remained the same. Within the services sector, sub-sectors have similarly shifted towards higher job creation and lower job destruction. The one exception was transports and communications sector, where both job creation and job destruction increased. Again, this is one of the sub-sectors with the highest shares of employment (around 10 percent). 46 Figure 5.6: A Shift Towards Higher Job Creation in the Manufacturing and Services Sectors ­ Except in the Textiles and Transports Sectors, which Account for Nearly Half of Employment All points below the diagonal indicate positive net job creation. Selected sub-sectors indicated with names Manufacturing and construction: Job creation and job destruction rates, 2002-2004 and 2004-2006 0.60 2002-2004 2004-2006 0.50 Basic metals Gross Job Destruction 0.40 Food Construction 0.30 Textiles Food Textiles 0.20 Construction Basic metals 0.10 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Gross Job Creation Services sector: Job creation and job destruction rates, 2002-2004 and 2004-2006. Selected sub- sectors indicated with names. 0.50 0.45 0.40 Gross Job Destruction 0.35 0.30 Wholesale 0.25 Retail Trsp&Com 0.20 Retail Trsp&Com Wholesale 0.15 Finance Finance 0.10 0.05 0.00 0.00 0.10 0.20 0.30 0.40 0.50 Gross Job Creation Source: Estimates based on Firm Registry data. 5.13 Small firms account for most of job creation. All sizes of firms saw a move towards higher job creation, and with the exception of the very smallest and the very largest firms, higher job creation was accompanied by lower job destruction. Net job creation rates are highest in the micro-enterprises and 47 small enterprises; generally, these firms also have the highest levels of job turnover (far "north-east" in the chart) because they also have high job destruction rates (Figure 5.7). 5.14 Only private sector firms have been creating jobs on a net basis. Private firms became more efficient job creators in the period 2004-2006, retaining the same level of job destruction and increasing job creation. However, job turn-over increased also among state-owned firms. Figure 5.7: Job Creation is Highest among Smaller Firms, and Only the Private Sector is Creating Jobs on a Net Basis Job creation and job destruction by firm size, 2002-2004, 2004-2006 0.50 1000+ 2002-04 0.45 2004-06 0.40 250-500 Gross Job Destruction 0.35 50-250 0.30 5-50 0.25 5-50 0-5 50-250 0-5 0.20 0.15 250-500 0.10 1000+ 0.05 0.00 0.00 0.10 0.20 0.30 0.40 0.50 Gross Job Creation Job creation and job destruction by ownership, 2002-2004, 2004-2006. 0.70 2002-2004 0.60 State 2004-2006 Gross Job Destruction 0.50 Mixed and other 0.40 Mixed and other 0.30 0.20 Public Private Private Public 0.10 State 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 Gross Job Creation Source: Estimates based on Firm Registry Data. 5.15 Importantly, positive net job creation between 2004 and 2006 was driven entirely by low wage jobs, some of which are likely to have been occupied by the poor. Gross job creation increased among all firms, whether they had lower or higher levels of wages. But the low wage firms were the only 48 ones to make a significant move "off" the diagonal in Figure 5.8 below, and towards positive net job creation. By definition, low income persons are likely to have occupied some of these new jobs. Figure 5.8: Significant Net Job Creation Only in Low Wage Firms Job creation by average wage (quintiles). Dark blue: 2002-2004; pink: 2004-2006 0.60 0.50 Gross Job Destruction 0.40 Lowest Second Lowest 0.30 Third Highest Fourth Second Fourth Third 0.20 0.10 Highest 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Gross Job Creation Source: Estimates based on Firm Registry data. 5.16 Only firms with negative productivity growth are creating jobs. From a poverty perspective, more employment in firms with low productivity levels is likely to mean more job opportunities for the poor. At the same time, without a boost in output among these firms, evidenced in positive labor productivity growth, there is no basis for real income growth for the employees. However, sorting firms by average productivity growth reveals that in the groups of firms by first, second and third quintiles of employment, average labor productivity growth is negative, and only firms with negative labor productivity growth (especially the second quintile) are creating jobs on a net basis (Figure 5.9). 5.17 Overall, then, there is some evidence of a positive dynamic underneath FYR Macedonia's slow employment growth in the non-agricultural sector, but there does not seem to be a shift towards better paying or more qualified jobs. Net job creation has increased, witnessing a more efficient job creation process. Small, low wage firms are creating jobs, especially in the services sector. However, since these jobs are not connected with any labor productivity growth, they may not be sustainable over the long run ­ firms need to make a profit to survive over time ­ and do not provide any means for income growth. Inasmuch, the story emerging from the firm data is consistent with the poverty stagnation witnessed in FYR Macedonia, the increase in working poor, and the disconnect between poverty reduction and job creation at a sub-national level. 49 Figure 5.9: Firms with Low (Negative) Productivity Growth are Creating Jobs Firms by average labor productivity level (quintiles). Dark blue: 2002-2004; pink: 2004-2006. 0.60 Lowest 0.50 Lowest Gross Job Destruction 0.40 Second 0.30 Third Highest Third Second 0.20 Fourth 0.10 Fourth Highest 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Gross Job Creation Firms by average growth in labor productivity level (quintiles). Dark blue: 2002-2004; pink: 2004-2006. 1.00 Lowest 0.90 Lowest 0.80 Gross Job Destruction 0.70 0.60 0.50 0.40 Highest 0.30 Highest 0.20 Fourth Second 0.10 Third Second Fourth 0.00 Third 0.00 0.10 0.20 0.30 0.40 0.50 Gross Job Creation Source: Estimates based on Firm Registry. 50 CHAPTER 6. CONCLUSIONS: SUSTAINABLE JOBS AND SAFETY NETS 6.1 The Government of FYR Macedonia has the opportunity to transform the dynamics of economic growth and poverty reduction from one where the poor did not share in growth to one that utilizes the poor's labor assets. The current global crisis may delay measurable benefits of adopting policies that would have yielded returns in the near term. Nevertheless, these policies remain important to pursue because the cost of inaction could potentially keep a significant share of the population mired in poverty and delay FYR Macedonia's overall social and economic development. Moreover, the greater danger is that it would potentially slow down convergence with the rest of the EU in the long-term and even with the New Member States of the EU in the medium-term. 6.2 The period 2002-2006 represents a difficult period in FYR Macedonia's development. It was characterized by moderate economic growth and low-quality job creation. The economy grew at rates below those experienced in regional comparator countries. Many of the jobs that were created in FYR Macedonia implied underemployment, precarious work, and low and even falling pay ­ rather than economic dynamism. Moreover, social welfare programs (as distinct from social insurance) were not adequately targeting the poorest in society who were unable to meet their basic needs. Figure 6.1: Very Fast Job Growth would be Needed to Reach the Target Employment Rate of 48 Percent by 2010 Annual growth rates 2004-2006 Job growth rates needed to reach target for employment rate 14.0 14.0 11.9 12.0 12.0 10.0 10.0 8.5 8.0 8.0 6.0 5.5 6.0 4.4 4.0 4.0 3.0 2.2 2.0 2.0 0.0 0.0 All jobs Paid jobs Paid non-ag jobs All jobs Paid jobs Paid non-ag jobs Source: Estimates based on LFS. 6.3 The Government of FYR Macedonia in recognition of the importance of employment creation for accelerating the pace of economic development as well as the population's welfare has set itself the ambitious target of increasing employment rates to 48 percent by 2010, up from 40 percent in 2006. Achieving a jump in eight percentage points in four years would require significant, not to say heroic increases in job creation, as shown in Figure 6.1Figure 6.1. Between 2004 and 2006, the Macedonian economy created 47,000 jobs, of which 18,000 were 51 unpaid jobs. Assuming that working age population grows by about 0.6 percent in the period 2006-20101, some 140,000 jobs additional job would be needed, and even with a stagnant working age population, 120,000 jobs would be needed. These numbers translate into job growth rates of 5.5 and 4.9 percent per year. Moreover, the rate of employment creation would need to be even higher if the government targets the creation of paid jobs in the non-agriculture sector. THE WAY FORWARD 6.4 The government is pursuing a broad agenda and must do so in the face of both short-term and long-term challenges. However, from the perspective of reducing poverty, there are certain priority areas, that if tackled, could help to better alleviate poverty. A poverty assessment can only point to areas that the government may wish to focus its efforts on based upon what the analysis of survey data has revealed. But the nature of this type of report is that it cannot give advice on what specific policies the government should consider but, at best, what areas to target. However, given the wealth of information in other World Bank reports, the recommendations below draw upon these as well to lend some direction to the areas of focus. Below, are five important areas that the government may wish to develop strategies in order to have a greater impact on poverty. 6.5 First, given the strong relationship between economic growth and poverty reduction globally, policies which accelerate growth to above 5 percent per annum have the best chance of reducing poverty. Having policies which enhance shared growth is the most effective ­ and most cost efficient ­ instrument for raising the greatest number of people out of the poverty. The recent World Bank report, FYR Macedonia ­ Moving to Faster and More Inclusive Growth: A Country Economic Memorandum, discusses the benefits of increasing annual average growth by over two percentage points by implementing a series of structural reforms. Continuation of reforms in the judiciary, contract enforcement and land titling is necessary to further strengthen creditor and property rights. Reforms that would promote and strengthen competition (i.e. further trade and market liberalization, stronger regulatory bodies etc.) are also a key requirement for stronger economic growth. Public administration reform are needed to reduced the administrative costs of the public sector and promote investments. Moreover, World Bank analytical work on labor markets pointed out that the responsiveness of employment to growth was low in part because of the high cost of hiring labor due to the excessive social contributions. At the same time, the mix of macroeconomic policies should support macroeconomic stability and sustainability of fiscal and external balances. 6.6 Second, the government should continue its focus on enhancing competitiveness and improving the business environment despite the challenges posed by the current global crisis. Areas of importance for private sector development are easing regulations for closing a business, especially given the economic contraction and permits and licensing; developing a system for retraining laid-off workers, again important in today's environment; infrastructure development which not only yields medium- to long-term benefits but can also generate employment in the context of the economic crisis; and ensure that regulatory procedures are simplified. A second important angle is to pursue is that greater exploitation of the enterprise data could provide greater insights into which firm characteristics are most strongly associated with high net job creation rates. Through better understanding to what extent job creation and destruction in firms is driven by wages, productivity levels, and industry or service sector, the government could 1 The growth rate of 0.6 percent per year is taken from the projections in the ILO's Economically Active Population Estimates and Projections for 2006-2010. 52 identify more targeted and specific policies to accelerate job growth. Thus, it would be useful to complement the policy advice from the CEM with a better understanding of the micro determinants of growth.2 6.7 Third, policies which improve the coverage and quality of the education system and enhance the population's marketable skills can reduce poverty in the medium- to long-term. As discussed in this report, jobs alone are insufficient to decrease poverty, since there must be productivity growth to sustainably raise wages over time, especially at the lower end of the income distribution. While access appears to be decent, quality and affordability remain a concern. The government has already taken significant steps to improve education, e.g., making secondary education compulsory and introducing a Conditional Cash Transfer Program (supported by the World Bank) to encourage school attendance among recipient households. Implementation of a Professional and Carrier Development for Teachers should produce better trained and more motivated teachers. In addition to increasing instructional time at early levels of education, wider use of pre-primary education needs to be promoted. To maximize the benefits of the recently purchased IT equipment that authorities needs to devise content and ensure quality teacher training. The government must remain vigilant to set national policies (such as minimum service standards) and ensure these are complied with at the subnational levels given the decentralization that is occurring in the education system. The assessment and quality assurance systems at all levels of education need to be well-established. Though education reforms both in school and colleges or universities are being introduced,3 these will help new labor force entrants. Given the large stock of working aged adults, a significant share who have little or no education and are also poor, the government may wish to accelerate establishment of the adult education mechanisms, anticipated in the law for adult education. At the same time, support to vulnerable groups and students with special needs (provision of text books, expanded use of CCT etc.) should ensure full enrollment and avoid drop-outs. Similarly, budget financing of core adult education programs (based on regular labor market needs assessments) can help adult poor without skills. 6.8 Fourth, improving the social welfare system, especially the social assistance program, can help to reduce poverty by expanding coverage of the poor while reducing leakage to the affluent ­ all within the same budget envelope. The Government of Macedonia has already been bold in initiating reform of the safety net. Though instruments exist for transferring assistance to the poor, the system needs to be upgraded to standardize the application of the eligibility criteria, improve targeting, and reduce leakage to the non-poor. The government is tackling these issues and it is important to ensure that the system is regularly evaluated and monitored to measure improvements and remaining areas of weakness. The importance of having an effective social welfare system is particularly important given the high level of under- and unemployment in FYR Macedonia combined with slow job creation rates. The recently revised Law on Social Protection opens up possibilities for significant reforms. Still, adoption of a by-law that would define "countable income" to include all transfers is a critical element for its effective implementation. Further policy reforms may also be needed such as expansion of the use of CCT to other sectors (including labor market activation). The institutional capacities should be strengthened as well: 2 The country has made significant progress relative to other countries according to the Doing Business report, 2010. It has improved its rank to 32 compared to 69 in the previous year. Progress was noted in easing regulation for starting a business, closing a business, dealing with construction permits, protecting investors, employing workers, and registering property. 3 One of the recent and important reforms in the education sector recently introduced is the nine-year basic elementary school with compulsory English and informatics courses. 53 the ongoing activities for the establishment of the Centralized Database on Programs and Beneficiaries are important for this but the capacity of implementing agencies needs to be upgraded also. The workload of the Centers for Social Works is rather big while the internal capacity of the Ministry of Labor and Social Policy to assess the effectiveness of the programs remains limited. Closely related to the topic of social safety net, the large informal economy resulted in relatively low coverage of the pension system. As a result, over the medium- to long- run, a growing number of citizens will not meet the minimum requirements for receiving a pension and would have to rely on the social safety nets to provide their subsistence and the system should initiate activities to accommodate this. 6.9 Fifth, in order to better monitor progress in poverty reduction, FYR Macedonia should adopt an absolute poverty line and strengthen its household survey instrument. The government currently uses a relative poverty line to monitor poverty; however, this instrument does not measure absolute welfare. Thus, annual changes in relative poverty do not yield sufficient information to help the government monitor the impact of its policies or guide reforms. Moreover, there are some key issues that need to be addressed in the design of the household survey which are modest in scope but could yield large benefits in terms of quality and relevance. 54 REFERENCES Angel-Urdinola, D. and Macias-Essedin, V. Macedonia Employment Profile, 2004-2006, The World Bank, 2008 CEA: Center for Economic Analysis, Report on the Labor Market in Macedonia, March, 2005 Davis, Steven J & Haltiwanger, John C, 1992. "Gross Job Creation, Gross Job Destruction, and Employment Reallocation," The Quarterly Journal of Economics, MIT Press, vol. 107 (3), pages 819-63, August. Eurostat: People outside the labour force: the downward trend continues - Issue number 122/2007 Migration Policy Institute, 2007: Macedonia: At a Quiet Crossroads, By Joanne van Selm Rutkowski, J., Walewski, M., 2007. Taxation of Labour. In Gray, C., Lane, T., Varoudakis, A., (eds.): Fiscal Policy and Economic Growth: Lessons forEastern Europe and Central Asia: World Bank, Ch. 9:281­313 State Statistical Office, Household Budget Survey 2002, 2003, 2004, 2005, 2006 State Statistical Office, Labour Force Survey 2004, 2005, 2006, 2007 State Statistical Office, Statistical Yearbook 2004, 2005, 2006 TransMONEE 2008 database, UNICEF Innocenti Research Centre, Florence. UNDP, 2008: http://hdr.undp.org/en/statistics/ World Bank (2003), FYR Macedonia Country Economic Memorandum: Tackling Unemployment. World Bank (2008), Unleashing Prosperity: Productivity Growth in Eastern Europe and the former Soviet Union. World Bank (2009), Macedonia Moving to Faster and More Inclusive Growth: A Country Economic Memorandum. . 55 ANNEX A TO FYR MACEDONIA POVERTY ASSESSMENT IMPROVING THE MACEDONIA HOUSEHOLD BUDGET SURVEY The Government of FYR Macedonia has been forward looking in the continued support given to the household budget survey. These surveys are carried out annually by the State Statistical Office of the Republic of Macedonia. They are generally comparable over time and collect data on expenditures and income. They are nationally representative and are based upon the updating of the census figures. The household budget surveys are the main instrument for measuring poverty in FYR Macedonia. Consequently, it is important to ensure that their quality continues to be improved and the questionnaire is designed to help with improving the relevance of policy choices. In this annex we discuss three shortcomings of the survey which if rectified could have an important impact on the overall quality of the poverty estimates. The average household size differs from the census and increases over time in the household surveys. According to the 2002 census, the average household size in Macedonia was 3.5. Table 1 shows the average household size per year based on the household surveys and applying the weights provided by the State Statistical Office (SSO). The HBS 2002, 2003, and 2004 show a household size of about 4 which increases abruptly in 2005 and again in 2006 to 4.65. Given this jump in the household size, the poverty numbers are no longer comparable between 2002-2004 and 2005-2006. It appears that the increase is driven both by the change in the number of children between the ages of 5-14 and adults between the ages of 15 ­ 64 years. This change in the household size also makes the poverty numbers unreliable. For example, using the poverty line computed in 2002, the share of the population living below the absolute poverty line increased between 2004 and 2005 from 20.4 percent to 25.7 percent of the population. Table 1: Household Budget Surveys' Average Household Size and Composition, 2002-2006 Number of Household Children Children under Adults age 15 Adults age 65 Year size under age 5 age 15 and more and more 2002 4.10 0.20 0.81 3.29 0.48 2003 4.02 0.21 0.78 3.24 0.46 2004 4.00 0.23 0.82 3.18 0.42 2005 4.49 0.25 0.97 3.52 0.45 2006 4.65 0.21 0.97 3.68 0.43 Source: World Bank staff computations based on HBS 2002-2006. 57 There is an anomaly in the consumption per capita data around the 20th percentile of the population in 2006. The per capita consumption (in constant terms) increases abruptly at the bottom 20th percentile and is above trend until the 25th or 26th percentile (see Figure 1). As an illustration, in 2004, the increase in consumption per capita between the 19th and 21st percentile was 5 percent; while in 2006, the increase in real consumption between the 19th and 21st percentile of the population was 18 percent. This unexpected sudden increase near the poverty line reduces total absolute poverty. This change in consumption per capita is unusual not only for Macedonia (as compared to past years), but is seldom seen in other countries as well. Figure 1: Average Consumption per capita by percentile, 2004-2006 /a 120 100 in dinars per capita 80 60 Year 2004 40 Year 2005 Year 2006 20 0 1 3 5 7 9 11 13 15 17 19 21 23 25 Population percentiles Source: World Bank staff computations. a/ Only the consumption per capita averages are shown for the bottom 25 percentiles of the population. The survey questionnaire's recall period for certain expenditures is too short for capturing consumption fully for households. The recall period for all expenditures is 15 days. Thus, monthly payments for electricity, water, and other utilities, are not captured for all households but only for those who pay their monthly bills during the period covered by the diary. As another example, expenditures on school supplies may occur at only certain times of the years indicating the need for a much longer recall period. In the case of utilities, this means that for almost half of the sample, consumption is under estimated due to the format of the survey instrument. Though this does not undermine the information on household expenditures for the national accounts, it does adversely impact the accuracy of the poverty estimates. However, this does not mean that we can state that the poverty numbers are upwardly biased since the lack of information on monthly payments also biases the estimation of the poverty line. 58 ANNEX B TO FYR MACEDONIA POVERTY ASSESSMENT Table 1 - Overview of Main Cash Benefits in the FYR Macedonia, 2007 Number of Targeting Income Monthly Eligibility Spending % of beneficiaries threshold benefit (mil. GDP amount MKD) Individual. House- MKD holds MT 13.5 -33,34 1,825 Able to work, % avg.w MKD(single) ­ but unable to Social Financial 4,507 MKD support Assistance (SFA) (5+ member themselves HH) 1,692 0.50 219,063 62,997 MT Income per 2,704 Unfit to work fam. MKD(single) ­ Member 5,408 MKD Permanent should be (benef. +2) Financial less than Assistance mo. benefit 195 0.06 5,033 MT 26% avg. w; 176 ­ 4,216 Unable to 23% for MKD perform basic partial (complete); living disability 135 ­ 3,371 functions Assistance and MKD (partial older than 26 care disability) 829 0.25 19,483 Non-MT None Current avg. w Civil Disability * disability coefficient 43 0.01 371 One off financial Non -MT None max 2 mo. avg. Exposed to assistance net w sudden risk 39 0.01 1,616 MT + 16% avg. w; 621 MKD (up Poor children Child allowance categorical 32% for to 15 yrs.); 987 with a single parent MKD (15-18 working yrs) parent 265 0.08 32,392 19,006 Special Allowance Non-MT None 27% avg. w Disabled 209 0.06 4,843 4,747 Assistance for new MT up to 200% born babies of avg.w 22 0.01 505 Other 1 74 0.02 431 Total 3,368 1.00 283,737 Note: MT=means tested. Avgw=weighted average. Source: Project Appraisal Document on a Proposed Loan for FYR Macedonia Conditional Cash Transfers Project (World Bank, May 2009), page 24. 1 Including for example, financial assistance and housing for young adults (18-26) without parental care, equalization allowance for care of a disabled child (shorten working hours), etc. 59 ANNEX C TO FYR MACEDONIA POVERTY ASSESSMENT Constraints to Labor Force Participation in FYR Macedonia: A Qualitative Approach SUMMARY The labor market non-participation rate in Macedonia is relatively high and encompasses about 37.8% of the working-age population (aged 15-64). This phenomenon is mostly evident in women, 50.8% of which are found not to seek employment. The majority of them belong to the non ethnic-Macedonian communities and have primary education or less. The following study aims to analyze the relative influence of the objective (infrastructure, education) and subjective (personal beliefs, traditional norms) factors as possible constraints for participation of the low- skilled women in the labor market. In order to detect the specific setbacks for labor market participation faced by the low-skilled non-participating women, focus-groups were conducted with the aim to compare and contrast their opinions and beliefs related to employment to those of educated non-participant women, unemployed and non-participant males. For this purpose, 23 focus groups were conducted in 8 municipalities1 in the period between 2nd and 23rd April, 2008; encompassing a total of 233 individuals. The method of focus groups provided an in-depth qualitative data on the issue, different from the data provided by the typically used surveying methods. It enabled for certain issues (questions) to be raised several times, in order to "unravel" the initially provided socially desirable responses. General Findings The general reasons for inactivity identified in both men and women from different ethnicities were caused by their discouragement that they would be able to find employment because they "lack personal connections" and are not affiliated with the political parties in power. Apart from the general discouragement, several specific participation constraints were identified among women from different ethnic communities. While the inability to find employment because of reasons such as age and marital status were detected to be the most dominant discouraging factors among ethnic Macedonian and ethnic Serbian women, for the ethnic Albanian women it is the traditional norms related to female employment. The findings indicate that the low level of education influences the non-participation among ethnic Albanian, ethnic Turkish, and some Roma women, however, not on its own, but in combination with the conservative community values. Specifically, the traditions act by causing different "objective" constraints for inclusion in the work force. For example, many of these women do not continue their education after primary school because the community does not expect them to need additional education for being housewives. Also, they are given too 1 Skopje (Saraj, Shuto Orizari, Centar), Kumanovo, Tetovo, Gostivar, Sv.Nikole, Bitola, Debar and Struga 61 many responsibilities around the house because of the belief that this should be their primary role, and other such examples. This causes many to not consider the idea of working and completely accept the traditionally established norms. More educated women from these communities also face the "traditional" setbacks for participation, but they are less "tight" because of their raised awareness regarding the need to work and the wider array of professions they can choose from. Accordingly, the motivation to work is strongly related to the level of education within ethnic Albanian and ethnic Turkish women. Specifically, the ones with lower education typically do not want to work (or at least not in the formal labor market) or do not believe they would find a job with their qualifications. On the other hand, the more educated would like to work in the case certain conditions are met: state-sector position with full benefits and a good salary (over 200-250 EURO). These findings suggest that in addition to the clear presence of restrictive attitudes on part of their marital partners and wider families (external influence), the motivation and wish for obtaining employment (personal factors) are also weaker in this group. Therefore the highly educated women of these two ethnicities find that the most desirable type of job in society (state-sector) is actually a minimum threshold for them. It is interesting to mention that the reference wage for most ethnic Turkish and ethnic Albanian women, regardless of their level of education, appears to be higher than the current average wages in the labor market. As for the other communities (ethnic Macedonian, ethnic Serbian and partially Roma) the level of education was not perceived as a determining factor for non-participation. Actually, there were only subtle differences regarding the economically active and inactive women, regardless of their educational attainment. The vast majority of them stated they want to work but feel discouraged by the current situation in the labor market. However, while some of them have stopped actively looking for a job, others have not done this yet (although there is a high possibility of this happening if they remain unemployed in future). It can be concluded that the traditional role of the patriarchal "breadwinner" and head of the family is still prevalent in Muslim communities (especially with ethnic Albanians) and is embraced strongly by both genders. These values are held rather resiliently by some men and most of the women do not appear to mind them too much. Hence, they accept them as a composite part of their culture, which partially conflicts with their intentions of overcoming the traditional mentality. Even though these norms are more strongly embraced by the low-skilled women, many highly educated Muslim women, though striving for greater emancipation, still find it hard to give up on their role of being "the perfect housewives". This is mostly due to the fact that the respect for the elderly family members is one of the primary values, and "rebelling" against their decisions (for e.g. the decision that the wife should not work) is considered as highly unacceptable. When it comes to money management, the vast majority of ethnic Albanian participants (from both genders and with different educational attainment) agreed that the eldest family members (principally male but if there is no male than the eldest female) are in a position to manage the money and decide how it is going to be used, including the decision on how it is going to be distributed among other family members. This sentiment was also shared by many of the Turkish participants, who explicitly stated that this is the situation in "our Muslim households". These practices are similar regarding the management of productive assets. Even though women may be consulted regarding certain decisions, they rarely have the final word (unless they are the oldest family member). The educational attainment of the women does not make any difference, since these norms are rather strongly internalized by all. 62 Participants from other nationalities (Macedonian, Serbian and Roma2) generally shared their practices of family consultation regarding the management of finances in which every member's voice is heard and respected. However, a smaller number also felt that the position of bringing income into the family partially dictated the position of individual family members in the decision-making. Therefore, the ones that did not bring in critical amounts of income into the family budget found it natural that they were not central in the decision making. Similar practices of decision-making were found regarding the management of productive assets. Despite the fact that the property was typically registered on the name of the male family members, ethnic Macedonian women generally agreed that the decisions regarding the way it is going to be used are made in collaboration between the family members Despite the actual non-participation in the labor market, the female participants generally expressed a will for further professional enhancement of skills, but they also perceived many obstacles for this. For the ethnic Macedonian, ethnic Serbian and Roma women with different educational attainment it was mainly the financial burden, while for the ethnic Albanian women also the traditional values, according to which a woman should primarily stay at home. The low- skilled ethnic Albanian women also rarely perceive the value of attaining additional skills/education, since they do not believe they would need them in their everyday life as housewives. The lack of information regarding the possibilities for professional enhancement is also evident, along with the lack of belief that they will be able to enroll in the state-organized programs. In general, the low-skilled participants (especially from Albanian and Turkish origin) felt most comfortable visiting trainings organized by local female Non-Governmental Organizations (NGOs) and this practice should be further developed and utilized. Policy Recommendations In regards to the summarized conclusions, several possible solutions for overcoming the labor market non-participation will be analyzed, selected on the basis of their feasibility having the current political and economic situation in the country in mind. The recommendations primarily refer to increasing the educational opportunities of women and facilitating the implementation of certain existing laws. Additionally, certain problems resulting from the use of the Labor Force Survey (LFS) are discussed and recommendations on how they could be overcome are presented Recommendations for increasing economic participation in low-skilled women from minority communities: Since the educational attainment was found to be a rather important determinant for the inclusion of women into the labor market, obtaining further education appears to be the most "controllable" factor (i.e. easier to influence). However, the main challenge is to encourage girls/women to actually complete secondary school and empower them to seek employment. In addition to this, there should be special programs which would provide adult education (primary and secondary) ­ ideally free of charge or with symbolic contributions - to individuals who have not completed their formal schooling. 2 The case of the Roma is rather mixed, as there are strong Roma traditional values which are sometimes combined with Muslim traditions. And then there is the additional stream of Roma culture. Members of the household even though they may live together in the same house, actually live quite independently from one another when regarding money. There is no clear management of finances but all members of the household spend what they earn when they have money. 63 Implementing these activities in the male-dominated cultures will be difficult. The family patriarch has a key influence in many areas (who in the family is going to be educated further, who is going to work and which professions are "suitable" for women), and active involvement of the patriarch is necessary in order for such programs to yield results. However, considering the fact that there is a will among many women in these communities (especially younger ones) for further education and advancing their skills, they could act as mediators for reaching out to the low educated women from traditional families. The female NGOs also appear to have the capacities to get in touch with some of these women and work with them on changing the traditional outlook on female education and employment (both in the individual and the family circle). Moreover, a systemic solution for this problem could be achieved through making connections between the women with specific skills (which are not officially recognized) and the businesses in need of these skills. Here, certain NGOs in cooperation with the State Employment Agency can act as a "link" between the both parties by identifying the women and the skills they possess and communicating and matching them with the businesses' needs. Hence, they would be generating income by, for example, selling their hand-made products and not having to leave their home. Besides this, a part of the solution might be to train these women to start up their personal home-based businesses, as well as expand the access to credit. It is important to mention that low-skilled women from Muslim communities (and their families) do not actually object working per se, since they work in their homes all the time, but feel less comfortable working formally outside of their house and their community. Hence, more flexible ways of utilizing and upgrading their skills, as the one mentioned above, should be explored. For example, some women can lead a day-care center in their homes, since the care of children is one of the skills most report to posses and like doing. Not only would this provide employment for them but at the same time enable the other mothers from the community to seek employment, knowing that their children are in safe hands. Low education levels constitute one of the biggest problems for the appropriate inclusion of the Roma community in the labor market. However, the reasons for this are not only traditional, but very often due to financial constraints and the tendency to involve the children in the (informal) labor market at a young age. Despite this, the primary education of Roma has received the least financial assistance in the actions taken in the frames of the Roma Decade for improving the educational attendance of Roma children. Hence, providing financial assistance to Roma families for completing primary education (of children and adults) should be the initial policy concern, and the practice of relating the social assistance to families with the enrollment and continuation of the formal education of children may be a good incentive for parents. In thos regard, while punitive measures (for example fines for not sending ones children to elementary or secondary schools) are important and should sustain, they are not enough and need to be complemented with other motivating sustainable measures (e.g. free books, school material, student allowance etc.). Finally, organizing buses which will transport only women and their children might prove to be a good incentive for motivating some of them to get out of their immediate communities more in order to continue their education or seek employment. General Recommendations for Increasing Economic Participation The policy of "Equitable and Appropriate Representation of Communities in the Public Administration", which is being implemented during the past several years is likely to partially 64 solve the problem of non-participation (occurring as a result of discouragement) among some highly educated minority women. Moreover, the state-sector positions are expected to be suitable for the women that consider these types of jobs to be appropriate for them. However, minorities that do not have sufficient numbers of highly educated cadres (such as Roma) are again expected to encounter further problems in this regard. Nevertheless, despite the belief that the education would increase their employment prospects, the expectations from the labor market of many young and highly educated ethnic Albanian and ethnic Turkish women tend to be quite rigid. Since typically only state-sector employment is seen as an appropriate, it makes them quite selective regarding the type of job they would accept and the acceptable wage minimums. However, some are aware that these beliefs are old-fashioned and need to be changed, which speaks of their readiness to change the outlook on the jobs that are considered to be adequate for women. They may serve as catalysts for other women to "break" the traditional mentality and become aware that the fast-transforming labor market is in a need of flexible labor force that can adjust to the changing environment. For individuals that have been unemployed for more than 10 years, there is a risk that the skills they have gained during the schooling process or pervious employment are probably outdated and most importantly forgotten because they haven't been practiced for quite some time. The Employment Agency is currently organizing trainings in IT skills, English language and specific professions, but they are usually out of reach to the most "vulnerable" groups, such as the bulk of the non-participating women. They are rarely acquainted with the possibilities being offered and since they typically take place outside of their "immediate community", the possibilities for attending such trainings are very limited. Perhaps, in collaboration with the Worker's Universities around the country (or another similar institutions) the Employment Agency would be able to provide specific re-qualification and/or additional qualification programs for unemployed (including the inactive) by actively approaching the people and inviting them to select training according to their education, skills and preferences. Efforts should also be made to attract more mail attendants to the professional trainings by emphasizing that it should lead to better employment prospects. In addition, for individuals who have specific skills related to crafts (e.g. knitting, embroidery etc.) but do not have specific educational/professional qualifications, there should be system of accreditation developed to provide some kind of formal document. Also, training on developing job searching skills would be very useful for all. Many appear to be unable to "sell" themselves on the current labor market, despite the possession of sufficient professional skills. As a result of the constant rejections by employers, they feel victimized by the institutional policies. Furthermore, the labor inspections need to be strengthened and become more effective in order for the people to regain their trust in the rightful functioning of the institutions. Also additional independent monitoring mechanisms (independent from the courts and in cooperation with NGOs) could provide another avenue of action for citizens facing discrimination. Establishing cheaper procedures for applying in state institutions would also encourage many unemployed (some of which are inactive) to apply. Specifically, original and notarized documents could be requested only by the short-listed applicants, and not by everyone applying. Finally, more flexible ways of employment accompanied with minimal taxes for employees working less than 4 hours a day (i.e. less than part-time employment), should be reviewed. This 65 would hopefully legalize certain informal sector jobs (e.g. manual workers, cleaners etc.) and at the same time enhance the respecting of certain workers' labor rights since the high tax wedge for low wage employments has been considered by experts (e.g. Rutkowski and Walewski (2007)) to be a hindrance for the formal employment of low-skilled workers. Recommendations for Adjusting the LFS Since the validity of the LFS as a screening method for the economically inactive population proved to be problematic to a certain extent, an adjustment of certain questions should be taken into consideration. Firstly, although during the screening phase many of the male participants provided answers to the LFS which indicated that they are inactive, were actually found to work as manual laborers for a daily wage (although not necessarily during the previous week). Secondly, a certain number of the participants who reported they haven't looked for a job during the past month during the discussion stated they are actually looking for a job all the time. Specifically, they "keep their eyes open" in case an opportunity comes up, even though they may not take any active job-search method measure. Hence, Probably, the definition of "inactivity" provided in the LFS needs to be adjusted. First, an extension of the reference period during which a person has not conducted any income- generating activity might be considered. The current period of one week appears to be rather short. Secondly, a rephrasing of the question "Have you been seeking employment during the previous 4 weeks" needs to be considered because of the subjective interpretation of the terms "seeking" and "employment". The first term is usually understood as active job search and the second is typically identified with a regular full-time employment. Moreover, the irregular engagements utilized for income generation are usually not considered as "employment". 66 INTRODUCTION Despite economic growth in recent years and achieving an EU candidate status, Macedonia lags behind tackling the problem of structural and long term unemployment. More importantly low level of social development, conservative beliefs and what can generally be described as "traditions" affect the size of the labor force as many individuals (women in particular) do not even search for jobs and meaningful employment. According to the LFS criteria, these women are considered to be economically inactive. The vast majority of them are "officially" unskilled (i.e. have no education or only have primary education) and belong to the minority (primarily Muslim) communities. The main aim of this analysis is to understand in more detail the reasons why the phenomenon of non-participating in the labor market prevails among unskilled women from minority ethnicities in Macedonia. For this purpose, by conducting focus-groups, we have attempted to explore and analyze the relative influence of the objective (institutions, infrastructure, and education) and subjective (personal beliefs, traditional norms) factors as possible constraints for participation of these women in the labor market. In order to detect the setbacks for participation faced by these groups of women in particular, the study attempts to compare and contrast their opinions and beliefs related to employment to those of non-participating women with higher (secondary and university) education and unemployed women with different educational characteristics. Therefore, we have placed both participating and non-participating women in the same group with the aim to facilitate the discussion and provide for a dynamic expression of diversified and potentially conflicting opinions. The idea was to use these mixed groups to assess to which extent these barriers to female participation in the modern Macedonian economy are shared and/or unique to certain groups of women. In addition, several male focus-groups were conducted for two reasons. Firstly, to look at the male opinions on issues such as female employment and decision-making and secondly, to assess the gender component of unemployment by exploring and contrasting the reasons for non-participation in the labor market stated by men and women. This has enabled examination of the problem from different angles, stimulated fruitful discussion and provided extensive insight into the challenging issue of constraints to labor force participation of unskilled ethnic Macedonian women belonging to minority groups. After taking a closer look at this issue, an attempt was made to pinpoint the possible avenues for improving the present circumstances. 67 BACKGROUND BRIEF REVIEW OF THE LABOR MARKET IN MACEDONIA During the past 10-15 years Macedonia has often been singled-out from the other countries from South East Europe as having a very high unemployment rate. Though the general opinion of the public is that the employment rate experienced a sharp fall during the transition, World Bank (2003) found that that the transition has actually contributed with only 8 percent increase in the overall unemployment, whereas the stagnant situation of the long-term unemployed (persistent even before the transition) represents the lion's share of the overall unemployment rate. The study also argues that unemployment, in the sense of complete inactivity is rare and that unpaid work on family farms, petty trade, and other casual employment has cushioned the impact of the gradual disintegration of the formal job market for many Macedonians. Overall, the labor market is "static" because transition (levels and flows) between different employment statuses is stagnant: the employed remained employed and the unemployed remained unemployed for long periods of time (see Angel-Urdinola and Macias, 2008). The labor force participation in the last several years has been low for Europe and Central Asia countries (ECA) standards, with only 58 percent to 62 percent of the working-age population being economically active (see Table A1). The employment rate is very low and at times almost equal to the unemployment rate (according to International Labor Organization (ILO) standards). The size of informal employment has not been officially determined, but is estimated to comprise 30% of the unemployed people3 and to account for about 35% of GDP in 2004 (CEA, 2005). Table A1: Labor Market Indicators, Adult Population Aged 15-64, 2004-2007 Rates (%) 2004 2005 2006 2007* Activity rate 58.8 60.7 62.2 62.7 Inactivity rate 41.2 39.3 37.8 37.3 Employment rate 36.8 37.9 39.6 40.7 Unemp. rate according to ILO 37.4 37.6 36.3 35.1 Long term unemployment 85 86 86 82 Source: SSO and Employment Agency. * Data for 2007 refer to the fourth quarter since the 2007 LFS was not available during the time of preparing the analysis The largest number of unemployed people (over 80%) belongs to the group of long-term unemployed (defined as those who have been in unemployment for more than 1 year), part of who are believed to be discouraged workers that have stopped actively seeking employment. In addition, data from the National Employment Agency indicates that about 30% of the registered unemployed stated they only register for the health benefits and do not actually seek employment. 3 According to ILO Bureau of Statistics on the basis of official national data 69 THE ISSUE OF NON-PARTICIPATION IN THE LABOR MARKET Participation among the working age population in Macedonia in 2005 was slightly below that of ECA (65% in ECA vs. 61% in Macedonia), which occurred mainly due to lower than average female participation rates. Labor force participation rates among women are low for ECA standards and this makes Macedonia third country in the region (after Turkey and Kosovo) with the lowest female participation rate (see Table 2.2). The LFS data indicates that lower female participation rates are mainly driven by very low levels of participation of young rural unskilled women. In particular, the labor force participation of women from rural areas has been 39% in 2006. Among the reasons of female inactivity, household responsibilities are the most important. Approximately 55 percent of women who are not in the labor force are housewives. However, this proportion varies by educational level. In fact, 68 percent of women having primary or less education are out of the labor force due to housewife responsibilities, while the same share is only at 21 percent among women having higher or university education. (Angel-Urdinola and Macias, 2008). Table A2: Labor Force Participation Rates, by country, by gender Labor Force (15-64) Country Male Female Total Albania 75.7 54.7 65.1 Armenia 65.9 55.4 60.2 Azerbaijan 78.1 66.2 71.9 Belarus 72.3 66.4 69.3 Bosnia and Herzegovina 78.3 70.5 74.4 Bulgaria 62.6 52.4 57.4 Croatia 71.0 57.5 64.2 Georgia 76.1 52.4 63.7 Hungary 66.8 53.5 60.0 Kazakhstan 80.1 73.6 76.7 Kosovo 70.8 33.7 52.3 Latvia 71.9 63.0 67.3 Lithuania 72.4 65.9 69.0 Macedonia, FYR 75.0 49.2 62.2 Moldova 76.0 65.4 70.5 Poland 68.8 57.6 63.2 Romania 69.5 55.3 62.4 Russian Federation 75.3 67.1 71.0 Slovak Republic 76.4 62.4 69.4 Turkey 76.0 27.2 51.9 Ukraine 72.4 62.9 67.4 Uzbekistan 75.7 60.6 68.1 ECA Average* 73.0 57.8 65.3 EA-13 78.6 63.9 71.3 EU-27 77.6 63.6 70.6 Source: WDI Dataset for year 2005. * Un-weighted ECA average. Indicators for FYR Macedonia from LFS 2006. Indicators for EU-13 and EU-27 from Eurostat: EU LFS for 2007. The non-participation rate in Macedonia is relatively high and encompasses about 37.8% of the working-age population (aged 15-64). Even though it has reduced slightly during the past four 70 years (see Table 2.1 above), it is still well above the EU-25 average for 2006 - 29.4 % (36.7% for women and 22.0% for men), (Eurostat, 2007). Table A3: Inactivity Rates in Macedonia According to Age and Gender (2006) Total men women Inactive persons (15-64) 537214 180110 357103 rate 37.8 25.0 50.8 Age 15-24 209209 96938 112271 rate (%the age group) 61.1 58.0 70.7 25-34 72412 16086 56325 rate (%the age group) 23.4 10.2 37.2 35-44 58771 9096 49675 rate (%the age group) 19.8 6.0 34.0 45-54 71467 15454 56012 rate (%the age group) 25.2 7.0 40.0 55-64 125355 42536 82819 rate (%the age group) 61.2 43.2 77.7 Source: SSO, LFS, 2006. The share of prime-age population (aged 25-54) in the total inactive population (aged 15 and over) in Macedonia is 27.8% and incorporates 15.3% of the prime-age men and 35.1% of the prime-age women. As elsewhere in the world, non-participation is a phenomenon most evident within women out of prime-age Expectedly, the non-participation rates decrease as the level of education increases. However, an interesting gender-related trend can be noticed from Table A4. Namely, the female non- participation is higher that the male one in each group, except in women with higher or university education. This indicates that the economic inactivity among women should be primarily seen as an educationally-dependent phenomenon. Table A4: Inactivity Rates in Macedonia According to Educational Attainment and Gender; Population 15 Years and Older (2006) Level of education total male female No education or incomplete primary 177702 46189 131513 Rate in % 75.7 59.0 84.1 Primary education 311839 102014 209826 Rate in % 57.7 40.1 73.5 Secondary education 207954 99850 108104 Rate in % 30.6 25.7 37.2 Higher or University education 29307 16639 12668 Rate in % 17.7 18.9 16.3 Source: SSO, LFS, 2006. In regards to this, Urdinola and Macias (2008) have found that in Macedonia women who have attained secondary and tertiary education show between 24 and 50% higher probability of participating in the labor force in comparison with those women who have primary or less education. 71 The same authors conclude that other distinctive factors that can be related to non-participation of women in the labor force are the number of children and the ethnic background. Namely: · Each child living at home younger than 7 years old reduces female participation by 3%; · Female individuals from ethnic Albanian origin are associated with 35% lower probability of participating in the labor force with respect to female individuals from ethnic Macedonian origin. According to the 2000 LFS, the participation rates of women of ethnic Albanian origin were 11%, compared to 51% for female ethnic Macedonians. Bearing in mind these findings, the analysis that follows is focused on exploring the specific factors low-skilled women from minority communities in Macedonia perceive as hindrance for their inclusion in the labor market. 72 METHODOLOGY In this part, the characteristics of the sample are explained, as well as a presentation of the data- collection process, methods and definitions used for the study. The method of focus groups was selected as the main source for data collection primarily because of its ability to provide in-depth qualitative data on the issues related to female non-participation in the labor market, different from the data provided by the typically used surveying methods. It enables for certain issues (questions) to be raised several times, which usually "unravels" the initially provided socially desirable responses. Additionally, the method allows for opinions to be expressed openly in a comfortable atmosphere, which resembles casual conversation. The process of ones' opinions being confirmed or confronted by others facilitates his/hers emotional reactions, which can not be detected by the traditional surveying methods. A significant limitation of the method is the inability to generalize the findings, since the sample is not a perfect representation of the population and each individual does not respond to every question. DATA COLLECTION PROCESS A total of 23 focus groups (16 with female participants and 7 with male participants) were conducted in 8 municipalities in the period between April 2 and April 23, 2008 (see schedule of data collection in Appendix 1), with the assistance of 8 local NGOs. The main group for analysis (or treatment group) consisted of unskilled women not participating in the labor force, while the control groups were: non-participant skilled and semi-skilled women, unemployed women, and non-participant and unemployed men. However, in the groups, women and men with different educational and labor market characteristics were mixed in order to facilitate contrasting opinions. In order to enable the participants to express themselves freely on gender-sensitive issues, the female focus groups were moderated by female facilitators, while male groups by male ones. The process involving the focus groups interviews had several steps. First, participants were selected by our partner NGOs, following the Centre for Research and Policy Making (CRPM) designed parameters. The key characteristics and their interrelation for the target sample included: nationality, age (i.e. working-age) and gender. They were also provided with the part of the LFS aimed at determining economic (in)activity and were asked to select the participants by using the LFS screening questions. Once gathered at the focus group, the participants were asked to complete a questionnaire regarding basic demographic characteristics (which required about 5 minutes). Afterwards, an open discussion (lasting between 50 to 80 minutes) facilitated by the moderator was conducted based on a questionnaire previously developed for this study based on World Bank guidelines (see Annex 3). All discussions were recorded with a tape-recorder and three were also recorded with a video-camera. Groups were moderated in two languages, Macedonian and Albanian. All groups with female ethnic Albanian participants (and part of the ethnic Turkish participants) were moderated in Albanian; while the groups with ethnic Macedonian, Roma, ethnic Serbian and part of the ethnic Turkish women in Macedonian. The male groups were moderated both in Macedonian and Albanian, depending on the language spoken by the focus-group participants. In addition to the focus-groups, 10 face-to-face interviews were conducted in the period between May 1st and May 4th, 2008; with inactive women who had a low level of education from 73 traditional ethnic Albanian families. This supplementary was made due to the fact that the local NGOs had difficulties convincing women with this background (and their husbands or the heads of the family) to participate in the focus-groups. Therefore, CRPM field surveyors visited 10 women with the abovementioned characteristics in their homes and conducted in-depth interviews, using the same questions as the ones from the focus-groups' questionnaire. STRUCTURE OF THE SAMPLE The sample consisted of 233 people, of which 68 (29%) were male and 165 (71%) were female. A more detailed presentation of the characteristics of the sample is provided in Figures A1-A3 (Additional data are available in Appendix 2). Figure A1: Distribution of Sample According to Figure A2: Distribution of Sample According to Gender Ethnicity 1% 2% 29% 18% 18% Macedonian Albanian male Turkish Roma female 14% Bosnian Serbian 71% 47% Source: own elaboration. Figure A3: Distribution of Sample According to Educational Attainment 8,20% 2,60% 14,60% No education 40,30% Primary Secondary High University 34,30% Source: own elaboration. As it can be noticed from the tables, the sample was purposefully selected to include more women than men and more women from minority communities, since the focus of the study are women from ethnic minorities. The numbers of representatives from minority communities were selected on the basis of their share in the total population of minorities and their distribution in each municipality. 74 DATA COLLECTION METHODS A set of five questions from the LFS were used for screening the participants to identify whether they were active or inactive. The aim was to provide about 30% unemployed (but active) participants and 70% unemployed (inactive) participants in order to enable comparisons. Afterwards, a set of questions (See Appendix 3) was developed to be discussed in the focus- groups. Two versions of the questionnaire were developed - a male and a female one. Although they basically explored the same issues, some of the questions were adjusted in order to extract more honest responses on the gender-sensitive topics. The male questionnaire had 40 questions, while the female - 45 questions. The main topics that the questionnaire was aiming to cover were the following: · Previous employment, i.e. history of unemployment · Existing skills, motivation for re-qualification, additional education, reservation wages · Reasons for being unemployed and economically inactive · Cultural perspective of the problem of their unemployment · Source of the family monthly income (e.g. transfers, remittances, pensions) · Decision-making regarding property, financial and job matters · Problems emerging from the economic inactivity · Access to infrastructure · Perception of job opportunities; motivation for employment · Feelings of discrimination because of unemployment Box A1: Definition of Terms Unemployed population: Persons that during the week of the report have not worked for money (in cash or in kind) or for profit but have made attempts to find employment during the previous 30 days. Population not in the labor force: Persons that: during the week of the report have not worked for money (in cash or in kind) or for profit and have not made any attempts to find employment during the previous 30 days Problems/Challenges Encountered Certain problems were encountered during the data-collection process that one needs to have in mind while considering the results. The LFS as a valid screening method proved to be quite problematic: Although all participants responded they did not have any personal income during the previous week, the discussion revealed that this is not true for some (especially male) participants. Many of them, although registered as unemployed, were found to work as manual laborers for a daily wage (although not necessarily during the previous week). There are various explanations why the participants provided answers to the LFS which indicated that they are inactive: - Some may be concerned that if they report their income (temporary employment), they might loose their status as unemployed and hence the health insurance that goes with it; - The same logic follows the idea that those that receive social allowance may be afraid of loosing it; 75 - Since the work they are performing is typically in the informal sector, they might be afraid they will be sanctioned for that. A certain number of the participants who reported they haven't looked for a job during the past month during the discussion stated they are actually looking for a job all the time. They "keep their eyes open" in case an opportunity comes up, even though they may not take any active measures (e.g. they have not sent an application, did not go to a company or an institution, did not look for advertisements in the newspaper etc.). However, their friends/relatives etc. are acquainted with the fact that they are interested in working and are expected to inform them if they hear of an opening. Hence, the validity of the LFS as a screening method for the economically inactive population may be put in question. In reality, significant numbers of the inactive persons (mainly male) actually appear to be quite active and some would even pass as employed. Difficulty in convincing women from traditional Muslim communities to participate: The problem of convincing the women from traditional Muslim communities (mostly ethnic Albanian) was quite dominant in all municipalities. First of all, these women have limited contacts with the "outside world" and a limited social network. Second, in order to get them to participate in the focus-group discussions, one needs to first consult with their husband or the head of the family and ask for approval. Very frequently, they were not given approval, which prevented them from participating. Since only few women of this type were present at the focus group discussions, the CRPM team has decided to approach 10 women from these communities directly in their homes and conduct individual face-to-face interviews. This has improved the representativeness of the sample to a certain extent. Willingness to share personal financial/equity information: Another problem encountered is some of the groups were related to the personal dynamics of the groups. Sometimes a single group included several acquaintances, friends and neighbors, also largely due to the fact that many of the places that we visited were smaller towns and a vast majority of people in the same peer group happen to know each other (sometimes rather well). This made obtaining sincere straightforward answers related to the financial income of a household, remittance from abroad and other factors which in effect reveal the individuals' financial "value" to the group problematic. This became very obvious for some groups during the informal debriefing section of talking during the joint lunch. Different information was provided by the same person and the participants spoke more openly about sensitive questions, for example about the amount of remittance they receive. When possible an effort was made to include such a debriefing section in order to check/support the validly of the data obtained during the formal focus group sessions. 76 RESULTS This section presents the main findings of the survey and the resulting conclusions. An attempt is made to present the opinions of women by highlighting their ethnic background and educational attainment, as the two main factors affecting non-participation. Additionally, the male opinions are presented in parallel as a means to complement the female ones and provide a complete understanding of the phenomenon. HISTORY OF (UN)EMPLOYMENT Although different patterns of (un)employment were noticed, the female participants can generally be divided into four groups according to their experience with employment and/or unemployment: · Women who have never worked for a variety of reasons (typically a combination of tradition and low education), some who do not want to work and others who want but do not believe they could find adequate employment. They typically are mostly represented in the traditionally Muslim communities (ethnic Albanian, ethnic Turkish and somewhat less frequently Roma). These women represent the main group of interest for this analysis. · Women that have worked before the transition (before 1990). During the privatization process, their companies were liquidated and they were laid of as technological surplus. Since they still have not fulfilled the requirements to be eligible to receive pensions, the majority of them still want to work. However, they see their age (40 or more) as the main constraint for employment. They are mostly represented among the ethnic Macedonian women. · Women with different levels of education who work from time to time, mostly as volunteers, seasonal workers, typically in the informal sector. Mostly ethnic Macedonian, ethnic Serb and Roma · Younger women (20-35) who have never worked and have rather high expectations; i.e. seek job according to their educational qualifications and have a relatively high reference wage limit. Mostly ethnic Turkish and ethnic Albanian, but also ethnic Macedonian. All women (except for those that were still in the education process) described themselves as currently being housewives, with the difference that most of the non-participating unskilled women stated they enjoyed this "profession", while others do it in the absence of a "real" job. The situation with unemployment and inactivity in men is rather different. In reality, very small percentage of the male participants was "really" inactive. Although in the beginning, they did not declare any income or that they are actively searching for a job, the discussions usually indicated that many actually do work in the informal sector (as manual laborers, seasonal workers etc.). This was valid for men from all ethnic backgrounds. 77 MOTIVATION FOR RE-QUALIFICATION AND SKILLS UPGRADING Existing Skills A vast array of skills was mentioned by the female participants regardless of their age and ethnicity. The majority of low-skilled women from Albanian and Turkish ethnicity typically emphasized their "homemaking" skills, which primarily include: maintaining the household (cooking, cleaning), child care and different types of craft-work (knitting, tailoring and embroidery). The skills in doing the everyday domestic activities were also mentioned by all married educated women. Agricultural skills were also often mentioned, mostly by women living in rural areas with different educational backgrounds. Many of the younger and more educated women from all ethnic backgrounds stressed they have completed different trainings (mostly related to IT skills and foreign languages). Furthermore, the women with secondary vocational school and university education stated skills related to their profession (e.g. teaching, nursing, economical and administrative skills etc.). Motivation for Skills Upgrading Regardless of the skills they already possess, ethnic Macedonian, Roma and ethnic Serb women with various educational attainment appeared to be very open to the idea of additional education/qualification if it comes with the prospects of being employed. However, several have mentioned that they would first need to be given a guarantee of employment after finishing such a course and some emphasized that they would accept the offer for additional training only if it is paid by the employer (or someone else), because they are not in a position to take on such a financial burden. Overall, there was an enthusiasm present when the idea of additional education was mentioned. One Roma woman with low levels of education from Bitola exclaimed: "I may only have completed fourth grade but I can do everything I put my mind into; even learn to work on a computer." The opinions of the ethnic Albanian women, on the other hand, were rather divided. While some of them (mostly the younger and more educated) would enthusiastically accept the offer of additional education, others would hesitate for different reasons. The most widely stated reasons of the later group are given bellow: · They do not expect to get approval for additional training by their husband. In the words of a woman from Kumanovo: "This is not Skopje, in Kumanovo the husband dominates and if you don't listen to him others think bad of you" In addition, a younger woman from Tetovo expressed her opinion that "My husband won't let me attend such night classes, because he doesn't want me to walk at the night hours on streets". · They do not believe they would have the time for additional training, since the domestic responsibilities take too much of their time. Women who lived in extended family and had young children felt especially restricted by their role as housewives. · They do not actually want to work, so there is no point in upgrading skills. Large numbers of women that live in traditional communities, mostly in the villages, have gotten 78 used to being housewives and they do not see the need and purpose of their involvement in the formal labor force. Since the need for additional education is associated with formal employment, it is perceived as an unnecessary burden. The level of education in many but not all cases was a variable that was related to their opinion of skills upgrading. For example, some low-skilled women that have already had the experience of upgrading their skills (as the women from Struga that attended classes in making national costumes organized by a women NGO) reported they would be happy to get out of their homes and feel a different atmosphere. They reported their experience as a form of socializing and at the same time gaining new skills, and as a getaway from the feeling of isolation which appears after staying at home all day long. Regardless of their opinions on the perceived need for additional education, women from all ethnicities did not appear to be very aware of the possibilities for further education and professional development. Most of them were aware that knowledge of languages and computer work is essential today, but mostly the younger ones had actually gotten involved in these types of education. Only a few participants with higher education completed or attempted to enroll into the trainings organized by the Employment Agency (IT skills and English language) and the IT training organized by the Macedonian Government. Even though these trainings are delivered free of charge, most participants did not attempt to take advantage of this opportunity to upgrade their skills. Several have mentioned that they would not even attempt to get enrolled since they are certain that these trainings are fixed and only the members of the political parties in power are actually invited to participate. In addition, only several ethnic Macedonian and ethnic Serb participants reported that they have continued their professional education, in order to increase their chances of finding employment. The situation with men was somewhat different. Most did not really see the value in getting additional training, as they did not have time for additional education. Regardless of their ethnic background, the majority felt that the primary need was to spend their time earning money. Often, employment is not perceived as a means of long term personal development. Rather work is more narrowly viewed as a means of generating some money to help out the family budget, and hence not related to continuous professional improvement. Although, the issue of low educational levels and lack of qualifications was perceived by many as one of the reasons for their unemployment, the enhancement of one's skills was rarely associated with the possibility to increase ones competitiveness on the labor market. Unsurprisingly, only a few seemed particularly motivated or saw the value of additional education or skill advancement. One participant from Sveti Nikole even jokingly mentioned "If I get money for it, sure". These opinions probably influence the statistically lower returns to education of men compared to women, as identified by Angel-Urdinola and Macias (2008), who conclude that returns to education among women are 10 to 15 percent larger than among men at all education levels. REASONS FOR NOT WORKING / NOT SEEKING EMPLOYMENT Reasons Stated by Both Genders The main discouraging factors that prevented men and women from all ages and ethnic backgrounds to seek employment are nepotism and political party affiliation. Almost all participants who reported they would like to work unanimously agreed that they feel frustrated and disappointed because of the "unofficial" employment policies in the country. Specifically, if 79 one wants to find employment in a state institution (still perceived as the most secure form of employment), s/he needs to be affiliated with the political party in power. The newspaper adds for the job openings are perceived as a farce, and everyone knows that the positions have been filled beforehand. This is especially noticeable in smaller towns where people know each other closely. Hence, they feel discouraged and do not apply for a these positions, perceiving that they have no real chance of getting it. "We know one another in the town. When you say to someone that you want to apply for some position, they say ­ you shouldn't bother and spend money on that, this or that person is already employed there." Serbian woman from Kumanovo These practices have deteriorated their trust in the institutions and deepened their discouragement regarding the possibilities for finding employment. "The absence of trust in the governmental institutions; also the low moral values in the society, we just don't try to even apply any more." Ethnic Albanian woman from Tetovo "That kind of behavior from the institutions is a big demoralization for us and this is one of the reasons for our unemployment." Ethnic Albanian woman from Debar In addition to the "perceived" unfair competition, in order to apply for a position in a state institution one needs to spend quite a lot of money (for photocopies, notarization of documents etc.) which makes the participants question whether they should spend this amount at all, since they are not employed and money is already tight. Also, a unanimous conclusion was reached regarding the possibilities for employment in the private companies. All participants agreed that you need to have strong connections to get employed there, mostly referring to being a relative or a close friend of the employer. The Roma participants in particular strongly expressed their aggravation, which suggests that they are especially harshly hit by the lack of developed social networks. Another important discouraging factor for seeking employment in the private sector is related to the unregulated labor relations, or more specifically ­ the employers' disrespect of the existing labor regulations. For example, it is a common practice of many private employers to hire someone to work for 1-2 months on a trial/probational basis and then fire him/her, paying a probation wage or no wage at all. Women and men from almost every town agreed that this is a common practice in the private companies (especially the textile ones). Since most personal experiences related to working for private enterprises were resoundingly negative (late payment of salary, long working hours, no health benefits or social insurance), many of the participants had lost faith in these companies acting with caution and reservation when seeking employment in this sector. All things considered, in spite of the openly expressed criticism in the state institutions' employment policies and the lack of trust in their "impartial" functioning, a state-sector job is still considered by many as the most stable type of employment and is associated with a secure future. Private sector employment is not seen as attractive (or often not seen as a real employment) due to the lack of proper legal regulation of the labor relations and the small/irregular pay. This sentiment is held very strongly; it seems that working for small private employers is seen as 80 undesirable and unsustainable in the long term. The strong frustration with the lack of security and benefits indicates that for many it is only acceptable to have a job which will provide this long term security. Everything else is a partial and temporary solution. Reasons Stated Exclusively by Women Despite the above stated employment constraints shared by both genders, women emphasized several other reasons specifically related to their gender and the traditional beliefs regarding female roles. In this regard, there were significant ethnic differences. Interestingly, almost none of the ethnic Macedonian women, regardless of their level of education, said they do not want to work. This is probably so because the tradition of employment is very strong in their communities and not wanting to work is perceived as laziness. For illustration, the local NGO that assembled the participants for the focus groups in Sveti Nikole emphasized that most of the inactive participants felt embarrassed to admit they are not seeking employment. Therefore, many ethnic Macedonian and Roma women identified as inactive by the LFS screening questions said during the discussion that they are constantly seeking employment. Although they may not take any active measures to find one (following job adverts, asking friends, visiting companies etc.), they constantly "keep their eyes open" in case some opportunity comes up. For example, the majority were registered as unemployed in the Employment Agency and waited to be called in case there is a job opening. Many still relied on the Employment Agency as the basic source of employment opportunities, even though they were aware that only a few those who registed with the employment office are actually called and offered a job. Out of the ones who did not seek employment at the moment, most chose this option because they had to take care of their young children, were still in the process of acquiring education, or believed they are too old to work (i.e. were over 50 years old). A general discouragement was noticed among women from all ethnic backgrounds, which appeared to be the main reason causing their current inactivity. The main discouraging factors (with different prevalence among the ethnic groups) were: Age The women that had a regular working relationship before the transition found themselves especially hit by their current state of unemployment. As one ethnic Macedonian woman from Struga mentioned: "We are the generation that has suffered most from the transition. At this age (she was 50) no one wants to employ you. As soon as they look at you and at your working card4, there is no chance...". These women were used to working and providing for their families; it made them feel useful and fulfilled. However, after the transition when many companies were liquidated, they have lost their jobs and now feel that there is no hope for them. They believe that their biggest constraint for finding employment is their age. Bearing in mind the fact that every employer wants young 4 Working card is a document that every person that registers in the Employment Agency receives. It contains basic data about the person and all formal employment engagements that s/he has had (when and where it was). 81 workers (usually the age limit is set at 30 or 35 years); the "older" women (according to the labor market criteria) often feel discouraged regarding the possibility of finding their place in the current labor market. The constant rejections they have faced throughout the years made many of them discard the possibility of finding employment beforehand, which results in not trying at all. One Serbian participant from Kumanovo put it succinctly: "I've lost any hope of finding employment." Many older low-skilled women from ethnic Albanian and ethnic Turkish communities that have never participated in the labor market also stated that they do not believe they could find employment because of their age. However, contrary to rest of the women, they also did not believe they possessed the appropriate skills needed in the labor market. Marital Status The older women were not the only discouraged workers since many (mainly private) employers are reluctant to hire young newly married women, or women with small children. These are some of the factors that indicate to the employers that these (potential) young mothers would not be able to work long working hours because of the need to spend time with their children. Many women exclaimed that when they go to an interview, the questions regarding their marital status appear to be the deciding factor for (not)getting the job. "The first questions they pose to you when you go to an interview are: Are you married / planning on getting married?; Are you planning to have children any time soon? etc. If you answer ­Yes, than you can say good-bye to the possibility of getting the job." Serbian woman from Kumanovo (mother of a small child) These circumstances make many young women who have decided to start a family, give up on the possibilities of finding employment in the private sector and often decide to stop seeking employment until their children grow older. Physical Appearance Several participants (mainly of Macedonian and Serbian ethnic origin) have experienced discrimination on the basis of their looks when seeking employment in the services sector (e.g. catering, sales). They reported that private employers from this sector tend to look for attractive women and advice them to flirt with the customers. This was a definite hindrance for these women to continue to seek employment in these sectors. Disbelief in the Possibility of Finding an Adequate (mostly state-sector) Job A general discouragement was noticed among many (mainly ethnic Albanian and younger ethnic Turkish) women that they could find a job with their qualifications (regardless if they had primary or a university education). "There are no jobs for us" concluded an ethnic Albanian woman from Struga. Although they believe that theoretically there are no constraints for them finding employment, since they are physically and mentally healthy, in reality most of them can not see this happening, since there are none or very few adequate job openings in their towns. This is understandable if one bears in mind that for these women the term "appropriate job" refers to an administrative position in a state institution, which is difficult to get because of the above stated reasons (lack of political affiliation, relevant personal connections), sometimes combined with the low level of education. 82 Occupied with Unpaid Care-Work at Home A relatively large part of the ethnic Albanian women and a smaller number of the ones from other ethnic communities explained that their responsibilities related to care-work for certain family members do not allow them to work. These responsibilities can be divided into: - Taking care of their small children. For illustration - three of the ten ethnic Albanian women who came to the focus group in Kumanovo brought their young children along, as they did not have anyone to take care of them while they were out of the house. - Taking care of ill family members (mostly elderly) at home and providing them with a special medical treatment Specifically, when living in an extended family, it is normal for an ethnic Albanian woman to stay at home and take care of the elderly family members or her small children, regardless of the level of education she has attained. This is seen as a woman's primary obligation, and although the higher educational attainment can somewhat reduce the family pressure; many highly educated women nevertheless decide their main task is care-work. In contrast, in the ethnic Macedonian and ethnic Serbian families, mothers (and the community in general) are more often "allowing" for the mothers to leave the children in a kindergarten or with the grandparents in case they decide they want to work. There were some key employment constraints which were shared by members of a particular ethnic background, and seemed to be strongly applicable for their ethnic group. For many Roma women there is a pressure to marry quite young and have children soon afterwards, which prevents them for obtaining further education and decreases their opportunities in the labor market. While many Roma women conform to these traditions, the ones that have attained secondary education or more strongly object them. For example, a young Roma woman from Skopje said to us that while she is encouraged to find employment by women her age, older women in her family usually advice the young girls to continue with their life as it was before and not take any risks, because the ultimate goal is to get married and establish a family. Several of the Turkish women with low education reported that they would like to work, but their low level of education (completed primary school) does not provide them with many prospects for employment. Some admitted that looking back they feel bad for not continuing their education when they were younger. However, they are actively trying to provide different options to their children, for example by enabling their daughters to obtain better education. Ethnic Albanian women also reported certain reasons for not working that were typical for their community. These setbacks for participating in the labor market were experienced by most of the women, although the ones with higher education appeared to be more willing to "rebel" against some. These factors are usually intertwined and influence the women's decisions regarding employment in a systemic manner. They are as follows: · Having too many domestic obligations that they do not believe they would be able to fulfill if they had been employed ­ an opinion held by women with different educational attainment. · Not being used to the idea of employment, since they have spent all their lives being housewives. Hence, they prefer to stay home and take care of the household and the children. This opinion was typically held by low-skilled women living both in urban and rural areas. "I don't like to work and I could not stand to go out and work for 8 hours for someone (...) as a matter of fact nor would my husband like such thing for me." 83 Ethnic Albanian woman from Skopje · There is no need for them to work since they consider their current monthly income to be sufficient for a normal life. This opinion was held by several low-skilled women living in urban areas. "There has always been enough money at home, since all of my family members work" Ethnic Albanian woman from Skopje · The traditional (conservative) values which either fully prohibit the woman in the family to work or allow them to work only in certain professions and certain sectors (mostly state-owed). The women with lower educational levels appear to be especially affected by these "traditional" beliefs, since they do not have the necessary qualifications to work in the "appropriate" jobs. To illustrate, many women said that it is considered as unacceptable/inappropriate in the ethnic Albanian community for a woman to work as a waitress, a cleaner, or a bus driver for example. In the words of one ethnic Albanian woman from Kumanovo: "Yeah, I would work a cleaning job, its not that I have some high education, so I can not expect something more. So cleaning...it is something I love to do and believe me it's not an easy job at all. But here it is not that popular for women to go and clean other peoples' houses cause firstly, they don't pay much and secondly, people will gossip about it, treating it like a humiliation to do such a work." Most women agreed that the main reasons for these beliefs lie in the mindset of their community members. "Even if the husband earns only 200 denars per day, the wife should stay at home, prepare the meal, clean, to be around for the kids.... This mentality has been ruling for years now and it in implanted in the peoples mind so much that we need a lot of work to take it out of our thoughts, of our way of living. (...)The husband thinks that the others around him, the society, will talk and say different kind of things, humiliating things, if his wife is working." Ethnic Albanian woman from Kumanovo These traditional values are supported by the whole community and are openly expressed by the interviewed men. Specifically, many men from the Muslim communities (ethnic Albanian, ethic Turkish, and Roma) objected to the idea of "empowering" women and stood in support of the belief that men and women should have different roles in the home and in the community. Hence, since a woman has too many responsibilities around the house, she would not have time to perform additional tasks. In many cases, men did not have specific explanation as to why the wife should stay at home, but simply considered it to be her traditional role. "The position of the wife is to stay home and take care of the children." Ethnic Albanian men from Skopje "It is unusual for me for my wife to wake up in the morning and go to work, while I'm still in the bed. I would never let that happen." Ethnic Albanian men from Tetovo 84 "If the husband works, there is no need for the wife to work too, especially if the husbands' salary is good and if they live in a village." Ethnic Albanian men from Bitola Apart from the domestic obligations, they considered the lack of adequate education of their wives (having only primary education) and the lack of appropriate job opportunities (i.e. fulltime- jobs with benefits) as the main barriers to the female participation in the work force. It has already been noted, that low educational qualifications were not perceived as the major setback by the men in finding employment (mostly of informal and flexible character). However for the women in their communities it was noted as the main barriers for female employment, since for their communities it is not appropriate for women to work informally and in the private sectors. Some would nevertheless allow their wives to work privately only if she worked with them (i.e. with her husband), for example running a joint business. Still, most of the men from Muslim communities would not object a highly educated woman to work, provided she does not have many contacts with men and has appropriate transportation. The dominant belief is that these women should be able to find suitable state-sector jobs. The ethnic Albanian women with lower education (primary school or less), especially those from rural areas accept the traditional male domination in the family and do not perceive it as something that needs to be changed. However, the more educated and especially younger women admit that the mentality of their families is still conservative and needs to be changed. These opinions differed only among women from Debar and partly from Gostivar, who stated that their husbands are "open-minded" and the tradition of the woman not working has been forgotten long time ago in their communities. Even though their opinions were more emancipated, they admitted that the conservative values are still dominant in many ethnic Albanian families. Luckily, many of the more educated ethnic Albanian women believed there should be shift in the way of thinking among the people in their community. A woman from Tetovo exclaimed her opinion that "except the change of the system we should try as a society to influence this tradition so that it could stop one day, to stop this tradition of not working. (...) We should fight against this tradition to show them that women can be creative and hardworking as men are." CULTURAL PERSPECTIVE ON THE ISSUE OF FEMALE EMPLOYMENT In order to explore the issue of opinions and values towards female employment from a different angle, we asked the participants about which member of their family makes the decisions regarding their employment. Women from different ethnic backgrounds answered in different ways; the main differences were between the ethnic Albanian and ethnic Macedonian women. Specifically, most ethnic Albanian women stated that these decisions are made by the husbands, the fathers or the elderly family members, i.e. the ones that (in the words of one woman) "have more experience with the outside world". This was confirmed not only by the less educated women, but also among younger and highly educated women. An ethnic Albanian woman from Bitola explained that even though her husband does not object the idea of her working, her mother in law told her that she would expel her from the house if it happened. Moreover, the vast majority reported that most of the employed people in their families are male and some even said that other women may gossip if they inform them that they are planning to start working. "As for talking they always talk, they will probably look at us with irony." Ethnic Albanian woman from Skopje 85 In regard to this, and somehow surprising, it is especially important to note that even the low- skilled women who do not want to work support the employment of other women and consider it a step forward for their independence. They perceive employment as being related to higher education (e.g. secondary or university) and see employed women as being more educated, independent and civilized. "If a woman has skills and education, its not right for her to waste them and not use them in any job or activity, and the other positive side is that if she works she becomes independent from others, she becomes the master of her own life ­in the perspective of financial state" Ethnic Albanian woman from Skopje They perceive their inactivity as primarily being caused by the conservative beliefs and the low education levels, and not the lack of skills. Hence, the low education is not seen as being related to not having specific skills. "The older women aren't employed because the mentality was like that within Albanians and not because they don't know how to work" Ethnic Albanian woman from Skopje Some of them even emphasized that in many ethnic Albanian families, even woman with high education and different professional skills may be discouraged to work by her or her husband's family. Participants from other nationalities, regardless of their educational attainment, strongly confirmed that they make the decisions regarding employment on their own or in collaboration with the other family members. However, several Roma and Turkish women said that the practice of men or elderly family members to decide about the employment of the female family members sill prevails among some Roma and Turkish families, although this tradition tends to be dying out. As one low-skilled Turkish woman from Bitola said: "It used to be that young people asked for the permission of the elders for everything, but now, no one asks the elders for anything." Many men from ethnic Albanian and Roma communities expressed a rather strong opinion that they (or other elderly men from the family) decide who is going to work. In these two ethnic groups, in particular ethnic Albanians, the family culture is a definite deterrent for the wives and other women in the family to find employment and work. There was a very strong declarative tone in the idea that they (meaning the men) manage the family budget and determine if his wife can work. In short, they are the head of the household in a traditional sense. In those cases when men would consider "allowing" their wife/daughter/sister to work, they would first consider the type of work that needs to be performed, the working hours (typically not more then 5-6 hours a day), the employer (especially in the case of a private company) etc. Employment vs. Housework Many women that would like to work may find themselves constrained by the domestic obligations they are traditionally expected to perform. All participants, regardless of their education and ethnic backgrounds, confirmed that the lion's share of the domestic tasks (i.e. cleaning, washing, cooking, and care-work) are typically left to be performed by them (or if younger - their mothers). Hence, they were asked to assess whether they believe that an employed woman can also be a good housewife. 86 Almost all ethnic Macedonian, ethnic Serb, Roma and ethnic Turkish women stated that being employed does not prevent a woman from also being a good housewife at the same time. Moreover, many stated that a woman who is employed feels self-fulfilled, more confident and motivated, which makes her perform the domestic duties with more enthusiasm. In contrast, a woman who is not formally employed does not feel useful and lacks strength for performing the house obligations. However, some emphasized that the working hours can play a role here, meaning that a woman that works 12 hours a day (as is the case in many private companies/factories) and comes home exhausted would not be left with much time and energy for housework. Less that half of the ethnic Albanian women, typically with higher education, expressed a very similar outlook on this issue, perceiving employment as something which motivates the woman and gives her strength to be even better in her housework. In contrast, the rest of the ethnic Albanian women (with different educational levels) were of an opinion that an employed woman could never be very successful in the housework. As one of them said: "No, I don't think that a woman working a full time job can manage to do all other tasks at home too. Only if she does extra work at night. And I especially see the work undone if the woman has children." The opinions of men on this issue were quite divided. In general, many of the Roma and ethnic Albanian men expressed a rather strong approval of the traditional gender relations. There was a clear understanding that the conventional gender roles are reflected in work and financial matters. The women primarily take care of the household and sometimes may work (for Roma), while the men bring home the main income and manage the family budget. "If I am busy, a woman can't go out to the market and also do my laundry." Roma men from Kumanovo In contrast, the vast majority of ethnic Macedonian men did not see the female employment as contradictory to the women's domestic tasks. Moreover, they claim that in these times of financial insecurity there is a need for both partners to work. The men from Sv. Nikole even said that in their town, women are more represented on the labor market than men, which sometimes leaves the men to take over large chunks of the domestic tasks. MANAGEMENT OF INCOME AND PRODUCTIVE ASSETS Sources of Household Income Most participants shared that money was tight and that the family income was barely sufficient or insufficient to cover basic family expenditures and needs. This opinion was shared regardless of the ethnic and educational background of participants. Nevertheless, a vast number of sources of income were mentioned by both male and female participants. Large numbers (including almost all Roma participants) reported to receive social assistance from the state. However, this small amount (2.000 MKD) did not satisfy any of the households' financial needs. Additional sources that were mentioned were: · Wage (regular or irregular) of the marital partner. More than a half of the female participants reported to have a husband or a father (for the unmarried) who is employed and provides the largest part of the household income. Most of the low- skilled women that reported they do not want to work, said their husband earns enough money to cover the family needs. 87 · Pensions and/or wages from parents or in-laws, a source typically reported by the ethnic Macedonian participants with different educational levels. · Remittances from relatives living abroad (significantly more in the ethnic Albanian and Roma families), which are not necessarily money, but can be in the form of presents (clothes, cosmetics, jewelry etc.). However, since for the majority of families they are not provided on a regular basis, remittances are not considered as a significant and a reliable source of income. For the ethnic Macedonian families in particular, remittances were a relatively insignificant and rarely mentioned financial source. · Seasonal employment of various kinds (mostly in the summer period) and informal employment from time to time (e.g. catering, plumbing, painting houses, cleaning houses etc. depending on individual skills). In the ethnic Macedonian and Roma families, they are preformed by both men and women, while in the ethnic Albanian and ethnic Turkish families ­ strictly by men. Nevertheless, some ethnic Albanian and ethnic Turkish women with low education reported to sometimes sell the products they make at home (embroidery, knitting, tailored clothes), and provide additional income for themselves and the household. Overall, various sources of household income were mentioned and most households try to take advantage of as many resources as they can. Living in larger households also reduces the financial pressure and eases the current situation of earning none or scant irregular income. It can be concluded that in some cases the ethnic Albanian participants had a much more secure feeling due to their financial buffer (large family pooling resources and ample help from relatives abroad), which may cause their attitudes towards employment and types of employment they were willing to take to be much more relaxed. Another interesting fact that was encountered concerns the finding that certain number of low educated women from traditional communities, that are considered as inactive by all criteria, actually do make an income from time to time working from their home. Hence, complete inactivity (during a longer period of time) appears to be very rare even among the low educated women, as everyone tries to increase the household budget by utilizing his/her own skills. Possession of Property Only three (an ethnic Serb, an ethnic Macedonian, and an ethnic Albanian) women with secondary and university education stated that they have some property (house or an apartment) registered on their name. All others confirmed that their family's property is registered on the name of their husband, father or father-in-law (in some cases mother-in-law). Also, many of the men who were interviewed reported that their property is also not registered on their name, but on the name of their parents (especially when they were still living with one/both of their parents). This is primarily due to the difficult financial situation that these families exist in. It is sometimes impossible for the second generation (even after they have started a family of their own) to move out in rented premises or to buy something on their own. In the current situation the reality of low income and low credit viability is clearly compounded and made more restrictive by traditional social norms, especially among the Muslim communities. As an ethnic Albanian woman from Debar shared regarding inheritance norms in a typical ethnic Albanian family: 88 "The house and everything in it belongs to our husband, and if not on his name than on the name of the father in law, and if he is dead than under the name of the mother in law. It's a rule in our society as a respect for the elder member in the family." Despite the fact that they do not legally possess any property, most of the interviewed women, regardless of their educational attainment, did not appear to be prepared for the possibility of not living with the breadwinners (e.g. in case of a death or a divorce). They admitted that they have not considered the option of how they are going to manage financially if something like that happens. Some of the low-skilled women reported they would most probably seek help and financial assistance from other family members (e.g. one ethnic Albanian woman whose husband has died now relies on the financial assistance from his brother). Others nonetheless said they would try to find employment, no matter what kind, in order to survive financially and provide a normal life for their children. The later opinion was given by women regardless of their ethnicity. "We would have to find a job absolutely, if it is needed to look at every angle and at every door, no matter if it would be a cleaning job." Ethnic Albanian woman from Skopje Management of Finances Most participants said that because of the tough economical situation, they can not manage to save any money, especially not for their personal needs. However, the ones that do manage usually keep it in case of "difficult times". The practices of management of finances differed from one family to another. As expected the dominant culture in traditional Muslim households is for the elders firstly and then secondly for the oldest male to be responsible for managing finances. This allowed for the mother to have the last word in this matter when she was the eldest family member in the household. "We live in a community where the oldest one is the most honored, for example the son gives the money to the mother, and they again are there to be used for all the family, but for the respect they always give that to the elder one in the family." Ethnic Albanian woman from Debar While most of the women from Muslim households accept and respect this tradition, some of the more educated were angered by it. A Turkish girl from Skopje expressed her aggravation with the fact that her mother needs to bag money from her husband despite the fact that she works as well at home (i.e. in the unpaid sector) but the results remain to be unseen by her husband. Also, despite the fact that most ethnic Macedonian women, regardless of the educational attainment, strongly asserted that the management of family funds is made in collaboration among all family members, some nevertheless admitted that although they may be consulted on how the money should be spent - their husbands or fathers (i.e. the breadwinners) carry the final decision. One highly educated Serbian participant from Tetovo was especially aggravated by this method of handling the family funds and stated that she feels discriminated because although she does so many tasks at home, her husband does not count her "voice" when deciding on financial matters. 89 ACCESS TO INFRASTRUCTURE Almost all participants that lived in or near the smaller towns stated that they have all the infrastructural necessities near to their place of living. Even some of those who did not live close to a factory reported that there are regular bus lines and the transport does not represent a constraint for their employment. The infrastructure was mostly seen as a problem by women that lived in larger cities (as Skopje and Bitola) or the nearby villages. Specifically, many of the women from Bitola (all of which lived in one suburb) explained that they do not have a school and a kindergarten near by, as well as a bus station, which carries with itself many infrastructural constraints. Even though the infrastructural approach was not mentioned by any female participant as being the main constraint for employment; several ethnic Albanian men reported that one of the reasons why their wives do not work is the absence of adequate transportation (they should not ride in buses as men) which is not connected to infrastructure per se but more to social norms. This reason, combined with the low education of their wives was seen as a definite hindrance for participation in the labor market. MOTIVATION FOR EMPLOYMENT Surprisingly, except for the highly educated participants, there was not much awareness among women as to their preferred line of work. Employment was seen by most simply as a source of income and better living, and much less as a source of personal and professional development. Nevertheless, most had a clear idea of the reference wage for various types of work. Very few women (ethnic Macedonian, ethnic Serbian, Roma and ethnic Albanian) with different levels of education expressed entrepreneurship initiative, stating that they would like start a business of their own (a store, boutique, hair-dressing saloon etc.). "I would like to be a business women and have my own Restaurant or Boutique...This is the only business witch could be supported by my family." Ethnic Albanian woman from Skopje However, these women were not very optimistic that their "dreams" could come true, mostly because of the need to have a large amount of capital for staring a business. They were also skeptical of the possibilities for supporting financing of self-employment offered by the State Employment Agency. It can be tentatively concluded that low-skilled ethnic Albanian women are predominantly of the opinion that they do not want to work at all, but would rather be housewives and care-givers. Furthermore, those that do want to work usually tend to be quite selective as to what job they would accept and the minimum salary they expect. The ones who have lower education and come from traditional families have rather specific criteria as to what type of job they would accept (presented bellow): "I would work a job which wouldn't have to do much with a male contact but where only women work." 90 "I would like to work if it was something that could be done from home, so I could have my own money and kind of independency from the others, but I wouldn't work a job that needs for me to go out 8 hours per day." Ethnic Albanian women from Skopje As already mentioned, state-sector employment is seen as more appropriate for the ethnic Albanian and many ethnic Turkish women with higher education, which makes them more selective regarding the type of job they would accept, when compared to ethnic Macedonian, ethnic Serbian and Roma women. They also appear to be more selective regarding the possibility of an employment without a health and social insurance. Specifically, while almost all ethnic Macedonian and Roma women said they would accept to work without being officially registered (since they are aware that this is a very common practice within private companies), the majority of ethnic Albanian and some ethnic Turkish women would not accept this. It is interesting to mention that the reference wage for most ethnic Turkish and ethnic Albanian women appears to be higher than the current average wages in the labor market. This is true for the ones with high and low education. Hence, most of them stated they would not accept to work for a wage lower than 150-200 EUR (9.000-12.000 MKD), even though they were aware that the typical salaries in their towns rarely exceed the amount of 9.000 MKD. Roma women in general appear to be the least selective when it comes to working any type of job. Although this mostly refers to the older and low educated Roma women, it appears that this way of thinking has transferred to many of the younger and highly educated Roma women as well. However, there were certain criteria that the ethnic Macedonian and Roma participants would not go bellow. For example, almost no one would accept to work 10-12 hours a day in a textile factory for a monthly salary of 5.000 MKD, since it would not cover their basic living costs. The opinions regarding the reference wage of other participants differed from case to case. However, they appear to be mostly related to their economic situation and less on the level of education. Hence, as expected, the ones who have stable incomes from other sources tend to have higher expectations, i.e. would not accept a wage lower than 200-250 EUR. On the other hand, the ones who are in a more difficult financial situation and would not mind even a wage of 100 EUR (6.000 MKD). Even though the former usually included women with higher education, some low-skilled ethnic Albanian and ethnic Turkish women also had rather high wage thresholds. FEELINGS OF DISCRIMINATION Most of the discrimination factors were mentioned above in the description of constrains (i.e. discouraging factors) for participation in the work force. Only Roma participants with low education felt they were being discriminated against on the basis of their nationality, i.e. rejected immediately after an employer sees they are Roma. This is most probably due to the strong stereotype held by non-Roma individuals that Roma are lazy and unreliable workers. In contrast to this belief, many Roma participants expressed their will to work and were the least selective regarding the tasks they are prepared to perform. "I want to work. Give me work. I want to provide for my children..." Roma woman from Sv. Nikole 91 Overall, the vast majority of women did not report feeling discriminated because they are not employed, since it is a quite common situation (especially among women) in the country and they do not feel as being different from others. However, several ethnic Macedonian women with university education reported to feel frustrated by the unfair treatment on behalf of their husbands because they do not contribute to the household income. Hence, even though they contribute to the household in many different ways, there is a pressure exerted that since they do not bring money in the house, their voice "weighs" less. In addition, some younger and highly educated women felt they are being pitied from their friends and/or relatives because they are still unemployed. An interesting paradox can be noticed here ­ while in the ethnic Albanian community many women are being pressured not to work, in the ethnic Macedonian community, women are being pressured to work and contribute to the household income. 92 RECOMMENDATIONS The problem of non-participation in low-skilled women is rather complex and includes a vide array of affected parties and stakeholders. In this part, the major employment constraints will be summarized and several possible solutions will be presented, selected on the basis of their feasibility within the context of the current political and economic situation in the country in mind. The recommendations primarily refer to increasing the educational opportunities of women and facilitating the implementation of certain existing laws. At the end, certain problems resulting from the use of the LFS are discussed and recommendations on how they could be overcome are presented. RECOMMENDATIONS FOR INCREASING PARTICIPATION IN LOW-SKILLED MINORITY WOMEN As expected, the traditional role of a dominant patriarch is still prevalent within many Muslim communities. While this influence is strongest in the ethnic Albanian community, it is somewhat less represented in the Roma and ethnic Turkish communities. These values are held strongly by some men and most of the women do not appear to vocally object to these norms and accept them as a composite part of their culture. Apart from the fact that the traditional values act through preventing women to work or allowing them to work only certain professions, they also instigate an array of other constraints. For example, the family restricts women to continue their education after primary school because the community does not expect them to need additional education for being housewives. In turn, women who have completed only primary education (or less) usually do not want to or can not take part in the work force, as many professions are off limits to them and the available ones are not very attractive. Also, many appear to be overburdened with domestic responsibilities and since they still perceive the task of being a good housewife as very important, they do not even consider the possibility of pursuing formal employment. As it can be seen, the factors are interrelated in a vicious circle which is difficult to be break. Obtaining further education appears to be the most "controllable" factor (i.e. easier to influence). In this regard, the new governmental policy of making secondary education compulsory (to be implemented from September 2008) may improve the educational levels of the new generations and increase their possibilities for participation in the labor force. However, the main challenge is to encourage girls/women to actually complete secondary school and empower them to seek employment. Also, there should be special programs which would provide adult education (primary and secondary) ­ ideally free of charge or with symbolic contributions - to individuals who have not completed their formal schooling. This would also need to be a carefully designed program which would be proactive (the institutions seeking out and motivating the participants) and with specific targets for yearly quotas of new graduates. Implementing these activities in the male-dominated cultures will be difficult. The family patriarch has a key influence in many areas (who in the family is going to be educated further, who is going to work and which professions are "suitable" for women), and active involvement 93 of the patriarch is necessary in order for such programs to yield results. However, considering the fact that there is a will among many women in these communities (especially younger ones) for further education and advancing their skills, they could act as mediators for reaching out to the low educated women from traditional families. This would be a challenge since their social circles are rather limited. However, female NGOs appear to have the capacities to get in touch with some of these women and work with them on changing the traditional outlook on female education and employment (both in the individual and the family circle). More importantly, women appear to trust these organizations and rely on them when it comes to delivering trainings aimed at upgrading their skills. These skills can afterwards be utilized to generate income while not having to leave their homes (for e.g. by selling the hand-made products they make at home). Moreover, a systemic solution for this problem could be achieved through making connections between the women with specific skills (which are not officially recognized) and the businesses in need of these skills. Here, certain NGOs (in cooperation with the State Employment Agency) can act as a "link" between the both parties by identifying the women and the skills they possess and communicating and matching them with the businesses' needs. Besides this, a part of the solution might be to train these women to start up their personal home-based businesses, as well as expand the access to credit. It is important to mention that low-skilled women from Muslim communities (and their families) do not actually object working per se, since they work in their homes all the time, but feel less comfortable working formally outside of their house and their community. Hence, more flexible ways of utilizing and upgrading their skills, as the one mentioned above, should be explored. For example, some women can lead a day-care center in their homes, since the care of children is one of the skills most report to posses and like doing. Not only would this provide employment for them but at the same time enable the other mothers from the community to seek employment, knowing that their children are in safe hands. Low education levels constitute one of the biggest problems for the appropriate inclusion of the Roma community in the labor market. However, the reasons for this are not only traditional, but very often due to financial constraints and the tendency to involve the children in the (informal) labor market at a young age. Despite this, the primary education of Roma has received the least financial assistance in the actions taken in the frames of the Roma Decade for improving the educational attendance of Roma children. Hence, providing financial assistance to Roma families for completing primary education (of children and adults) should be the initial policy concern, since the drop-out rates from primary school are extremely high among individuals from this minority group. This would not be a stand alone effort but would be accompanied by an awareness raising campaign specifically targeting the parents to motivate and support their children in obtaining primary and secondary education. In this regard, relating the social assistance to families with the enrollment and continuation of the formal education of children may be a good incentive for parents. The combination of promoting the aims of the program in the population and also providing the real time facilitating means and incentives to act upon realizing these goals is a necessary effort. While punitive measures (for example fines for not sending ones children to elementary or secondary schools) are important and should sustain, they are not enough and need to be complemented with other motivating sustainable measures. Finally, organizing buses which will transport only women and their children might prove to be an efficient mean of motivating some of them to get out of their immediate communities more in order to continue their education or seek employment. 94 GENERAL RECOMMENDATIONS FOR INCREASING PARTICIPATION The following recommendations could be applied to tackle the inactivity and unemployment within the non-participating women with higher education, as well as the long-term unemployed and discouraged individuals. Some are related to women from specific ethnic groups, while other can be applied regardless of the gender and ethnic origin. First of all, the fact that some young women from the minority Muslim communities continue to secondary school and university is encouraging, since the educational attainment would very probably increase their participation in the labor force. Also, the policy of "Equitable and Appropriate Representation of Communities in the Public Administration", which has been in implementation for the past several years is likely to partially solve the problem of non- participation (occurring as a result of discouragement) among some highly educated minority women. Moreover, the state-sector positions are expected to be suitable for the women that consider these types of jobs to be appropriate for them. However, minorities that do not have sufficient numbers of highly educated cadres (such as Roma) are again expected to encounter further problems in this regard. Despite the belief that the education would increase their employment prospects, the expectations from the labor market of many young and highly educated ethnic Albanian and ethnic Turkish women tend to be quite rigid. Since typically only state-sector employment is seen as an appropriate, it makes them quite selective regarding the type of job they would accept and the acceptable wage minimums (over 200-250 EUR). However, some are aware that these beliefs are old-fashioned and need to be changed, which speaks of their readiness to change the outlook on the jobs that are considered to be adequate for women. They may serve as catalysts for other women to "break" the traditional mentality and become aware that the fast-transforming labor market is in a need of flexible labor force that can adjust to the changing environment. For individuals that have been unemployed for more than 10 years, there is a risk that the skills they have gained during the schooling process or pervious employment are probably outdated and most importantly forgotten because they haven't been practiced for quite some time. The Employment Agency is currently organizing trainings in IT skills, English language and specific professions, but they are usually out of reach to the most "vulnerable" groups, such as the bulk of the non-participating women. They are rarely acquainted with the possibilities being offered and since they typically take place outside of their "immediate community", the possibilities for attending such trainings are very limited. Perhaps, in collaboration with the Worker's Universities around the country (or another similar institutions) the Employment Agency would be able to provide specific re-qualification and/or additional qualification programs for unemployed (including the inactive) by actively approaching the people and inviting them to select training according to their education, skills and preferences. Efforts should also be made to attract more mail attendants to the professional trainings by emphasizing that it should lead to better employment prospects. In addition, for individuals who have specific skills related to crafts (e.g. knitting, embroidery etc.) but do not have specific educational/professional qualifications, there should be system of accreditation developed to provide some kind of formal document. Also, training on developing job searching skills would be very useful for all. Many appear to be unable to "sell" themselves on the current labor market, despite the possession of sufficient professional skills. As a result of the constant rejections by employers, they feel victimized by the 95 institutional policies. Hence, gaining certain self-presentation skills may prove to be effective for an increase in their self-confidence and actually increase the possibility of finding a job. Having in mind that the discouragement usually comes as a result of the unequal treatment of individuals regarding different criteria (political affiliation, age, marital status, nationality etc.), the employers' respect of the anti-discrimination articles of the Labor Law is put in question. The same can be said for the articles regulating the protection of worker's rights (e.g. regular monthly wage, health and social benefits, 8 working hours daily etc.). Principally, the labor inspections need to be strengthened and become more effective in order for the people to regain their trust in the rightful functioning of the institutions. The existing forms of protections, such as the ombudsman office need to be strengthened as they are not being efficiently utilized by the citizens. Also additional independent monitoring mechanisms (independent from the courts and in cooperation with NGOs) could provide another avenue of action for citizens facing discrimination. Establishing cheaper procedures for applying in state institutions would also encourage many unemployed (some of which are inactive) to apply. Specifically, original and notarized documents could be requested only by the short-listed applicants, and not by everyone applying. Finally, more flexible ways of employment accompanied with minimal taxes for employees working less than 4 hours a day (i.e. less than part-time employment), should be reviewed. This would hopefully legalize certain informal sector jobs (e.g. manual workers, cleaners etc.) and at the same time enhance the respecting of certain workers' labor rights since the high tax wedge for low wage employments has been considered by experts (e.g. Rutkowski and Walewski (2007)) to be a hindrance for the formal employment of low-skilled workers. RECOMMENDATIONS FOR ADJUSTING THE LFS As elaborated in the part about the problems/challenges encountered, the LFS as a valid screening method proved to be quite problematic. Firstly, although during the screening phase many of the male participants provided answers to the LFS which indicated that they are inactive, were actually found to work as manual laborers for a daily wage (although not necessarily during the previous week). Secondly, a certain number of the participants who reported they haven't looked for a job during the past month during the discussion stated they are actually looking for a job all the time. Specifically, they "keep their eyes open" in case an opportunity comes up, even though they may not take any active job-search method measure. Hence, the validity of the LFS as a screening method for the economically inactive population may be put in question. In reality, significant numbers of the inactive persons (mainly male) actually appear to be quite active and some would even pass as employed. Probably, the definition of "inactivity" provided in the LFS needs to be adjusted. First, an extension of the reference period during which a person has not conducted any income- generating activity might be considered. The current period of one week appears to be rather short. Secondly, a rephrasing of the question "Have you been seeking employment during the previous 4 weeks" needs to be considered because of the subjective interpretation of the terms "seeking" and "employment". The first term is usually understood as active job search and the second is typically identified with a regular full-time employment. Moreover, the irregular engagements utilized for income generation are usually not considered as "employment". Adjustments in this direction are expected to improve the validity of the LFS to a certain extent, although it is important to emphasize that it is normal for the data provided by this method to differ from those gathered by the qualitative methods. 96 1.3.1 APPENDIX 1: DATA-COLLECTION PROCESS Table A1.1. Schedule of the data-collection process Town Date No. of focus Partner organization /municipality groups (No. of participants) Sveti Nikole 2.04.2008 2 (19) Women's Organization of Sveti Nikole Debar 9.04.2008 2 (19) Mladinski Gragjanski Komitet Tetovo 3.04.2008 and 3 (30) ADA ­ Action Towards the Future 5.04.2008 Kumanovo 7.04.2008 and 3 (29) Roma Rights Forum - ARKA 9.04.2008 Skopje 11.04.2008 and 3 (29) Center for democracy and solidarity 12.04.2008 Gostivar 18.04.2008 4 (40) ADI - Association for Democratic and 22.04.2008 Initiatives Struga 19.04.2008 3 (29) Women's Organization - Aureola Bitola 23.04.2008 3 (28) Bairska Svetlina Skopje 1.05 ­ 4.05 2008 (10) (individual interviews) 97 APPENDIX 2: STRUCTURE OF THE SAMPLE Table A2.1: Distribution of sample according to gender and age Gender Age male female Total 18-25 27 72 99 % of Total 39,8% 43,7% 42,5% 26-35 17 42 59 % of Total 25% 25,4% 25,3% 36-45 15 26 41 % of Total 22,1% 15,8% 17,6% 46+ 9 25 34 % of Total 13,1% 15,1% 14,6% Total 68 165 233 % of Total 100% 100% 100% Table A2.2: Distribution of sample according to gender and ethnicity Gender Ethnic background male female Macedonian 15 25 % of Total 22,1% 15,1% Albanian 22 88 % of Total 32,1% 53,3% Turkish 14 18 % of Total 21,0% 11,0% Roma 15 28 % of Total 22,0% 16,9% Bosnian 1 1 % of Total 1,4% 0,6% Serbian 1 5 % of Total 1,4% 3,1% Total 68 165 % of Total 100% 100% Table A2.3: Distribution of sample according to level of education Level of education Frequency Percent No education 6 2.6% Primary 94 40.3% Secondary 80 34.3% High (2 years after 34 14.6% secondary) University 19 8.2% Total 233 100.0 99 Table A2. 4: Distribution of sample according to place of living Habitat Frequency Percent town5 190 81.5% village 43 18.5% Total 233 100.0 Table A2.5: Distribution of sample according to marital status Marital status Frequency Percent Married 138 59.2% Not married 93 39.9% Divorced/widow/er 2 0.9% Total 233 100.0 Table A2.6: Distribution of sample according to number of children Number of children Frequency Percent One child 31 13.3% Two children 65 27.9% Three children 32 13.7% Four children 9 3.9% Five children 2 0.9% Six children 2 0.9% No children 92 39.5% Total 233 100.0 Table A2.7: Sample according to registered unemployment status Registered as unemployed Gender in the Employment Agency male female Yes 59 93 % of Total 86,7% 56,4% No 9 72 % of Total 13,3% 43,6% Total 68 165 % of Total 100,0% 100,0% 5 Note that some of the participants living in rural habitats close to the town stated they live in the town 100 APPENDIX 3: QUESTIONS FOR DISCUSSION IN THE FOCUS GROUPS PART 1: Questions for Identifying the Participants These questions are posed to all potential focus groups participants to determine their economic (in)activity Did you do any work for payment or some other kind of income during the previous week? Yes (Stop because the person is employed) No (Go to question 2) Even if you did not work, did you have a job from which you were temporarily not working during the last week because of annual leave, maternity leave, illness, work dispute, paid absence? Yes (Stop because the person is employed) No (Go to question 3) If the person is not currently employed: Have you ever work for pay or any kind of income? (including unpaid family workers) Yes (Go to question 4) No (Go to question 4) Do you want to work? Yes (Go to question 5) No (Stop because the person is inactive) Have you been seeking employment during the previous 4 weeks? Yes (Go to question 6) No (Stop because the person is inactive) What have you been doing in order to find an employment during the previous 4 weeks? (three answers are possible) ....Registering in the employment office ....Tried to start an activity as self-employed ....Contacted private employment agencies ....Placed advertisement ....Answered advertisement ....Contacted employer directly ....Checked with relatives or friends ....Participating in testings, interviews, etc. ....Other (specify) ________________________________ ....You haven't done anything (Stop because the person is inactive) Could you start with your job in the next two weeks if one were offered to you? Yes No (Stop because the person is inactive) 101 PART 2 Each participant separately is asked the following questions: What is your age? _______years Ethnicity: ____________________ Highest educational level attained:_____________________________ Are you religious? Yes No Are you married? Yes No Is your marriage registered? Yes No Do you have children? Yes No If yes, how many? ___________ Specify age and sex for each child ______________________________ ______________________________ ______________________________ ______________________________ Are you currently registered as unemployed in Employment Agency? Yes No (The following questions are posed only to the unregistered) Why aren't you registered as unemployed? _______________________________________________________________ _______________________________________________________________ Have you ever been registered as unemployed? Yes No When was that? __________________________________ PART 3 Introduction by the moderator and presentation of participants Focus-group discussion on the following topics: Previous employment, i.e. history of unemployment 102 Have you been employed before? If yes, what happened? How much time has passed since your last employment? What have you been doing during that time? Skills they possess, motivation for re-qualification, additional education, reservation wages What skills do you have / What can you do well (e.g. cooking, sewing, knitting, cleaning, crop- growing, child care, computer, clerical skills, occupational skills...)? Do you think you could manage to find a job with your qualifications/skills? Do you know about the opportunities for upgrading your skills? If you have the possibility of getting a job, but you need to upgrade your skills, what would you do? / Do you consider upgrading of your skills, additional education, evening classes etc.? Would you have time for that? Based on your experience, what do you think is the most frequent monthly wage offered to a person with your qualifications? Reasons for being unemployed and economically inactive (for the inactive) What are the reasons you currently do not seek employment? (e.g. Don't want to, don't need to ­ have enough money, don't have time because of housework, looking after children, taking care of an ill family member; I'm not allowed to work; I don't have sufficient qualifications; transportation difficulties) (for the unemployed) In what ways have you looked for a job? (e.g. searched adds; registered in the Employment Agency; asked friends, relatives; placed add; contacted employers...) For how many positions have you applied to during the last month? What kind of positions were they? What was the reason for not getting the job/s? Were you given any explanations or what is your impression for the reasons, e.g. you lacked professional skills, everything depends on the political affiliation or on the connections on has etc.? Were you offered a job, but you refused because of certain reasons? What were the reasons? Cultural perspective of the problem of their unemployment. Traditional values related to employment, their opinions about them Is it typical for your community for women to stay at home and not look for employment? (for women) Who in your family decides whether you are going to work or not? (for men) Who in your family decides which members are going to work? (for women) What would other women from your family say if you tell them you are planning to find employment? 103 Is there a female member of your family that works? If not, why not? What do you think of women who work? Do you feel that employed women have time for the everyday housework and child care? (19 and 20 only for women) If you were working, who would take care of the children and of household responsibilities? Would this person charge any money for the services? If yes, how much (monthly)? Source of the family monthly income (e.g. transfers, remittances, pensions) Who brings the money home? Are other household members working (formally, informally)? What do they do? Can you buy the necessities (food, drinks, clothes) from your household's income? Do you have relatives living abroad that send you money? (for women) How do you afford the things important for one woman (cosmetics, jewelry, clothes, materials for your hobbies- knitting, embroidering etc.)? Decision-making on productive assets (i.e. land, car, cash). What assets does your household possess? Are some of them registered on your name? Can you decide on how the money should be spent / how the land can be used (e.g. what seeds to plant) / renting the house/rooms etc.? Do you have your own savings? Do you manage the family savings? Who buys the food in your household? Who pays the bills? Who decides what to buy/ which bill to pay? Problems emerging from the economic inactivity (administrative, economic, health care etc.) (this question is only posed to women): How would you manage financially if you did not live with your husband / father / the household breadwinner? (e.g. he dies, you get divorced)? Have you ever had the need to raise a bank loan? If yes, did you encounter any problems? Access to Infrastructure Do you live walking distance (5 to 15 minutes away) from a nursery/child care center? Do you live walking distance (5 to 15 minutes away) from a bus stop? Do you live walking distance (5 to 15 minutes away) from the market? 104 Do you live walking distance (5 to 15 minutes away) from a factory? Perception of job opportunities; motivation for employment Do you want to work? If yes, if you had the opportunity what would you like to work as? Do you think that if you started looking for employment now, you would be able to find an appropriate job for you? If someone offered you a job that you could perform well, would you accept the offer? Would you accept a job that is different from your educational back ground? Would you accept a job if you feel that the wage is not appropriate to the work that you are supposed to perform (lower wage)? Would you accept to be employed informally (no health and social insurance)? Would you tell me what would be the acceptable monthly wage level for you? Do they feel discriminated against? Do you feel that because you do not work, you are being treated differently than men/women that do work? From your: family, friends, Istitutions? 105 22 ' 23' E SERBIA FORMER YUGOSLAV REPUBLIC OF MACEDONIA " 'N ( \ \ - BULGARIA 4, "N ® ® SElEOEO CITIES AND TOWNS MUNICIPAliTY CAPITALS' NATIONAl CAPITAl J " ~ THE CITY Of SKOPJE RIVERS 1 MAlN ROADS RAILROADS ~\ MUNICIPAl BOUNDARIES p,e.CIYO....,.; l -' INTERNATIONAl BOUNDARIES « ' In mos1 cases, the nOlYlej of the m.mjcjpoMjes ~ are identical'" thej, capjtols_ W1.ere they z [ djffer, Phe mlNljcipo/jly is.hewn in green j",lje « '" ~ « ,.;,~ ~ GREECE 4 1'N f 4 !' N \' o 10 20 30 1G1omeIe... ~ /"- ' ! ! I ! r ' ~' i i i i i W > o 5 10 15 20MiIM j;: "I § ,.~ ( .1-- 22 'f