Macedonia, FYR: Focusing on the Poor

Poverty is not a new phenomenon in the country. Prior to 1990, the former Yugoslav Republic of Macedonia was one of the poorest Yugoslav Republics heavily subsidized by transfers from Belgrade, and greatly reliant on internal markets for trade.¹ Although literacy was almost universal (94 percent according to the 1994 Census) and life expectancy (72 years) at par with many higher-income transition economies, infant mortality rates compared unfavorably with Central and Eastern Europe norms, and were particularly high in rural areas and for ethnic minorities.² The poorest households were mainly rural and derived their income primarily from agricultural sources.³ A cradle-to-grave protection system was also in place with the intention of providing individuals with full protection against income loss and other lifetime risks. The system provided guaranteed employment, child allowances, free education and health care, and pensions upon retirement. However, open unemployment was high compared with regional norms, and was concentrated among new entrants to the labor force. The socialist worker management system gave the employed (or insiders) extensive powers to push up wages, restrict hiring and termination practices, and keep out younger and less skilled workers.

The transition exacerbated the poverty situation in the country. Independence from Yugoslavia, the loss of federal transfers, the external conflict among its neighbors, and the economic transformation from a socialist system to a market economy led to a sharp economic contraction. Inflation rates soared, real wages declined and unemployment increased. Not surprisingly, poverty increased from 4 percent of the population in 1991 to approximately 20 percent in 1996. The increase in poverty was not uniform. Poverty rates remained largely unchanged between 1990-91, but increased sharply between 1993-95. The main reason for the increase in poverty was a major decline in real consumption. Inequality in consumption grew, but contributed far less to poverty growth.

In 1996, the country overcame the deep recession of the early 1990s and realized a small positive growth rate (1 percent) for the first time since independence. However, the incidence of poverty increased further between 1995-96. As real consumption remained roughly constant over this period, the main reason for poverty growth was a growing inequality in the distribution of income.⁴ The modest growth realized in 1997 and 1998 may have reduced poverty over 1996 estimates, but these gains may well be eroded, at least over the short term, by the contraction of economic activity resulting from the Kosovo crisis.

The diverse characteristics of poverty. In 1996, according to the official poverty line, approximately 20 percent of the population was poor. An overwhelming share (two thirds) of the poor lived in rural areas. About 60 percent of the poor lived in households headed by individuals who worked.⁵ Of these, almost two-thirds were wage/seasonal workers, and the rest were farmers. The remaining non-working poor were roughly evenly divided between pensioners and the unemployed/transfer recipients. The report identifies three distinct groups of poor:

• The traditional poor. Rural, agricultural households with more than three children, as earlier, have among the highest incidence of poverty in the country. These households are larger in size (with many children) and are characterized by the very low education status of household heads and, particularly, of female household members. The poorest are those that survive exclusively on agricultural income.

• The new poor. The transition has changed the face of poverty in the country. Poverty remains a rural phenomenon, with nearly two-thirds of the poor living in rural areas. However, non-agricultural households (now two-thirds of all poor) have replaced agricultural households as the largest group of poor. Among socioeconomic groups, the unemployed/transfer recipients have emerged as the new group of poor and have the highest incidence of poverty, followed by farmers, and wage earners. Pensioner-headed households have the lowest poverty rates overall.

• The chronic poor. The elderly poor, particularly non-pensioners, the disabled, aged farm households without pension or other fixed income, and individuals not in households (institutionalized elderly, disabled, orphans) emerge as the most disadvantaged groups in society. These groups are least capable of work and are, therefore, the least likely to benefit directly from economic growth.

The common characteristics of the poor

In 1996, the common characteristics of poverty included limited asset ownership (durables and land), a large household size, and a high share of food in total consumption. However, low education attainment, labor force status, poor living conditions and low health status are also closely linked to poverty. While ethnicity is likely an important dimension of poverty, the sources of data for the report were not adequate to explore this important issue.⁶

• Low levels of education: Nearly 80 percent of the poor live in households headed by individuals with primary education or less. Rural household heads report a much lower education status compared with urban households. Most worrisome, poverty rates are much higher in households with non-attending children (25 percent) vs. those with children attending school (10 percent). This means that poverty among these households may persist in the future.

• Labor force status. Unemployment, low-wage share, and lack of enterprise ownership are the main labor market determinants of poverty. Low-wage share, indicative of low-wage jobs, is an important indicator of welfare in FYR of Macedonia in both rural and urban areas. However, unemployment is a strong poverty marker only in urban areas. In contrast, remittances and the ownership of a household enterprise reduce the chances of being poor in rural, but not urban, areas.

• Poor living conditions. Living conditions (very poor sanitation, lack of water supply) are a hallmark of poverty in FYR of Macedonia. These conditions are particularly worse for the rural poor but are also difficult for peri-urban residents. Electricity is universally available in both rural and urban areas, but the poor often cannot afford to pay for this service.

• Low health status. Low education levels (particularly for women) coupled with poor living conditions contribute to the low health status of the poor. The higher infant and maternal mortality rates in rural vs. urban areas indicates that this problem is more severe in rural areas.

• Low-wage less-educated workers have lost in both relative and absolute terms during the transition. The returns to education are high in the country. Well-educated and top-paid workers have gained in both absolute and relative terms over less-educated and less-paid workers.

• Private sector growth has helped employ a large number of the low-wage poor. The private sector has grown mostly in trade, services, and in agriculture. Over half of low-paid jobs are in the private sector.⁷ The private sector also employs the top-paid workers in the country.

• Poor informal sector workers complain of the occasional and seasonal nature of work, the inability to get wages on time, the problems of surviving the winter, and the difficulties of starting up small enterprises (fees, limited collateral, and high interest rates). Seasonal workers have the highest poverty rates among all workers.

• The majority of the unemployed are new entrants to the labor force. Unlike many countries in the region, layoffs comprise a small proportion of the unemployed, indicating that limited entry, not exit, is the reason for unemployment.

• The duration of unemployment is extremely high (relative to other countries in the region). Unemployment of long duration discourages labor force participation, depreciates labor force skills, and signals a "less productive worker" to the employer.

• New private sector hires match the characteristics of the unemployed. These are less-educated, less-skilled, and mainly urban workers. The only exception are women, who are least likely to be hired by the private sector. Women are also more likely to receive lower wages for the same work, particularly in the private sector.⁸

• Indirect costs: The need to earn a living to support their families is an important reason for not participating in the secondary education system for the rural poor. Some secondary age children work. This is more characteristic of rural areas than urban areas.¹⁰

• Direct costs: The cost of attending school is another reason for lack of participation in the school system. The poor spend more on education per enrolled child (relative to their income) than the rich. Books and supplies are the most important components of private costs in primary and pre-school education. Transportation costs dominate private spending at the secondary and post-secondary levels, while admission fees are important at the tertiary level.

• Primary education is the most effective¹¹ public program. Public spending on the program is also egalitarian. Primary education subsidies comprise nearly 13 percent of total consumption of the poor. The poor receive a share of education spending that is roughly equal to their share in the population.

• Secondary and higher education spending is less effective than primary education spending; it is also strongly non-poor. The top 20 percent of the population receives nearly half of all spending on higher education and scholarships; and over 20 percent of total spending on secondary education. The low effectiveness and efficiency of public spending on education for the poor merits considerable concern.

• The pension system has protected pensioners against poverty. Poverty rates for pensioners are among the lowest compared with other population groups. However, many elderly women are not covered by the pension system and, therefore, have high poverty rates. This is mainly the result of low labor force participation of women in the past, which means that they do not receive pensions.

• The protection of pensioners against poverty has come at a very high fiscal cost imposing a large tax burden on current workers. This decline in the number of contributors due to early retirement, growing unemployment, and a very generous pension system (relative to other countries in the region) is the source of current financial problems of the system. In the future, an aging population will lead to increasing pension fund deficits.

• Benefits have often not been paid on time. There are often two- to three-month delays in benefit payments to the poor.

• The new decree may have reduced inclusion errors (many ineligible households receiving assistance) which, in 1996, were found mainly in urban areas. Income is difficult to measure in a transition economy, where income sources are variable and records and registration procedures are in a nascent form. As a result, the new program may have to be fine-tuned to correct any remaining problems. The report finds that the exclusion errors (eligible households who do not receive assistance) of the social assistance program are very small.

• Information difficulties constrain delivery of benefits (to clients) and verification of claims (for social welfare officers). Social assistance recipients claim that lack of information about the program, delayed benefits, and onerous documentation requirements limit the program's effectiveness. Social workers also relate difficulties in monitoring and penalizing false claims.

• A growing number of (mainly Macedonian) elderly are becoming institutionalized as family structures break down either due to financial problems or as a result of the migration of younger family members from rural to urban areas.

• Disabled children and adults are institutionalized under poor conditions and receive limited care and attention. Their education possibilities are limited.

• Orphans are also a vulnerable group. There is a growing problem with older institutionalized children (18 or so) who are having difficulty integrating into society, particularly given the lack of jobs. On the positive side, there appears to be some demand for adoptive children in FYR of Macedonia. However, adoptive parents are not screened properly and post-adoptive monitoring is limited.

• Disability allowances for children and foster care allowances have often been delayed due to budgetary constraints.

• The promotion of economic growth that increases employment and real wages to help alleviate poverty among the new poor. Easing rigidities in financial, land, and labor markets would help ease constraints for the promotion of small and medium enterprises.

• Investment in human capital (both its quality and availability), particularly of the rural poor, to promote economic growth and alleviate long-term poverty;

• Well-targeted, financially viable cash transfers (with adequate work incentives), for the short-term unemployed;

• Investment in community infrastructure (water, sanitation, electricity), particularly in rural areas, to improve living standards. This investment in physical infrastructure, together with investment in education and health sector reform,¹² would help improve the health status of the population.

• Community-based care (and cash/in-kind transfers), based on an assessment of the financial and economic feasibility of these programs, for the chronic poor. Such programs could potentially focus on children-at-risk, the disabled, and non-pensioned elderly.

• Continuous poverty monitoring to identify changing poverty patterns and to assess the impact of the government's poverty alleviation programs.

• Reduce high labor costs. The government has to take important steps in this regard: e.g., hiring restrictions and excessive termination restrictions on small entrepreneurs should be reduced, and the court review of the rationale for small-scale terminations should be severely curtailed.

• Reduce remaining labor market rigidities and adverse work incentives. These include: (i) eliminating or significantly reducing special benefits to the unemployed with 25 years of contributions. (ii) further reducing the duration of unemployment benefits.

• Reduce high payroll tax rates and contain the growing informalization of the economy by reforming the social insurance systems (health, pensions, and unemployment insurance). Arresting the informalization of the economy would help improve tax compliance, allow an enforcement of (less-restrictive) labor contracts (including addressing discrimination against women), and promote timely payment of wages in the private sector,

• Minimum wages, if introduced, should be kept low as possible. There is no public minimum wage in the country and this policy should continue given the high level of unemployment of unskilled workers. Minimum wages tend to discriminate against the hiring of less-skilled, poorer workers and perpetuate poverty among this group. However, if public minimum wages are introduced, they should be kept as low as possible.

• Increase participation of the poor in the education system The participation of poor students at all levels of the education system should be increased, but efforts should focus on the rural poor, and on secondary and higher education. Efforts should also focus on integrating minority populations (e.g., gypsies) and those with disabilities (cost effectively) into the education program. While qualitative reports pointed to a potential cultural, security constraint to the participation of Albanian girls in secondary education, it was difficult to assess ethnic differences using available data.

• Reduce the direct costs of the education system. Reducing the costs of education for the poor is one way of increasing their participation in the education system. The direct costs are essentially textbooks and books/supplies at the primary school level, transportation/boarding for the poor at secondary and higher education levels, and fees at the tertiary level. Introducing merit and needs-based scholarships and redirecting boarding subsidies towards the poor might be considered for increasing the enrollment rates of poor students.

• Improve efficiency of education spending, including targeting the poor. Real spending on education should not be reduced. However, there is scope for (i) shifting public spending on education from non-poor to poor through greater cost recovery in secondary and higher education; (iii) shifting public spending over time away from financing wage costs to non-wage investment costs; (iv) shifting social spending from current benefits (mainly pensions) towards spending on education.

• Reduce disincentives to work. The level of the social assistance benefit is established on the basis of fiscal considerations, with the poverty line as a point of reference. However, minimum benefits also need to coordinate with wage developments. Currently, the minimum income for a single person household is 17 percent of average wage, but for a four-person household (the average family size) the minimum income guarantee is 40 percent of wages. Given the growth in low-wage jobs, this level of benefit (almost equal to the minimum pension) reduces incentives to work. This benefit scale reflects recent reductions, but will have to be adjusted downward in the future if it constrains the beneficiary outflow from the system. In addition, while the recent changes in the decree may have reduced inclusion errors, further tightening of eligibility conditions may be required to further reduce false claimants from gaining access to the system.

• Work incentives may be difficult to administer and should be monitored carefully. The introduction of a four-year term limit to social assistance, and a gradual reduction in benefit over this time, may be difficult to enforce and administer. The ability of this incentive to keep individuals off social assistance rolls will have to be monitored carefully and fine-tuned over time.

• Pilot proxy means test approach to social assistance benefits This approach targets social assistance to the poor based on simple, identifiable characteristics of poverty, thereby reducing the administrative burden of the social assistance system. Empirical estimates indicate that such a program could help identify the bulk of the poor in the country, but an assessment of its fiscal costs and targeting efficiency vis-a-vis. the current program and potential implementation issues would require further work and a pilot test.

• Monitor poverty incidence and the effectiveness and efficiency of cash transfer programs. The incidence of poverty, program effectiveness and efficiency measures, based on the Household Budget Survey, should be investigated on a regular basis. A periodic sampling of client concerns in all programs through a social assessment of poverty should be instituted for all main social programs. This social or client-based assessment could be carried out by the Ministry of Labor and Social Policy, with the involvement of other ministries, and with the assistance of research institutes and the Statistical Office. It could also be investigated through an add-on questionnaire along with the Household Budget Survey.

• Improvements in household survey sampling efficiency and measurement. Measures include improvements in the sampling efficiency of the survey, the measurement of consumption (less detailed), and the separation of the consumption and income modules to reduce under-reporting of income.