Bulgaria: Poverty during the Transition

Bulgaria FY99 PA
•	Read the Full Text (7Mb PDF)

The transition from a planned to a market economy has not been easy in Bulgaria. Macroeconomic performance has been worse than the average for Central and Eastern Europe (CEE). Output has dropped by more, and inflation has been higher than other countries in the region. As a result, households have seen a sharper contraction in their standard of living. It was only in 1998 — when output had begun to recover and, simultaneously, inflation had been controlled — that prospects for reversing the decline in living standards had emerged. Promoting a recovery in living standards by continuing these positive macroeconomic trends, accompanied by structural reforms and programs to reduce poverty, is now one of the greatest challenges confronting the government.

Extent of poverty in Bulgaria

In early 1997, over 36 percent of the population in Bulgaria was living in poverty. Moreover, poverty increased sharply over the period 1995-97, when macroeconomic conditions took a sharp turn for the worse. The bulk of the rise in poverty can be explained by the fall in consumption and incomes. However, a significant share is due to widening inequality in earnings and, consequently, consumption. Since this time, the government has implemented a wide range of reforms stabilizing the economy, leading to a recovery in growth and an anticipated reduction in poverty.

The poverty assessment relies on data from two specially commissioned multi-purpose household surveys conducted in 1995 and 1997,¹ as well as research undertaken by the government, Bulgarian and foreign researchers, and international agencies. To measure poverty, the poverty assessment uses a poverty line equal to two-thirds of average consumption in Bulgaria in 1997 (as measured by the BIHS).² While the precise level of poverty is a function of the chosen poverty line, there can be little doubt that there was a substantial increase in poverty between 1995 and 1997 since levels of consumption declined substantially between the two years, and people at levels of consumption in 1995 suffered a clear deterioration in their standard of living.

Who is at risk of poverty?

The analysis shows that about 36 percent of the population, or some 3 million people, were living in poverty in 1997. A more detailed analysis of poverty rates among different sub-groups of the population reveals that poverty rates are higher in rural areas. Over 41 percent of rural residents are found to be poor compared with 33 percent in urban areas. Poverty also varies by region. The regions with the highest poverty rates in 1997 were Sofia and Plovdiv, followed by Russe and Sofia city. Together, the poor from these four parts of the country account for 58 percent of all poor. The higher-than-average poverty rates in Plovdiv, Sofia region and Sofia city are in marked contrast to 1995, when poverty rates in these areas were lower than average. Poverty rates are also found to be higher among ethnic minorities,³ and large households. Also vulnerable are elderly, single women, especially those in urban areas outside Sofia city and in rural areas.

Characteristics of the poor

The level of education and, therefore, earning opportunities of the poor appear remarkably lower than for the population at large. As a result, labor force participation rates are lower and unemployment rates are higher among the poor.⁴ Due to their greater employment opportunities and higher earnings, the non-poor rely more on wages as their main source of income, while poor people depend more on social transfers, including pensions, or a mix of incomes.

The poor have lower access to public services compared with the rich, especially in rural areas. The poor also have lower access to proper sanitation and telephone services. They rely more heavily on coal (for both heating and cooking), wood, and kerosene as sources of fuel. In rural areas, 76 percent of the poor use coal or kerosene stoves for cooking compared with 62 percent of the non-poor. In urban areas, where access to district heating is greater, 55 percent of the poor still rely on wood or coal, compared with 35 percent of the non-poor. Within the city of Sofia itself, 20 percent of the poor use wood or coal for heating, compared with 14 percent of the non-poor.

Not surprisingly, the poor allocate a larger amount of their budget to food (72.3 percent compared with 68.5 percent for the non-poor) and consume larger amounts of cheaper staple grain commodities. In fact, they spend nearly 29 percent of their budget on cereals compared with 14 percent for the non-poor. They also spend a larger proportion of their budget on energy and fuel (14.2 vs. 11.9 percent). Although out-of-pocket spending on education does not appear prohibitively high, enrollment rates in primary and secondary education among children from poor households are below average.

The nature of the BIHS, which followed the same households over 1995 and 1997, allows us to examine the characteristics of those who remained poor over the period covered by the two surveys. Residents of rural areas, ethnic minorities, large households with heads who had primary education or less, and households more reliant on social transfers or a mix of incomes are over-represented among those who are found to be poor in both survey years. In general, households found to be "persistently" poor between 1995 and 1997 are similar to those found to be poor in 1997. The only exception has to do with residence in Sofia city. While residents of Sofia city show higher-than-average poverty rates and are, consequently, over-represented among the poor in 1997, they are under-represented among individuals who find themselves to be "persistently" poor. This suggests that the higher poverty rates observed in Sofia city in 1997 may be an aspect of the 1996-97 crisis and may well change in the future given that Sofia city residents are relatively well-placed to benefit from economic growth (due to, for example, their significantly higher educational achievements).

Labor market developments and poverty

Looking more closely at the labor market we see that the labor market in Bulgaria is fairly stagnant. Employment rates have risen by a fraction in the past four years (from 43 percent to 45 percent of all people aged 15 or older), and unemployment has remained high. Although unemployment has declined recently, it may increase again (at least in the short run) as the restructuring process accelerates.

The turnover of the unemployment pool is low. An unemployed person has a low chance of finding a job.⁵ As a result, many of the unemployed become discouraged and withdraw from the labor force. Average job search duration is long and, thus, the incidence of long-term unemployment is high. This is a worrisome feature of the Bulgarian labor market, as the long-term unemployed face increasingly lower chances of finding a job due to erosion of their skills.

The problem of unemployment in Bulgaria is accentuated by a skills "gap." Many of the unemployed will not be able to find a job even if there are enough vacancies because their educational attainments fall short of those required by employers. Young, low-educated persons, especially those belonging to ethnic minorities, face the highest risks of unemployment. The link between low educational attainment and unemployment is particularly strong.

Many of the unemployed in Bulgaria do not look intensively for a job and confine themselves to visiting labor offices. At the same time, they have excessive wage expectations; that is, the wages they are prepared to work for are much higher than what they can bargain for. These unrealistic expectations may contribute to the persistence of high unemployment. The receipt of unemployment benefits does not appear to reduce the intensity of job searching.

As mentioned above, one factor accentuating poverty in Bulgaria has been the increase in earnings inequality. Between 1995 and 1997, there was a sharp increase in the incidence of low pay. Low pay in Bulgaria is mainly associated with low educational attainment, but also with working in a low paying industry, such as agriculture, trade, and the social services. The incidence of low pay is higher in the private sector. More generally, private sector jobs are less attractive than public sector jobs, although this may be reflective of a greater degree of under-reporting of earnings in the private sector as a form of payroll tax evasion. In particular, the private sector offers substantially lower educational premia, which is peculiar to Bulgaria, since in other transition economies well-educated workers are better off in the private than in the public sector.

Many of the poor do work, but their "earning power" is weaker than that of the non-poor. Thus, the "working poor" account for a substantial proportion of all poor. To a large extent, this reflects the fact that the poor are employed in low-productivity industries, such as agriculture, or work in low-paying jobs in the social services. In families with workers, poverty is a consequence of low earning capacity and low effective labor supply. Both factors tend to go hand-in-hand: households with lower earning power tend to have a lower number of earners.

The incidence of public spending

Bulgaria's system of cash transfers (both social insurance and social assistance), as currently structured, is unable to fulfill either an income replacement or a poverty relief function satisfactorily, nor — given the large numbers of beneficiaries — is it able to raise individual benefits to meaningful levels without upsetting the fiscal balance. Despite relatively high costs, the quality and impact of services within some unemployment programs and social assistance is open to question. Moreover, the present system is probably acting as a hindrance to economic growth, inter alia by raising labor costs in the formal sector and by encouraging increased informalization of the economy.

In practical terms, while the incidence of cash transfers is higher among the poor, and the average benefit received (with the exception of pensions) is higher for poor households, cash transfers are a relatively small share of total household expenditure. They have a relatively small impact on poverty: only about a third of (ex-ante) poor households are moved out of poverty as a result of social benefits. This can be explained by a combination of factors: low incidence, low benefit levels, and, to an extent, poor targeting (over a third of households receiving various forms of social assistance, and 58 percent of those receiving child allowances were not poor before the receipt of the benefit).

Turning to in-kind transfers, we see that education spending is, more or less, equally distributed across the population, with the poor benefiting the most from spending on kindergartens and basic education. Health spending, however, favors the rich because the bulk of public health spending occurs on hospitals, which tend to be used less frequently by the poor.

Conclusions and Recommendations

A. Ensuring macroeconomic stability and undertaking structural reforms

Given the magnitude of output decline (over 30 percent since 1990) and the limitations imposed on the government by the need to maintain fiscal discipline, it should be clear that sustained growth, which raises wages and reduces unemployment, will be critical to reducing poverty in Bulgaria. It will be necessary to establish and maintain a sound macroeconomic environment conducive to growth, reduce inherited price distortions, and create conditions conducive to private sector development. The new government, which has been in place since April 1997, has taken a number of steps to restore macroeconomic stability, and has initiated a program of structural reform, including privatization in the banking, agriculture and financial sectors. A key first element of a poverty reduction strategy for Bulgaria would, thus, consist of maintaining the current policy environment and carrying through much-needed reforms.

One fact highlighted earlier is the greater incidence of poverty in rural areas in Bulgaria. The growth of the non-agricultural sector, coupled with agricultural reforms that provide titles to land, develop land markets, increase the supply and timeliness of agricultural inputs, and improve marketing arrangements will all have an important role in raising average incomes in rural areas, thereby reducing poverty. It will be important to ensure that any initiatives to raise agricultural productivity (e.g., extension services) do not bypass the poor. An investigation into the factors that restrict access to the use of credit, or inputs, by poor agricultural households would also need to be undertaken to guide policy on how to (if possible) reduce these distortions.

Major sectoral reforms are also likely to occur in the energy sector. Given the existing structure of household spending, it will be important to design suitable safety net schemes to protect the poor from unaffordable price increases, and to prevent the adjustment of prices to economic levels from having an adverse impact on the poor

Whatever combination of policies and measures is adopted by the government, it will be important to monitor poverty on a regular basis, as well as to monitor the effectiveness of poverty alleviation programs. Perhaps the first step would be to finalize a poverty line, drawing upon the recommendations of the joint MOLSP/ILO/UNDP study, or using the poverty line used in this study (which is close to the lower poverty line of the MOLSP/ILO/UNDP study). Follow-up work on national and regional trends in poverty and the incidence of public spending will be necessary. Closer analysis of a number of social protection initiatives is also warranted. The capacity to undertake such an analysis will need to be developed.

B. Encouraging growth in employment

Unemployment in Bulgaria is high and is one of the major causes of poverty. Economic growth is essential to increasing the demand for labor and creating new jobs. At the same time, sound labor market policies can encourage employment growth.

Policies for encouraging the growth of employment, relative to the development of skills and qualifications of individuals, include:

Long-term reform of the education system: In order to take full advantage of a growing economy individual skills, especially of the unemployed, will need to be upgraded. Bulgaria does have a well-educated, skilled labor force. As in other economies, there is scope for improvement and over the long term, there are at least two areas of reform in the education sector that would assist in expanding the labor market opportunities of the population: (i) greater emphasis on raising the educational achievement of the population as a whole, by increasing secondary enrollments; and, (ii) greater emphasis on general secondary education at the expense of obsolete or non-transferable vocational skills.

Raising enrollments among children from poor households: While education reform is probably best considered a medium-term objective, the fall in enrollment rates among children from poor households is of immediate concern. Reversing these negative trends will be important if children from poor households are not to be at a disadvantage in the future labor market.

More and better labor market information: While the report finds no evidence that the system of unemployment benefits has significant disincentives for job searching, the unemployed do seem to have excessive wage expectations, which may get in the way of their accepting jobs that are offered to them. This suggests that there is a role for public employment services to provide more and better information on what employment prospects exist and at what rates of remuneration, thereby facilitating a reduction in unemployment.

Improvements in active labor market programs (ALMPs): In addition to running the employment services, the government also conducts three ALMPS: (i) temporary works, (ii) wage subsidies, and (iii) training. Given the limitations on resources,⁶ it would be important to evaluate existing programs from an outcomes, targeting, and cost-effectiveness perspective. The findings could then be used to guide policy development. International evidence suggests that ALMPs are often most effective when they are used to assist the most disadvantaged groups in the labor market. From this point of view, the emphasis that the new Act on Unemployment places on ALMPs aimed at the youth, the long-term unemployed, and ethnic minorities, is to be welcomed. The recently announced review of ALMPs by the government is a very positive step.

Factors that would encourage employment generation include:

Reduction in social insurance taxes: High payroll taxes associated with social insurance is likely to have a strong negative impact on employment creation in Bulgaria, particularly in the formal sector.⁷ A significant reduction of these tax rates will primarily depend on the successful implementation of pension reform (pensions are the largest source of tax), as well as the continued reduction of unemployment insurance contributions. Likewise, the 6 percent payroll tax that will be imposed by the Health Insurance Fund from mid-1999 will need to be offset by corresponding reductions in other payroll taxes.

Encouraging greater flexibility in the market for labor: The Labor Code governing labor market institutions in Bulgaria has had several revisions and is generally well-suited to the needs of a market economy — avoiding some of the most restrictive provisions found in some neighboring transition economies (e.g., prohibition of fixed-term, part-time and additional work, restrictions on enterprise-level collective bargaining, unduly onerous standards for firings and layoffs, etc.). Nonetheless, further improvements can be made, including: reducing requirements with respect to maternity and child leave; reforming sick-pay regulations; and easier termination for misconduct or economic reasons.

Poverty can be expected to fall given a return to economic growth and a rise in average incomes. Simulations indicate that with uniform growth of 3.0 percent per annum for 5 years, the proportion in poverty would fall from 36.0 percent to 27.6 percent, a reduction of over 23.0 percent. At a higher rate of growth, say 5 percent, the proportion in poverty would fall to 25.1 percent after 5 years, a reduction of over 30.0 percent.

The simulations ignore changes in inequality, or the fact that not all individuals may benefit equally from growth. Changes in distribution, as well as growth, will have an impact on poverty. With no change in inequality, poverty will fall from 36.0 percent to 33.7 percent, if there is one-time growth of 4 percent; 32.0 percent, if growth is 6 percent; 30.9 percent, if growth is 8 percent; and 29.9 percent, if growth is 10 percent. With worsening inequality, the poverty-reducing impact of growth is diminished. A widening of the distribution by as much as occurred between 1995 and 1997 would require rates of economic growth of 6 to 8 percent to merely maintain current poverty levels.

C. Strengthening the social safety net

These simulations draw attention to the fact that a strong social safety net that targets those who are least likely to benefit from growth will be essential to reducing poverty. The system of social assistance in Bulgaria prior to 1998 suffered from some key drawbacks:

fragmentation;
inadequate funding;
poor targeting; and
poorly defined financing responsibilities.

As a result, overall coverage of the system is very low. The overlap of programs with different targeting mechanisms undermines the ability to reach the poorest groups in the population. Over a third of households receiving social assistance are not poor (ex-ante). The overall level of spending on social assistance, especially on income support, is low: spending on all social assistance was 0.7 percent of GDP; and spending on income support (the Basic Minimum Income (BMI) program) was 0.07 percent of GDP in 1996. This should be raised to a level of 1.5 percent of GDP to provide a well-targeted basic minimum income guarantee.

Reforms in four areas, some of which have been underway since 1998, would increase the poverty reduction impact of the social assistance system:

Reduced fragmentation: Following the passage of the Social Welfare Act early in 1998, the government has taken key steps to reduce fragmentation, including combining eligibility criteria for the BMI program and the program of in-kind support for utilities. However, there is scope for further consolidation. In principle, this could proceed to the point where the means-tested BMI program becomes the principal source of social assistance.

Increased funding: In addition to expanding the envelope for social assistance, as indicated above, more resources could be made available for the poorest by looking for savings within programs. Increased funding could also be sought by cutting back on untargeted programs. Perhaps the most significant program in this context is maternity leave for uninsured mothers, which costs over twice as much as the BMI program (0.2 percent of GDP versus less than 0.1 percent for BMI), and which benefits rich and poor alike. Finally, increased NGO participation in the provision of social assistance and social welfare is likely to further expand funding, as well as raise the standards of care for clients in social care institutions.

Improved targeting: In general, social assistance in Bulgaria appears to be relatively well-targeted. However, there is scope for improvement. Discretionary social assistance payments, for example, often tend to be regressive (i.e., they accrue disproportionately to the non-poor). The government has already taken steps to reduce the scope for leakage from the system by altering the provisions for one-off discretionary payments to households. There is a need to further analyze the extent of leakage from the BMI scheme in order to determine and reduce the extent of leakages.

Clear responsibilities: The decentralization of social assistance in Bulgaria has generally not worked well. However, the new Social Welfare Act is a step in the right direction. It clarifies institutional roles and helps strengthen the administrative capacity of the system. In order to improve the viability of the scheme, the BMI program will be funded on a cost-sharing basis between central and municipal budgets, and incentives for municipal authorities to meet social welfare payments on time will be introduced. The central government budget will provide earmarked funds toward program costs, as well as an additional contribution to the general municipal budget if social benefit payments are made on time. Implementation deserves careful monitoring to ensure that the program develops on the lines envisaged.

Notes:

The surveys were carried out by Gallup International, Sofia with technical support from the Ministry of Labor and Social Protection, the National Statistics Institute, and the World Bank. They shall be referred throughout as the Bulgarian Integrated Household Survey (BIHS) with the relevant date — 1995 or 1997.
Since 1991, when poverty was first formally recognized in Bulgaria, the country has lacked an official poverty line. The absence of a meaningful official poverty line has complicated the task of determining the level of and monitoring changes in poverty. A recently concluded joint project of the Ministry of Labor and Social Policy and ILO/UNDP has however made specific recommendations on the subject. The poverty line used in this report (BGL 45,466 in June 1997 prices) is close to the lower poverty line (BGL 40,000) recommended by the joint team.
Over 40 percent of Bulgarian Turks and 84 percent of the Roma were found to be poor in 1997.
The unemployment rate among the poor (33 percent) is over twice as high as among the non-poor (15 percent). Only one-third of the working-age poor are employed, compared with almost one-half of the non-poor.
According to the BIHS, the probability that an unemployed person will find a job within twelve months is around 6.2 percent. This is roughly one-sixth of the yearly exit rate from unemployment observed in CEE countries such as Poland, and less than one-eight of those prevailing in high unemployment Western European countries such as Spain.
Total spending in 1996 on all ALMPs was less than 0.25 percent of GDP.
Social insurance taxes in Bulgaria currently average 47.5 percent of net wages.

Permanent URL for this page: http://go.worldbank.org/NZPDUXA640