Two factors have led to an increase in poverty incidence (headcount) in Hungary: a decline in overall national income and a more unequal distribution of income. Poverty in Hungary is "shallow": incomes are densely concentrated around the poverty line. Therefore, a small increase or decrease in income levels can lead to a substantial change in the incidence of poverty. Growing income inequality has also played a part in the rising incidence of poverty. The poorest 10 percent of the population have suffered the largest decline in real income.
Hungary does not have an official or even widely used poverty line. Within the benchmarks utilized for the social transfer system, only the minimum pension has a poverty connotation. It is also used as the basis for various social assistance programs. The minimum pension in 1993 was worth only 73 percent of its real 1989 value and is used in this analysis to identify the very poorest individuals in Hungary. There is a strong tradition in Hungary of using relative poverty lines, often set as a fraction (1/3, 1/2, 2/3) of mean income/expenditure. This approach is also used.
Income from employment, self-employment, and business (but primarily wage income) is the main contributor to individual and household income. Of sources of household income, employment income (wages) was already the most unequal, and its dispersion across the income groups increased markedly in the transition years. Aggregate labor incomes declined by almost as much as the decline in GDP during 1989-93, but the decline was much more a function of job losses than of real wage declines. As such, those who kept a regular job fared reasonably well, particularly in so far as formal-sector jobs provide access to second jobs, an additional source of income. Highly educated employees in private sector companies, often associated with foreign investment, have seen substantial real wage gains. In contrast, less educated and semi-skilled workers and public-sector employees have experienced real wage declines. But, the transition has affected most severly the unemployed, casual workers, those on extended child care on state support, and those who have withdrawn from the labor force altogether (in many cases on disability pension arrangements). Their access to regular wage income has essentially ceased.
Among the socioeconomic groups, the deepest poverty is among households whose head is unemployed, temporarily employed, or dependent on child care benefits as the main source of income -- about five percent of the population. Approximately one-fifth of such households live below the minimum pension. Poverty among pensioners is slightly above average but not a serious phenomenon. Very elderly female pensioners living alone, however, are among the very poorest. In households affected by unemployment, poverty is deeper if the head of household is unemployed than if another member of the household is unemployed. If the head is unemployed and does not receive unemployment insurance benefits, then incidence of poverty is very high: more than 40 percent of such households are below the minimum pension.
Demographically, the incidence of poverty is lowest in nuclear households with one or two children. It rises steadily with the number of children and is especially deep in households with two adults and four and more children, and in households with three adults and three and more children. Among these households, one in five lives below the minimum pension. Poor children live primarily in villages, and in households whose head is poorly educated and does not have more than a temporary attachment to the labor market. This indicates the likelihood of inter-generational poverty. Hungary displays the normal inverse U-shaped lifecycle pattern of poverty, but the age effect is stronger among the young than the old, reflecting the relative well-being of pensioners and the high unemployment rate among the young. There is a distinct gender dimension to poverty. Poverty is deeper among female- headed households, especially if they are single adults with children.
Regionally, Budapest has the lowest incidence of poverty, while the predominantly rural North and South Plains have the highest incidence. This disparity is consistent with the finding that poverty is deeper in villages than in cities.
The Cash Transfer System
Hungary offers a generous array of cash transfer programs. The transfer system includes contributions-based social insurance, income supplement entitlements, and means-tested social assistance. This report finds that, in 1993, cash transfers were widely distributed in Hungary, benefitting no less than seven million individuals and 91 percent of households. Of the six transfers examined, pensions, family allowance and social assistance were found to benefit the largest number of households. Pensions alone reached 52 percent of households, and the family allowance reached 44 percent of households. One of the most striking findings was that the aggregate value of cash transfers was very uniform across households. (While this was not an unexpected outcome for some programs, such as pensions and family allowance, because of the design of those programs, such a flat distribution should not occur with means-tested programs like social assistance.)
As the share of labor income in aggregate disposable household income has declined in the transition years, cash social transfers that, in the aggregate, maintained their real value through 1993 have become even more important. In 1993, they provided around 38 percent of gross household income, accounting for 20 percent of GDP. In the past two years, public expenditure on cash transfers has fallen, and since the number of recipients has remained largely unchanged, benefit amounts have declined. Different cash transfers have played different roles during the transition and some have been more effective than others at preventing poverty. In general, social insurance-based transfers tied to previous wages and designed to replace wage income (pensions, unemployment insurance benefits and the child care fee) have been most effective at maintaining incomes, largely due to program size and the high value of payments. (These transfers alone account for 15 percent of GDP). Social assistance has been less effective at alleviating poverty, due primarily to poor targeting and inadequate transfer amounts for those most in need.
In a static sense, cash transfers have prevented many households from falling into poverty. Indeed, cash transfers have held about 60 percent of households above two-thirds of mean household expenditure, or 45 percent if pensions are excluded. On the other hand, because of their wide distribution and their relative "blindness" to need (or household income level), cash transfers leave many transfer recipient households in poverty. More than one-quarter of households who receive social transfers remain below the poverty line.
Of considerable interest, both as a counterfactual to the recent past and to the present time when expenditure reduction is needed for macroeconomic stabilization reasons, is the extent to which cash transfers could be more effective poverty alleviation tools. (It is acknowledged that they have other important objectives.) It is clear that they could do a great deal more to help the poor. There are two fundamental aspects that prevent more effective poverty alleviation:
- There are too many beneficiaries, but there are also unintentional exclusions, mainly from discretionary (social assistance) programs. Programs would do well to be more narrowly focussed (targeted).
- Because of the large number of beneficiaries and the flatness in their distribution (with respect to household income), cash transfers are of an inadequate amount when they are most needed. Benefits remain largely unreconstructed from the past when there was little income differentiation and uniform payments made sense. That is no longer the case, and the level of benefits needs to be much more carefully tied to need.
Poverty Strategy -- Restructuring Social Transfers
Dealing with Shallow Poverty
The appropriate policy response to the phenomenon of shallow poverty is complicated because people in this poverty category are a heterogeneous group -- included here are some pensioners, low-paid employees, some of the unemployed, and some of the households with a large number of children. The appropriate policy responses for each of these groups differ. Moreover, evidence suggests that the movement of people (households) in and out of shallow poverty is substantial, and incomes around the higher poverty line are very close together. Differentiating people at this level of income has little validity.
The resumption of economic growth, creating new jobs and generating productivity-based real wage growth would be an effective cure for much of the population in this income range. In addition to benefiting those active in the labor market, it would probably attract discouraged workers back into the workforce, and provide a revenue base for higher pay-as-you-go pensions. Beyond this, however, specific policy interventions in the form of reformed cash transfers could be needed for large families.
Households with three or more children are a population group that are over-represented in shallow poverty. This number of children is closely correlated with the low educational level and poor labor-market status of the household head. As such the number of children can also be used as proxy for other variables closely associated with poverty but which may be more difficult as a characteristic for targeting. This interaction between many children in the family, poor educational attainment, and poor labor-market status creates inter-generational poverty. As a poverty group, these households should be high on the priority list of Hungary's policymakers.
The family allowance is the main cash transfer program for preventing poverty among children; it has clearly made an important contribution to this goal in recent years. But previous research has shown that the family allowance can do even more, and many recommendations have been forwarded to improve its targeting. (The recommendations have focused on taxing the family allowance, targeting by categorical indicators, using means tests, and applying a combination of these.) This report confirms that the family allowance could play a much more critical role in alleviating poverty. Targeting the family allowance more effectively could virtually eliminate poverty below the higher poverty line.
Measures approved by Parliament in 1995 (Act CXXII) tackle universal entitlement to the family allowance for the first time and attempt to target the transfer to needier households with children. The retention of the family allowance in all families with three and more children is a particularly important feature of the reform. These measures represent a major step forward.
Helping Those in Deep Poverty
Among the pockets of deep poverty, one group is particularly vulnerable, the long-term unemployed, who have exhausted their wage-related unemployment insurance benefits and have been unable to find a job. It is unlikely that economic growth will lift these people out of poverty without other policy interventions -- this has been the experience of Western Europe. In theory, the long-term unemployed who are active job seekers are entitled to a means-tested social assistance program, introduced in 1993 for this particular target group. Yet, the level of payment, up to 80 percent of the minimum pension, is a very modest amount, and quite insufficient to have any meaningful impact on severe poverty. This program needs urgent review.
Pensioners are a particularly important population group, numbering almost three million. Cash transfers paid as pensions have both a significant absolute and relative impact on poverty: they keep more than 60 percent of recipient households out of poverty. But this comes at a high cost, currently borne by the working population, and it leaves one group of pensioners -- elderly women -- in deep poverty. While still meeting their other objectives (income smoothing in old-age, and mandatory saving), pensions could be made more effective at combating poverty. One mechanism would be to raise the level of the state-provided minimum pension and ensure that it becomes a genuine minimum such that no one entitled to a pension would receive less. This scheme could easily be affordable without any additional expenditure if the value of state-provided pensions received by top income households (pensioners) were reduced.
GYED and GYES (child care fee and child care allowance) were merged in April 1996, and became contingent on means tests, whose income thresholds would be similar to the family allowance. The replacement allowance is equal to the minimum pension. The analysis in the report finds women on child care leave among the poorest, and although GYES is one of the most effective programs at reaching the poor, its benefit level is insufficient to lift these families out of poverty. Some GYED and GYES recipients will return to work, assuming that they have a job to return to, and someone to take care of their children. In this case, the income/expenditure outcome would be positive. For those who do not (cannot) return to work, the poorly educated and low skilled, the situation for some (especially single parents) could deteriorate. Additional interventions might be necessary.
Hungary's social assistance programs are the least effective of all cash transfers at alleviating poverty. Given their function as the last layer of the social safety net, this is a serious failing. The programs suffer from two major problems: benefit leakages and omissions, and low-levels as well as large variations in payments. Too many people are currently receiving too little social assistance to ensure any meaningful poverty alleviation where it is most needed.
Some of the shortcomings of the social assistance system are embedded in the design of the 1993 Social Service Welfare Administration and Social Services Act (the Social Act for short). There are three basic problems with the 1993 Social Act: the income ceiling is too high, ineffective and outdated standards for assessing eligibility, and a lack of a national minimum level of support (only local minimums exist). What exists, therefore, is a mixture of some modern, Western approaches to social assistance, overlaid on subjective, case-by-case eligibility criteria that are remnants of the previous system.
The categories of citizens entitled to a nationally mandated social assistance program could be extended to include others with a high probability of belonging to poor segments of society. Benefits should then be set at levels ensuring the beneficiaries are lifted above the lowest poverty line, but without creating undue disincentives to work. The system would have to be reasonably coherent and administratively simple so that benefit levels could be made more substantial and the potential target group reached more effectively. At the same time, the income ceiling governing eligibility could be lowered. A maximum of twice the minimum pension and possibly only 1.5 times it (preferably on an equivalency rather than per capita basis) would be appropriate.
The government has invested considerable effort in a robust and comprehensive database. This study utilizes two household data sources: the 1993 Household Budget Survey (HBS); and the 1992-93-94 Household Panel Survey (HPS). The 1993 HBS, which is the main data source, is part of a tradition of budget surveys undertaken by the Central Statistical Office (CSO) since the early 1950s. The surveys are conducted every two years, and 1993 is the first HBS that fully incorporates western economic concepts. The 1993 sample is about 9,000 households, selected in a two-stage stratified design, and covering the whole non-institutional population in Hungary. Although not possible to quantify, the HBS probably fails to capture some of the very poorest, namely the homeless. This is a common problem with income/expenditure surveys. The HPS's main advantage is that it follows the same households over time, thus permitting study of the dynamics of poverty. It is this feature of the data that is utilized in this report. The drawback is that the sample size is only 2,000 households, so that the amount of disaggregation in the analysis is limited.
Hungary: Priority Poverty Indicators
|Poverty Lines (1989 Values)|
|Upper Poverty Line (minimum pension)|
|Lower Poverty Line (1/2 mean expenditure)|
|Nominal GDP (Billion Forint)|
|GDP Per Capita (US$)|