The data recently published by Israel's National Insurance Institute (NII) on the number of poor and the increase in the rate of poverty-stricken children do not present an authentic picture of Israeli society. The NII profile is only partial; it stresses poverty among children, a rate influenced by the measurement methodology used, but does not adequately relate to one sector of the population that is unwillingly entrapped in the cycle of poverty resulting from contraction of Israel's labour market. We are referring to the non-Jewish (primarily Arab) sector of Israeli society. In addition, the poverty index employed does not sufficiently differentiate between the working poor and those excluded from the labour market, circumstances that grant the latter greater leisure but may also enable them to reap greater benefits from this situation than do the working poor.

Determination of the poverty level on the basis of available income, that is, total household income less taxes plus state-distributed direct transfer payments, is deficient due to the fact that only monetary income is considered; generous government transfer payments in services, such as health, education, and housing, are ignored. Eligibility for a considerable number of these services is determined by a means test together with the number of the family's children; thus, members of large poor families benefit more than do members of small poor families. Because the scope of transfers has been expanded during the last decade (e.g., the National Health Insurance Law, 1995), ignoring the value of these services not only biases estimates of the level of poverty upwards (especially among children), it also biases estimates of the very incidence of poverty upwards.

New address: 50, Reines St., 64587 Tel-Aviv, Israel, Tel: 972-3-5271552, Fax: 972-3-5271554

The basic assumption at the core of the NII's poverty index is that each additional child induces additional family expenses at a decreasing rate, an assumption that is commonly accepted as valid. It is clear that the cost of caring for two children is less than twice the cost of caring for one child. The problem with translating this assumption into practice in Israel lies in the belief held by local authorities that each child adds comparatively more to an Israeli family's expenses than does a child in any other OECD country. Considering the inflated cost of each additional child, the level of income required to classify an Israeli family as poor is likewise inflated. In consequence, families with children are over-represented among Israeli families categorised as poor. Recent research conducted by the Hebrew University's Falk Institute indicates that the proportion of small families considered poor (single-member families and couples, primarily the elderly) in Israel has risen considerably in contrast to the proportion of families with many children, which has declined.

In addition, the image purveyed by the NII's poverty index is essentially static. It does not trace the history of the poor; instead, it relays a "snap shot" of poverty at any one point in time. The question therefore arises as to whether the annual poverty index refers to a constant population or whether, perhaps, it also captures individuals who are momentarily in a state of financial distress. This means that the latter's inclusion among the poor represents a sporadic episode in their lives, a phenomenon policy makers consider to be far less troubling on the macroeconomic level. The NII has never dealt with this issue; in effect, no long-term data about poverty-stricken individuals is available given that two-thirds of the poor remain outside the labour market. With respect to the working poor, the Director of State Income has found that the poverty cycle is quite volatile, and that the level of hard core poverty is considerably lower than that reported on the basis of annual income surveys. Among the Jewish population, 25% of those listed as poor in 1983 remained poor in 1995. Moreover, more than 70% of the hard core poor belong to the non-Jewish sector, which represents only about 20% of Israel's total population.

Finally, it is important to note that the index is a comparative instrument. That is, some level of the income received by 50% of the population automatically falls below a specified criterion, a statistical effect that makes individuals in the highest and mid-income groups (i.e., those belonging to the top five income deciles) appear to be relatively richer. This means that even if the purchasing power of individuals in the bottom five deciles remains the same, the poverty rate automatically rises.

All the alternative approaches to the measurement of poverty have identified one group as consistently suffering from poverty despite its progress, expressed in higher levels of education, reductions in family size, and increase in the rate of participation in the local labour market. This group belongs to Israel's non-Jewish or Arab sector of the population.

According to the latest report issued by the NII, whereas the rate of poverty among families in the general population is 18%, the rate among non-Jewish families is 42%. Although the proportion of non-Jewish children among all poor children is on the decline, non-Jewish children continue to represent the largest single group (about 40%) in this population. Contrary to other poor sectors, non-Jews participate in the labour market at relatively high levels: 64.2% (males) in 1999. Furthermore, in comparison to the 25% of the Jews who remained poor between 1983 and 1995, 41% of the non-Jews remained poor during the same period. According to a study conducted by Israel's Ministry of Finance (cited previously), the probability of a Jew remaining chronically poor is 2% while the probability of a non-Jew remaining chronically poor is 19%. Again, contrary to other sectors, the economic effect of education on poverty among non-Jews is marginal: Among 600 poor children aged 16-18 in 1983, 70% of those who remained poor at age 28-30 in 1995 were non-Jews.

Every index currently used to measure poverty indicates a steady trend of increase in poverty and income disparities. A review of Israel's benefits programs indicates that direct transfer payments have burgeoned over the years. Total expenditure for guaranteed income has almost doubled in the last decade although population has grown by only 31%. In the early 1990s, average state expenditure per capita equalled NIS 3,528; by 1999, it had reached NIS 5,157, an increase of 44%. During the same period, GDP per capita increased by 13%, and private consumption per capita increased by 30%.

Considerable additional evidence has been gathered indicating that an increase in direct transfer payments contributes to the growth in income gaps. Since guaranteed income legislation was passed in 1980, the number of individuals receiving this benefit has grown by 1,200%; at the same time, population grew by 55%. We should note that most of the increase has resulted from the rise in the number of individuals "unclassifiable" with respect to other eligibility categories. The number receiving guaranteed income benefits in 2000 reached 125,000. Of these, 52% were aged 31-50. During this period, income inequality also increased significantly (according to the NII, the Gini Index grew from 0.32 in 1980 to 0.36 in 1999). This implies that despite the widespread use of this instrument, income gaps have continued to grow.

Regarding income per se, Israel's minimum wage in 2000 equalled NIS 2,798 per month. Application of the income model employed by the Director for State Income (1997) indicates that about 32% of those assessed (i.e., individuals having some source of taxable income) earned below the minimum wage. The guaranteed income benefit for a family with two children equalled NIS 3,142 in the period surveyed. Furthermore, those eguaranteed income benefits are also eligfor other benefits, including a 90% reduction in kindergarten costs, rent assistance valued at NIS 1,170, a 70% reduction in city property taxes (also paid by renters), reduced public transportation fares, exemption from television ownership fees, and exemptions from doctors' fees. The basket (including the guaranteed income benefit) is valued at more than NIS 5,000 annually. In effect, these benefits provide a negative incentive to a considerable proportion of potential labour market participants. Stated differently, the combination of low wages and a policy of high transfer payments, some of which are obviously indispensable, induces a situation where many do not find it in their interest to work.

The NII's poverty index is problematic with respect to the way in which it measures poverty in Israel, but not in its description of the general trend, whether in terms of the growth in income gaps or the rate of comparative poverty. Nonetheless, modification of the direct transfer payments system does not appear appropriate as a means to resolve the intensifying dilemmas. While direct transfer payments may only marginally improve the quality of life, attention would be better directed to the labour market, in which poor participants receive few social benefits and Arab members are often excluded. That is, a more appropriate solution involves increasing the number of high value-added jobs and access to vocational training.