הרקע הכלכלי למחאה החברתית בקיץ 2011

Transcription

1 TAUB CENTER FOR SOCIAL POLICY STUDIES IN ISRAEL POLICY PAPER SERIES THE ECONOMIC BACKGROUND OF THE SOCIAL PROTEST OF SUMMER 2011 Michael Shalev Policy Paper No הרקע הכלכלי למחאה החברתית בקיץ 2011 מיכאל שלו נייר מדיניות מס' *** All errors are the author s own. The views expressed herein are those of the authors and do not necessarily reflect the views of the Taub Center for Social Policy Studies in Israel. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit is given to the source. The Taub Center was established in 1982 under the leadership and vision of Herbert M. Singer, Henry Taub, and the American Jewish Joint Distribution Committee, the Center is funded by a permanent endowment created by the Henry and Marilyn Taub Foundation, the Herbert M. and Nell Singer Foundation, Jane and John Colman, the Kolker-Saxon-Hallock Family Foundation, the Milton A. and Roslyn Z. Wolf Family Foundation, and the American Jewish Joint Distribution Committee. For more information go to or write to us at info@taubcenter.org.il.

2 The Economic Background of the Social Protest of Summer 2011 Michael Shalev * Abstract The social protest of summer 2011 was largely one of younger Israeliborn Jews. The centrality of this group in the protests may be explained by trends in their socioeconomic position. Analysis of the period between 1995 and 2010 shows that in the five years prior to the summer of 2011, the income of the typical working family headed by a young Israeli-born Jew, relative to all Israeli households, declined to unprecedented levels. The main cause was wage erosion among young adults. This decline also occurred among those with higher education, and their chances of attaining an income in the highest quintile fell substantially, especially among women. Among young Arabs and ultra-orthodox Jews, both individual wages and household income already very low in earlier years declined even further. Of all the population groups, only Russian-speaking immigrants improved their relative income position. Overall in the period studied, there was a decline in the value of those advantages that previously assisted young families in Israel to attain a middle-class standard of living: higher education, two working partners, residence in the Tel-Aviv area, and being an Israeli-born Jew. At the same time, the rising cost of housing has made income erosion a bigger problem. The proportion of young adults living in their parents homes increased, and the share of young home-owning families fell. * Prof. Michael Shalev, Outgoing Chair, Taub Center Social Welfare Policy Program, Department of Sociology and Political Science, Hebrew University. The author is deeply grateful to Haim Bleikh, a Taub Center researcher, who invested vast amounts of time and creative energy in this project and made vital contributions to the design and implementation of the research. 161

3 162 State of the Nation Report S ummer 2011 was an exceptional time in Israeli history. What began as a tent protest focused on housing costs expanded quite quickly into profound criticism of public policy in education, health, taxation, transportation, and more. Central to the protest leaders demands was the reshaping of social priorities in order to reduce economic inequality in Israel. At the same time, the protests appeared to be indicative of economic hardship among younger Israelis, especially in the middle class. These young people as well as their parents expressed frustration, anger, and disappointment at their inability to meet the rising cost of living and a growing sense that the State of Israel had abandoned them. This chapter examines changes in the living standards of young Israeli adults (aged 25-34) from the mid-1990s to the present. The analysis is based on two main variables that reflect standard of living: income and housing. Income is measured both by individual earnings and household income; and the housing indicators relate to individual living arrangements (e.g., with parents vs. independent) as well as the question of home ownership (versus rental) among young families. The results consistently point to declining living standards among younger individuals and families. Some of the measures indicate a steady long-term decline whereas in others the decline is more recent, yet all analyses show a clear decline in the year or two preceding the protest. Moreover, a major impetus for the protest was the sense that economic hardships had reached the mainstream of Israeli society. This widespread sense is confirmed by the results, which show that the decline has affected even those young Israelis who are relatively advantaged Israeli-born Jewish university graduates from central Israel a fact likely to explain their intense involvement in the protests.

4 The Economic Background of the Social Protest of Summer Data and Measures This study is based on the annual income surveys conducted by the Central Bureau of Statistics (CBS) between 1995 and Each annual survey samples about 6,000 individuals and 1,300 young families (with families defined as households consisting of a married couple with children, 1 young families as families headed by an individual aged 25-34). In addition to having a wide sampling that allows breaking the population down by separate variables, the income survey includes individual income information within households allowing an analysis of within household incomes. Since the income information on the selfemployed was not included for the first years of the income survey, all of the empirical analyses in the chapter relate to salaried employees only for both individual and household incomes (i.e., only families headed by a salaried employee were included). 2 Social and political cleavages within Israeli society influenced individuals degree of participation in the social protest. The protest was largely driven by secular, Israeli-born (non-immigrant) young people. Public opinion polls (Haber, Heller, and Hermann 2011) indicate that members of certain groups Arabs, 3 Haredim (ultra-orthodox Jews), immigrants from the former Soviet Union (FSU) expressed reservations about the protest despite these groups relatively limited access to economic resources The information collected in the income survey does not enable identification of cohabiting couples with children or same-sex relationships. Samples since 1997 have ranged from 5,500 to 6,400 individuals and from 1,200 to 1,400 families (rounded to 100s). The 1995 and 1996 surveys were more limited, sampling only 3,000 individuals and 700 families. Data for East Jerusalem residents are missing for certain years and therefore excluded from this study. The terms Arabs and Arab Israelis are used interchangeably to refer to the same population.

5 164 State of the Nation Report Most of the following analyses will distinguish between Israeli-born Jews, Arabs, and immigrants from the FSU. In addition, Israeli-born Haredim will, for the most part, be treated as a separate group. 4 Young immigrants who were not part of the mass wave of immigration from the FSU have not been included in the study due to their heterogeneity. Table 1 shows the relative size of the groups at various points in time. As the data indicate, the population of non-haredi Israeli-born Jews has declined in relative size due to growing numbers of immigrants and Haredim, although it still constitutes nearly 60 percent of the younger population. Table 1. Major groups in Israel s young population (ages 25-34) Israeli-born Jews (not including. Haredim) 63.9% 60.9% 58.9% FSU immigrants 11.0% 13.4% 13.1% Haredim 4.4% 4.7% 5.9% Arab Israelis 17.8% 16.8% 16.8% Other immigrants 2.8% 4.2% 5.2% Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. 4 The CBS data underlying this study do not make it possible to identify Haredi individuals with certainty. Included in this category were all people living in a household in which at least one male member s last educational framework was a yeshiva. Because the population of FSU immigrants barely includes Haredi Jews, no distinction was made between Haredi and non-haredi individuals within that population.

6 The Economic Background of the Social Protest of Summer This study s reliance on CBS income surveys presented several challenges. First, some variables were not included in the survey, or were included but only partially. An example of a variable not available in the surveys is current educational status, without which employed students could not be distinguished from employees who had already graduated. Two examples of variables for which only partial information is available in the surveys are Haredim and household living arrangements. Another difficulty was that over the research period, some 15 years, a variety of changes occurred: in the sample, the variables included in the questionnaire, and their definitions (details are provided later in the chapter, when appropriate). In addition, even though the samples are relatively large, breaking down the findings by age, education, survey year, and other variables could decrease the accuracy of results with regard to the total population. For this reason, the study s findings are best approached cautiously, with an eye to general trends rather than detailed findings for specific years. To minimize the effects of sample errors and technical changes on long-term trends, in most comparisons the annual results have been grouped into sub-periods. 2. Preliminary Findings and Division into Sub- Periods The approach adopted here assumes that the degree to which individuals experience satisfaction or a sense of deprivation in relation to their income depend on two variables: relative attainments and purchasing power. People judge themselves relative to others and are sensitive to the purchasing power provided by their income, relative to their aspirations and expectations. Expectations are not studied here directly, although towards the end of the chapter the issue of housing which figures heavily in young people s complaints about the cost of living in Israel is examined. The main part of the study deals with changes in individual and family incomes relative to the total population, making it possible to

7 166 State of the Nation Report assess differences between groups of young people as well as where each group is situated in the overall income distribution of the population. The relative income position of young working families was calculated in two stages. In the first, the typical young family was identified, defined as the household with the median income relative to all young families. In the second stage, this median income was located in the income distribution of all Israeli households. For these purposes, all households (excluding pensioners) were divided into one hundred equal groups according to their income. The method is similar to that used in creating income deciles except that in this case, percentiles are created. In 2010, for example, the median income of young families was NIS 5,917. In the same year this level of income was in the 47th percentile of all families (the income of the families found in this percentile ranged from NIS 5,871 to NIS 5,979). As in other studies of household income inequality, all calculations were adjusted for family size by calculating standardized per capita income. 5 To illustrate the importance of a relative perspective in understanding trends in young working family incomes in Israel, Figure 1 compares the trends obtained using our preferred measure ( relative income ) with the standard approach, based on average real income in shekels ( absolute income ). The method used in this research replaces the average with the median, since the mean is always skewed upward by those with high earnings, thereby failing to reflect typical group members. In addition, the proposed method (based, as noted, on relative calculations) takes into consideration the possibility that even if a given group enjoys a rise in income that exceeds the inflation rate, satisfaction with income is determined by gains relative to other groups. The analysis in Figure 1 relates to Israeli-born Jews and does not include Arabs, FSU immigrants or Haredim. Comparison of the trends that are found using each of the two income measures reveals strikingly 5 The OECD s equivalence scale has been adopted in this study, in which the standardized number of persons per household is the square root of the actual number.

8 The Economic Background of the Social Protest of Summer divergent results over the past few years. Since 2005 the real income of younger families has remained stagnant followed by a slight decline, whereas their relative income over the same time period fell dramatically by ten percentiles to a level previously unseen in the series. In contrast to the moderate decline indicated by the absolute measure, Figure 1 6 shows that in relative terms the incomes of young Israeli-born Jewish families eroded. It follows that on average, the incomes of the rest of the population were rising faster than consumer prices. Figure 1 Real income versus relative ranking young working families non-haredi Israeli-born Jews, Percentile 64 Income in 2010 prices Income (NIS) 7, Income rank in percentiles 60 7,000 6, , , ,000 Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. 6 In this figure and in all additional tables and charts in this chapter young refers to ages Data relating to household income represent disposable income per standard person (as noted, using the equivalence scale of the OECD). The data refer to households consisting of a married couple with at least one child, in which the head of the household is a salaried employee between the ages of 25 and 34. Total income by percentile was ranked relative to all Israeli households headed by an individual under age 65.

9 168 State of the Nation Report Most of the families that did better than the young families analyzed in Figure 1 belonged to older age groups, although as will be shown later in the chapter, one segment of the younger population immigrants from the FSU were able to improve their relative position significantly. It is important to emphasize that the erosion in the incomes of young Israeliborn Jews relative to the rest of society occurred in a period characterized, according to their claims, by particularly steep increases in the prices of products and services relevant to young families, especially and most sensitively in housing (a subject that will be looked at more closely later in the chapter). Another component of the method of analysis that is worth noting is the division of the 16 years into time periods. An important element in this division is not to give undue weight to years of economic plenty or scarcity, which naturally can affect the relative achievements of the younger population. Figure 2 reproduces the trend over time in the relative income (in percentiles) of young families, but with the addition of the composite State of the Economy Index calculated annually by the Bank of Israel. This measure (shown in the lower portion of the figure) is an indicator of economic activity in Israel. In the chart substantial differences can be seen in changes in the relative income of young families in the first two and last two periods ( versus ). The first two periods are characterized by short-term fluctuations (probably induced by recessions, indicated in Figure 2 by red dots) but a rising trend. Despite cyclical declines, between 1995 and 2005 the relative position of young families rose from the 56th percentile to the 64th. The third period ( ) is characterized by the steep erosion noted earlier, and cannot be explained by macroeconomic factors, whereas the final two years ( ) show no change. By 2010, the year before the social protests erupted, the relative income position of young families had reached an all-time low, and stood at the 54th percentile.

10 The Economic Background of the Social Protest of Summer Figure 2 Changes in the income of young working families income of Israeli-born Jews*, in national percentiles, Percentile Income ranking in percentile State of the Economy Index** State of the Economy Index * Not including Haredim. ** Red dots signify years of economic recession. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. 3. Developments in Individual Wages For most young working parents, the main determinant of the household standard of living is their income from work. According to the Bank of Israel, in the course of the last two decades, the real wages of Israelis rose steadily between 1993 and 2001, then fell sharply as a result of a recession. 7 Since 2002, average real wages have 7 Bank of Israel, 2011 Annual Report.

11 170 State of the Nation Report remained unchanged, except for minor cyclical fluctuations. How did the wages of young people fare in this situation? To see the complete picture, factors that influence wages in Israel like education and geographic location (Tel-Aviv and the center where nearly half of young households are found, versus other areas) must be considered. 8 The analyses that follow encompass all wage earners, including those employed part-time (common among mothers and students). Consequently, the calculations therefore refer to hourly wages. Figure 3 presents the rate of change between the two last periods of the study ( , ) in two variables: nominal earnings per hour and actual inflation. It shows clear indications of erosion of the wages of young people in particular those with higher education in the center of the country, whose nominal earnings actually declined by a few percentage points at a time of significant price rises (9 percent). In contrast, among older workers (35-54 years-old), the average rise in wages was similar to the price rises in all areas. Also amongst those without a higher education there is a gap between generations, although more moderate, and the value of the earnings of young people eroded more than that of the older groups. It should be noted that if the changes are calculated from an earlier point (starting in 2000; results not shown) a similar picture emerges, although it is even clearer. While the cumulative inflation rate went up by some 17 percent, only the hourly wage rate of those with higher education between the ages of rose at a similar rate while the earnings of all other groups declined. It is reasonable to assume that this changed the relative position of young people in the wage structure. 8 Individuals with higher education are defined as ones who attended an accredited institution of higher education and completed at least 15 years of study. The reason for this indirect definition is that not all income surveys included a detailed question about the highest diploma acquired. Central Israel was defined as including the following districts: Sharon, Petach Tikva, Ramla, Rehovot, Tel-Aviv, Ramat Gan, and Holon.

12 The Economic Background of the Social Protest of Summer Figure 3 Recent trends in wages and prices hourly wages of salaried employees, percent change to Not higher education 8.3% 9.1% Higher education 10.3% 9.7% 6.0% 6.4% 4.5% 3.3% 4.1% 3.6% 1.9% Price index % Central areas Other areas -2.2% Central areas Other areas Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. The position of the typical young wage earner in the overall hierarchy of hourly wages can be estimated using the same approach used earlier to evaluate the position of the typical young family. The income of the typical younger worker is defined as the median income of young workers. This median is compared with the incomes of other Israelis divided into percentiles. The division into years and population groups is the same as before, with one exception: since the sample did not include enough young Haredi wage earners, this group was not included in the current comparison.

13 172 State of the Nation Report Figure 4 shows that young Jews who are not immigrants or Haredim have the highest position in the national wage hierarchy and Arabs are positioned lowest. Over the last decade, these two sectors experienced little change in their relative position: the situation of Arabs remained fairly constant, while the highest group experienced a measure of erosion (by three percentiles) but remained in a relatively high percentile. At the same time, the hourly wages of young FSU immigrants increased substantially their position on the earnings scale rose by 11 percentiles. Figure 4 Relative earnings of young adults median hourly wages of salaried employees, in national percentiles Israeli-born Jews* FSU immigrants Arabs * Not including Haredim. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. Alongside the findings in Figure 4, it should be remembered that every sector is comprised of many parts, and individual education is highly important for wage levels. Figure 5 separates the data for each group of young people (with the above-noted exception of Haredim) by education level (with or without higher education) and by gender. The

14 The Economic Background of the Social Protest of Summer overall trend is clear. Over the fifteen years examined, Israeli-born Jews and Arabs at all education levels and of both genders experienced declines in their relative wages. In contrast, immigrants from the FSU, especially those with higher education (about one-third of this group), experienced the opposite trend. Whereas in the late 1990s the wages of FSU-born men and women with higher education were lower by percentiles than those of their Israeli-born counterparts, two decades later the gap had nearly closed, falling to 4-5 percentiles. What does the data reveal about the Arab minority? Here the results are very different for those with and without academic studies. In the first period, the academically-trained group was very small and selective (less than 10 percent in the first five-year period) and presumably had the advantage of filling senior positions. As a result, this group began at a similar or even higher relative position than Israeli-born Jews. Since then, however, Arab professionals of both genders suffered a steeper decline than their Jewish counterparts, although among women with higher education, Arabs still enjoy a slight advantage over Jews. Meanwhile, Arabs without higher education have always been located far below Jews in the wage hierarchy. Returning to immigrants from the FSU, what can explain their remarkable advances, so exceptional against the general background of declining relative incomes among young people?

15 174 State of the Nation Report Figure 5 Earnings rankings ages per work-hour, percentile of median person, by gender and sector* A. With higher education Israeli-born FSU immigrants Arabs Men B. Without higher education Israeli-born FSU immigrants Women Arabs Israeli-born FSU immigrants Men Arabs Israeli-born FSU immigrants Women Arabs * Not including Haredim. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics.

16 The Economic Background of the Social Protest of Summer The 1990s wave of immigration from the FSU included many academically educated individuals. In addition to employment difficulties stemming from language deficits, many immigrants were employed in occupations in their country of origin that were unsuited to the needs of the Israeli economy. Others with more compatible occupations (like doctors) were often relegated to less attractive segments of the Israeli labor market (Sussman and Zakai 1998). Many had to undergo occupational retraining to find employment, while others resigned themselves to less lucrative jobs for which they were overqualified (Raijman and Semyonov 1998). For these reasons and others (for example, the fact that many immigrants from the FSU resided in geographically peripheral areas), many immigrants, regardless of gender or education, began low on the wage scale but had ample potential to improve their position once they had mastered the language and become better acquainted with the local labor market. Note, however, that rather than following the members of one age cohort over time, the current analysis examines the same age group (25-34) in different periods. Since the relevant age group spans ten years while each sub-period spans only two to five years, the study population at any given time consists of individuals from several consecutive cohorts. The findings therefore reflect the characteristics of new cohorts, as well as processes of immigrant absorption experienced by members of earlier cohorts. It is fair to assume that cohort succession had a positive effect on the wage attainments of young immigrants from the FSU. Earlier cohorts, whose members arrived in Israel as adults, and thus were Russianspeaking and Russian-educated, were replaced over time by later cohorts whose members arrived at a younger age and were partly raised and educated in Israel. By the mid-2000s, most academically-educated immigrants aged had studied in Israeli high schools or universities or colleges. In addition, by that time many immigrants had moved away from the periphery, increasing the percentage of FSU immigrants living

17 176 State of the Nation Report in the country s central regions from 38 percent in to 48 percent in This analysis of the percentile rank of the typical wage-earner in each group is based on the median young person s salary. However, changes in the median do not necessarily reflect what is going on with all members of a group. Theoretically, it is possible that the median earnings of Israeli-born Jews fell mainly because of the growth of lowwage employment, leaving the more advantaged members of this group undisturbed. To find out whether high-achievers also suffered a decline in their relative wage, the percentage of young adults whose hourly wages are high enough to be included in the top 20 percentiles (the top quintile) of all employees is examined. This analysis is limited to those best positioned to succeed: those with higher education, aged in their early thirties (i.e., old enough to be on the road to success). Figure 6 shows trends in the position of those aged in different population sectors, for men and women separately. The most striking results pertain to well-educated Arabs: there was a tremendous decline in the proportion of young Arabs in the top wage quintile, especially among women (from one out of two to one in eight). Israeli-born Jews experienced a much more moderate but still significant decline throughout the survey period, which became more severe in the last two years of the study, just before the social protests. By contrast, immigrants from the FSU experienced sharp increases. This rising trend stopped recently, but without turning into a decline.

18 The Economic Background of the Social Protest of Summer % 40% Figure 6 Young adults with higher education in the top quintile of hourly wages as a percentage of all those aged 30-34* A. Men 55% 53% 52% 46% % 36% 43% 34% 30% 28% 20% 16% 20% 16% 10% Israeli-born FSU immigrants Arabs B. Women % 40% 37% 30% 33% 30% 29% 20% 25% 24% 23% 22% 14% 13% 13% 10% 0% Israeli-born FSU immigrants Arabs * Not including Haredim. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics.

19 178 State of the Nation Report Even more striking are the gender-related aspects of the results. Not only did the percentage of women making it to the top quintile decline in each successive period (with the exception of Arab women over some of the periods), but the decline was steeper for women than for men. (In the case of immigrants from the FSU, women s advance to the top quintile was slower than men s.) The decline experienced by Israeli-born Jewish women was double that of their male counterparts; and similarly, the increase enjoyed by male FSU immigrants was almost double that of their fellow female immigrants. At least for those women born in Israel, this result may help explain their central role in the social protests. The differences between sectors are especially striking in the case of Arabs and immigrants from the FSU. Among Arabs, the decline is explained in large part by the contrast between the rapid growth in the percentage of academically educated individuals and the fact that job opportunities are largely restricted to the Arab communities, increasing competition among educated applicants and causing some (especially men) to accept relatively low-paying jobs offered by Jewish employers. By contrast, the higher education acquired by immigrants from the FSU increased in value over time, among other reasons because they now have a better range of opportunities, because they face less discrimination than their Arab counterparts and because more of them now live in the center of the country. Several factors may explain the decline in the relative earnings of young professionals in their early thirties (Shwed and Shavit 2006). First, the doubling of the percentage of young adults with higher education means that some of these individuals have less ability and motivation than in the past. In addition, a growing proportion are now educated at colleges whose degrees confer less value in the labor market than do those of universities. Second, over time some occupations (e.g., primary school teachers) now require degrees, without experiencing comparable increases in relative pay. Third, as there are more educated workers in the labor force, competition is likely to drive down their wages. Fourth, since age at graduation and the age at which stable careers begin have

20 The Economic Background of the Social Protest of Summer both risen, members of more recent cohorts are likely to have less work experience and tenure. Finally, simply having a diploma no longer meets the requirements of an advanced economy. It is possible that some graduates of colleges and universities suffer from a mismatch between their skills and those sought by employers. 4. Developments in Family Income The findings presented early in this chapter revealed that there has been serious erosion in the relative earnings of young working families over time. For several reasons it is important to focus on the situation of families when analyzing the young population. First, young adults with their own families constitute the largest group in Israel s young population, totaling 51 percent (on average over the 15-year period). Second, based on past trends, most young adults who have yet to start families of their own are likely to do so in the future. Third, this group is the most homogeneous. It is more difficult to analyze the earnings of all households headed by a young person because of their heterogeneity there are households of singles, cohabitants, couples without children, single parents, and more (for further details see Section 5). Fourth, focusing on this group is of paramount social importance because they are likely to be the most economically vulnerable. As the social protests underlined, difficulties of coping with the cost of housing and education are likely to be especially pronounced for families with young children. This section provides additional information on how the young families included in the survey were identified in practice. It also introduces several important factors in addition to wages that were analyzed in the previous section which determine the incomes of young households.

21 180 State of the Nation Report The young working families included in this survey were selected using a three-stage process (left-hand side of Figure 7). First, all households headed by a young adult (aged 25-34) were identified. 9 Second, only households headed by a married individual with children were selected. Finally, only households headed by currently employed salaried employees were selected. This excluded from the survey young families headed by non-salaried members of the labor force (including the self-employed). The right-hand column of Figure 7 presents key characteristics of the sample (divided by sector) at different stages of the selection process. The results show that Haredim (and to a lesser extent Arabs) were disproportionately included in the final sample because they tend to start families at younger ages. Nevertheless, they were also more likely to be excluded because of their low rates of employment for the head of household, especially among Haredim. In the end, the share of Arabs and Haredim in the sample was higher than their initial share while the share of Israeli-born Jews was lower and the share of immigrants from the FSU was proportionate to their share in the population. 9 In households with a married couple, the head of the family is usually defined by the Central Bureau of Statistics as the higher-earning spouse, regardless of gender.

22 The Economic Background of the Social Protest of Summer Figure 7 Selection stages and characteristics of the sample of young working families Stages of sample selection Total population aged 25-34** Total 100% Sample characteristics* Distribution of year-olds Arabs Haredim FSU 18% 5% 13% Israeli-born 64% Stage 1: Select head of household Percent married with children Haredim FSU 93% 42% Arabs 66% Israeli-born 44% Stage 2: Select married with children Percent salaried employees Haredim FSU 58% 90% Arabs 72% Israeli-born 83% Stage 3: Select salaried employees Total 100% Final distribution of the sample Arabs Haredim FSU 20% 7% 13% Israeli-born 60% * Averages for ** Not including immigrants who were not from the FSU. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. Table 2 provides data on some important determinants of family income. The analysis presented in the previous section focused on the earning power of individuals, but earning power is only one determinant of the earnings of families with children. The data in the table relates to several additional features of households that may influence their earnings. First it relates to total working hours which depends primarily on whether both spouses work or not. Second, since household income is adjusted for family size, having more children in the family dilutes the earning power of the parents. Finally, gainful employment is not the only

23 182 State of the Nation Report potential source of household income. For most families, transfer payments from the government are the most likely additional source. Table 2. The attributes of different sectors of young working families* Characteristics Israeliborn FSU** Haredim Arabs Working partner 67% 69% 21% 17% Median number of work-hours (per couple) Mean number of children Transfer payments as percent of household income 9% 12% 33% 13% * Averages for ** Immigrants from countries other than the FSU. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. The figures show that the incomes of both young Israeli-born families and FSU families benefit from their relatively low number of children and high likelihood (more than two-thirds) of being dual-earner households. The opposite is true of young Haredi and Arab families (they have relatively large families and both spouses are less likely to work) and serves to drive their family incomes down, although to some extent this is compensated (mainly among Haredim) by enhanced transfer payments.

24 The Economic Background of the Social Protest of Summer Trends in the Relative Income of Young Families This section presents trends in the incomes of young working families relative to disposable household income in the total population (in percentiles). Household income is defined as total income from all sources, after tax and other mandatory deductions, divided by the standardized number of persons in the household. The analysis begins by comparing trends in wages and trends in household income for each of the four population groups included in the study. 10 As Figure 8 shows, the relative position of young Israeli-born Jewish families in the household income hierarchy is six percentiles lower than the relative position of heads of household from the same group in the individual wage hierarchy. Nevertheless, the two indicators exhibit a very similar trend of moderate increases followed more recently by decline. In contrast, until recently the relative incomes of young FSUimmigrant families were superior to the relative wages earned by the heads of these families. During the 2000s, however, this superior position was undermined; as a result, the rapid labor market advances made by young FSU immigrants are only modestly reflected in their relative household income. 10 Due to a limited number of cases, the analysis excludes Haredi wage earners. To make wage and household income data commensurable, the present analysis (unlike the previous analysis of wages) does not include all wageearners aged 25-34, only those who are household heads.

25 184 State of the Nation Report Figure 8 Relative earnings and family income among young people median individual and household in each sector, in national percentiles Israeli-born - wages 60 Israeli-born - income FSU immigrants - income 45 FSU immigrants - wages Arabs - wages 30 Haredim - income Arabs - income Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics. The data in Figure 9A-9D may help clarify this trend. On the one hand, as more FSU-immigrant households moved to the center of the country and continued to work long hours, one could expect their family income to increase. On the other hand, changes in the other two determinants of family income caused that income to fall. First, the number of children in households headed by an FSU immigrant increased. Second, the percentage of FSU immigrants eligible for special benefits earmarked for new immigrants fell over time, shrinking household income from transfer payments among this group.

26 The Economic Background of the Social Protest of Summer Figure 9 Characteristics of young working families, by sector A. Average number of children Israeli-born FSU immigrants Haredim Arabs B. Transfer payments, as a percent of disposable income % 31% 15% 16% 9% 10% 8% 10% Israeli-born FSU immigrants Haredim Arabs Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics.

27 186 State of the Nation Report Figure 9 (continued) Characteristics of young working families, by sector C. Average number of working hours per week both partners work Israeli-born FSU immigrants Haredim Arabs D. Residence in center of country, as a percent of all working young families 46% 53% 48% % 37% 32% 12% 11% Israeli-born FSU immigrants Haredim Arabs Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics.

28 The Economic Background of the Social Protest of Summer Among the two economically disadvantaged sectors, Arabs and Haredim, relative household income has been much lower than relative wages. 11 This has been the combined effect of their low percentage of dual-earner households and their high birthrates. The two groups also differ, however, in several important respects. As Figure 9A shows, members of Haredi households work considerably fewer hours than their Arabs counterparts, and their families are significantly larger. Had it not been for the very substantial contribution of transfer payments, household income would be much lower for working Haredi families than for their Arab counterparts. Turning now to trends in the household income of young working families, Figure 8 shows that over the last 15 years, young Israeli-born Jew and FSU immigrants experienced continuous improvements, though this trend was reversed in the two years prior to the social protest of However, since the long-term improvement was greater for FSU immigrants and the recent erosion smaller, the gap between this group and (non-haredi) Israeli-born Jews fell from 12 percentiles in the late 1990s to only 8 percentiles in Far below both of these groups in the household income hierarchy are the Arab and Haredi populations, which occupy similarly disadvantaged positions. Whereas Haredim started out significantly higher in the late 1990s, young Arab families experienced remarkably little change. Figures 9A to 9D illustrate the many contrasts between the four population sectors with respect to the various determinants of household income (other than wage levels): average number of children per household, transfer payments as a share of total family income, total hours worked by the spouses, and place of residence in Israel. As already noted, some of these determinants changed over time. An interesting question remains: would the four groups still differ in terms of income 11 As already noted, estimated hourly wages are not shown for Haredim due to small sample sizes. However, data for the two most recent periods (when samples were more representative) place Haredi employees (many of whom are female teachers) 7-10 percentiles higher than their Arabs counterparts.

29 188 State of the Nation Report and relative ranking if they all shared the same characteristics except wage levels? This will be the final topic of the present analysis after an examination of several important issues relating to changes in the earnings of young working families over time. Is Declining Income Limited Mainly to the Youngest Families? In choosing to define young as ages 25-34, this research may have exaggerated the situation of most young families, which are headed by persons in the upper levels of this age range. Due to the tendency for age at marriage and at the birth of a first child to rise, by fewer than 30 percent of the heads of young working families were under the age of 30. Declines were especially large among Russians and Arabs; among Israeli-born Jews the fall was less severe (from 28 percent to 24 percent) and among Haredim there was actually an increase. In view of the numerical importance of young families headed by Israeli-born Jews, if the economic situation of the very young families in this group has been deteriorating faster than in the older age group, this might be responsible for the overall decline found in the year-old age group as a whole.

30 The Economic Background of the Social Protest of Summer To assess this possibility, trends in the relative position of young (non- Haredi) Israeli-born Jewish families were measured separately for the and age groups, and compared with the next age group (35-39). As Figure 10 shows, the decline in family income began earlier and was much steeper for families in the youngest age group. In all three age groups, family income increased between the first and the second periods and fell between the second and the third, but only the age group suffered a dramatic decline: a fall of 10 percentiles since This important finding may imply that most (though not all) of the rising economic vulnerability of very young families is actually a temporary setback, which becomes much less severe after the heads of these households reach their thirties. To test this possibility requires following cohorts of new families over the course of the life cycle. Figure 10 Relative income of young working families by age groups median Israeli-born Jewish family*, in national percentiles * Not including Haredim. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics.

31 190 State of the Nation Report Does the Relative Income of Young Families Improve as They Get Older? Because the sample of families investigated in the income surveys changes each year, it is not possible to follow the same members of a cohort as they age. But a good indication can be obtained by tracking adult family members from the same birth cohort in different survey years. Starting in 1995 (the first year of the study), in each five-year period all those aged in the first year constitute a birth cohort, and so on making for a total of four cohorts, the first one of which is observed in all 15 years and the last one (those who were years-old in 2010) only in the final year. The only population group which is large enough for this to be technically feasible is Israeli-born Jews (excluding Haredim). For the purpose of this analysis the selection of households was based on identifying young families by the age of the female spouse In order to increase the number of cases per cohort and minimize contamination of cohorts over time by the addition of couples who married and had children after the first year in which the cohort was observed, instead of selecting households in which the head was aged the selection was based on wives in the same age range, whether or not the wife was classified as the household head. (Note however that the results of a parallel analysis based on household heads were substantively similar to those reported here.) This procedure takes advantage of the fact that most women marry at a younger age than men. As a result both spouses in the selected households tended to be older than those who would have been obtained by selecting according to the age of household heads. This in turn raised the household income level, mainly for newly-observed cohorts which would otherwise have been composed of younger individuals. In an attempt to avoid the aforementioned problem of contamination, in principle households were only selected if the couple was already married in the first year the cohort was observed. However, due to the way year of marriage is grouped in the source data, the desired restriction could only be fully implemented for the first and last cohorts. The cohort that joined in 2000 includes newly married members through 2003, and the 2005 cohort includes them through The analysis is based on roughly cases per year, except for the most recent cohort observed in 2010, with only 245 cases.

32 The Economic Background of the Social Protest of Summer In order to understand Figure 11, consider the four bars indicating household income for households in the first cohort of females aged in 1995 (meaning they were born in ). The bar on the left (in pale yellow) shows that the median income of these households during the first observation period ( ) was in the 59th percentile of all Israeli households. The bars to its right show that in later periods the median income of these households climbed to higher percentiles, probably as a result of higher pay due to more experience and, possibly, additional formal training. Figure 11 Relative income of young working families: tracking four cohorts median Israeli-born Jewish family*, in national percentiles Born Born Born Born * Not including Haredim. Young families were selected on the basis of the age of the woman in the household. Source: Michael Shalev, Taub Center and Hebrew University. Data: Central Bureau of Statistics.

33 192 State of the Nation Report A similar pattern is observed for the next cohort (born ), although the observation period was of course briefer. The two most recent cohorts that also have shorter observation periods displayed worse income trajectories for young families. Strikingly, the initial relative income position of the most recent cohort (born , observed in 2010 only) was percentiles lower than those of all three preceding cohorts at the same stage. 13 In addition, the second-to-last cohort (first observed in 2005) did not enjoy the same surge in relative income that earlier cohorts experienced between their first and second periods of observation. Only time will tell whether future cohorts of young families will continue to suffer from lower initial incomes and less improvement over their life cycle compared with young families started before the 2000s. In the worst-case scenario, the economic opportunities facing successive cohorts of young families will continue to decline, without being offset by improved attainments over time. Is There Less Room at the Top for Young Middle-Class Families? A good deal of the public discussion started by the social protests of 2011 centered on what many defined as a crisis of the middle class. The focus of this chapter so far on the condition of the median young family fits a literal definition of middle. An alternative approach is to focus on two characteristics often associated with the middle class: higher education and relatively high incomes. Accordingly, this section evaluates changes over time in the success of young families in reaching the highest income quintile, distinguishing between those with and without higher education. As explained earlier, any such analysis must take into account the fact that Israelis tend to complete their higher education relatively late in life. 13 This finding should be treated as tentative due to the relatively small number of cases on which it is based.