Miri Endeweld and Haya Stier - Intragenerational Wage Mobility and Inequality in Israel 1990-2015

Intra-generational Wage Mobility among Young Adults, 1990-2015 Haya Stier,*Miri Endeweld** *The National Insurance Institute (NII) and the Hebrew University in Jerusalem **Tel Aviv University 1

Background Job search theories suggest that the early labor market years are a period of rapid wage growth as well as extensive job mobility (Burdett, 1978; Jovanovic, 1979a, 1979b).

New entrants to the labor force can look forward to about 40 years of work. Their wages are expected to double along their careers and they are likely to change jobs ten times. But the pace of these changes is far from even. The first ten years of a career will account for 66 percent of lifetime wage growth (Topel and Ward, 1992).

Changes in the labor market over the last three decades which resulted in stagnant wage growth affected mainly young workers (age 18 to 29) (Schmitt, 2008).

2

Background - 2 • The period investigated, 1990-2015, is characterized by global and local changes which have led to the abandonment of traditional labor relations; a decline in the power of trade unions and the introduction of flexible work arrangements.

• Changes in Israel: (1) A massive immigration flow from the Soviet Union in the early 90’s; (2) Increased exposure to import sharpened the differences between industries; (3) Increase in the share of foreign workers starting in 2000; (4) Sharp decline in the percentage of unionized employees; (5) Increase in employment rate and decrease in unemployment, especially in recent years.

(6) In 2011 – social protest where more the 400000 mainly young people demanded “social justice”.

3

Purpose and research questions The purpose of the study is to examine and compare the wage trajectories of young people who entered the labor market at different points in time during the period 1990-2015.

We examine their absolute and relative mobility and how they changed during their first years of employment.

We also compare the wage trajectories of different social groups of young people, in order to understand whether changes in opportunities and working conditions contributed to increased inequality and polarization in the Israeli society.

4

Database • The research is based on administrative data obtained from the tax authorities, which includes the population of all employees in Israel.

• This dataset includes information on wages as well as sociodemographic characteristics of individuals.

• The study follows the wages of young people aged 24-30 who entered the labor market at various points in time: 1991, 1995, 2000, 2005, and 2010, and examines their absolute and relative earnings mobility over the years.

• All wage data were adjusted to 2015 prices.

• Extreme values (too low and too high wages) were excluded.

5

Method • Examination of absolute mobility – following the real wage trajectories in five-year intervals for various time points (91,95,00,05,10).

Relative mobility: • Following the change in deciles which were calculated on the basis of the entire wage distribution (for all employees) each year.

• Aggregative mobility indices (correlation coefficients, Shorrocks index).

• Logistic Regressions to examine the probability of upward mobility.

6

Absolute Mobility

7

, Monthly wage by gender, 2015 prices years intervals 5 after 20 after 20 after 15 after 10 after 5 years years years years years

wage at entry point Men

15,543

13,800

12,786

11,962

7,983

4,328entered LM in 1991

15,060

13,140

11,624

10,101

4,861entered LM in 1995

13,811

11,327

8,799

5,383entered LM in 2000

12,337

8,886

4,736entered LM in 2005

9,984

4,629entered LM in 2010 Women

10,049

8

8,548

7,385

6,906

5,052

2,838entered LM in 1991

8,902

7,436

6,298

5,745

2,953entered LM in 1995

8,099

6,672

5,327

3,431entered LM in 2000

7,450

5,853

3,339entered LM in 2005

6,429

3,275entered LM in 2010

• The real wage level gradually declined from 2000 to 2010. • The gender gap is consistent.

Initial Monthly Wage of Young Adults as a Percentage of GDP per capita 6.0%

5.6% 5.1%

5.0%

70%

Men 5.0%

Women 4.4%

4.0%

3.7%

3.1%

3.0%

3.2%

3.5%

2.0%

9

57% 50%

50%

49%

47%

45%

2005

2010

30% 20%

1.0%

• •

60%

40%

3.1% 2.5%

0.0%

Initial Monthly Wage of Young Adults as a Percentage of the Average Wage

10%

1991

1995

2000

2005

2010

0%

The initial wages declined consistently The situation of young people has deteriorated over time.

1991

1995

2000

Wage Increase rate after 5 working years 140%

Women 120%

95%

100%

80%

60%

78%

116%

Men

108%

88%

84%

96%

75% 55%

63%

40%

20%

0%

• 10

Not only that women's wages are lower, their wage growth rate is also lower than that of men.

Monthly real wage - changes by gender groups, 5 years intervals Change rate (commulative) after 25 after 20 after 15 after 10 after 5 wage in years years years years years entry point Men 259% 219% 195% 176% 84% 4,328 entered to LM in 1991 210%

254%

170%

139%

108%

4,861

entered to LM in 1995

157%

110%

63%

5,383

entered to LM in 2000

161%

88%

4,736

entered to LM in 2005

116%

4,629

entered to LM in 2010

201%

160%

143%

78%

2,838

Women entered to LM in 1991

201%

152%

113%

95%

2,953

entered to LM in 1995

136%

94%

55%

3,431

entered to LM in 2000

123%

75% 96%

3,339 3,275

entered to LM in 2005 entered to LM in 2010

*Lower mobility among women in all intervals. *The high mobility after 5 years in 1995-2000 is understandable (increase in GDP in 2000). But how can we explain the high mobility in 2010-2015? Could this be an impact of the protest?

11

Wage growth of young workers entering the labor market in 1991, by population groups Men Haredim

Women

Jewish

Arab

Ethiopian imm.

Immigrants

16000.0 14000.0 12000.0 10000.0 8000.0 6000.0 4000.0 2000.0 1991 12

1995

2000

2005

2010

2015

Wage growth – Male vs. Female – all periods

16000.0 10000.0 9000.0

13000.0

8000.0 7000.0 10000.0 6000.0 5000.0 4000.0 3000.0

Women ente red to LM i n 1995

13

7000.0

Men entered to LM in 1995 Men entered to LM in 2000

Women ente red to LM i n 2000

Men entered to LM in 2005

Women ente red to LM i n 2005 Women ente red to LM i n 2010

2000.0

Men entered to LM in 1991

Women ente red to LM i n 1991

Men entered to LM in 2010 4000.0

Arab vs. Jewish

9000.0

14000.0

8000.0

12000.0

7000.0

10000.0

6000.0 5000.0

8000.0

4000.0

6000.0

3000.0 2000.0 1000.0 0.0

14

Arab entered to LM in 1991 Arab entered to LM in 1995 Arab entered to LM in 2000 Arab entered to LM in 2005 Arab entered to LM in 2010

Jewis h entered to LM in 1991

4000.0

Jewis h entered to LM in 1995 Jewis h entered to LM in 2000

2000.0

Jewis h entered to LM in 2005 Jewis h entered to LM in 2010

0.0

Immigrants 16000.0 14000.0 12000.0 10000.0 8000.0 6000.0 4000.0 2000.0 0.0 Basic wage

Immigrants entered to LM in 1991

Immigrants entered to LM in 1995

Immigrants entered to LM in 2000

Immigrants entered to LM in 2005

Immigrants entered to LM in 2010

after 5 years

after 10 years

after 15 years

after 20 years

after 25 years

Immigrants have nice wage growth in all periods as well as growth in the initial wage over time 15

Hi-Tech vs, Food and Trade 12000.0

20000.0 18000.0

10000.0

16000.0 14000.0

8000.0

12000.0 6000.0

10000.0 8000.0

4000.0

Trade & Food entered to LM in 1991

6000.0

Hi-Tech entered to LM in 1995

Trade & Food entered to LM in 2000

4000.0

Hi-Tech entered to LM in 2000

Trade & Food entered to LM in 2005

2000.0

Trade & Food entered to LM in 1995 2000.0

Trade & Food entered to LM in 2010 0.0

Hi-Tech entered to LM in 1991

Hi-Tech entered to LM in 2005 Hi-Tech entered to LM in 2010

0.0

HT - Initial wage generally increases. T&F – wages remained in the same level. Slopes are significantly steeper in HT. 16

Public Administration 16000.0 14000.0 12000.0 10000.0 8000.0 6000.0

Public admin entered to LM in 1991 Public admin entered to LM in 1995 Public admin entered to LM in 2000 Public admin entered to LM in 2005 Public admin entered to LM in 2010

4000.0 2000.0 0.0

Basic wage

after 5 years

after 10 years

after 15 years

after 20 years

after 25 years

• An example of the development of wages of unionized workers, which is different from what we have seen so far: more arranged; the lines don’t cross each other; differences in the initial wage in different starting points and relatively consistent growth.

17

Impact of education on mobility Started in 2010

Started in 2005

Started in 2000

Not Not Not educated educated educated educated educated educated 4,515

4,195

5,138

4,899

6,895

5,606Entry wage

153%

71%

117%

42%

84%

39%Wage after 5 years

199%

71%

132%

49%Wage after 10 years

170%

67%Wage after 15 years

*At least 3 academic years

18

As expected, the wage growth of educated is much more rapid than that of the non educated at every point in time.

But over time the returns for education decreased: the gap between educated and non educated in 2000 at the starting point is 23% compared to 7% in 2010 (tradable vs. non tradable sectors?).

Relative Mobility Comparison of 4 decades: 1991-2000 1995-2005 2000-2010 2005-2015

19

Correlation coefficients for the deciles between the entry point and 10 years later 0.400

0.350

0.300

0.250

Pearson Spearman

0.200

0.150

0.100

0.050

0.000 1991-2000

1995-2005

2000-2010

2005-2015

Gini coefficient and Shorrocks mobility index at each entry point – after 10 years 2005-2015

21

2000-2010

1995-2005

1991-2000

Index

0.329

0.336

0.338

0.308 Gini coef (average wage)

0.114

0.104

0.100

0.114 Shorroks index M

0.320

0.321

0.319

0.291 Gini coef (average wage)

0.118

0.111

0.115

0.126 Shorroks index M

0.300

0.295

0.297

0.280 Gini coef (average wage)

0.135

0.134

0.123

0.135 Shorroks index M

0.330

0.339

0.338

0.310 Gini coef (average wage)

0.113

0.103

0.100

0.113 Shorroks index M

0.254

0.235

0.271

0.232 Gini coef (average wage)

0.156

0.164

0.133

0.170 Shorroks index M

0.329

0.336

0.338

0.308 When 1991-2000=100.0:

0.114

0.104

0.100

0.114 Gini coef (average wage)

0.320

0.321

0.319

0.291 Shorroks index M

Pop. Group All Men Women Arab Jewish

All

Relative wage mobility (deciles) of young workers entering the labor market in 1991, by population groups Men Immigrants 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 1991 • 22

•

1995

Women Haredim

2000

Jewish

Arab

2005

Ethiopian imm.

2010

2015

Main relative mobility occurred in the first 5 years and almost stopped growing after 10 years; There is a floor effect with the exception of Arabs (their slope is the flattest)

Starting deciles and increments after 10 years, pop. subgroups Entered 1991 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

4.4

3.0

3.7

3.7

Entered 1995

3.0

3.5

3.4

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

4.6

3.0

Entered 2000 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

23

4.4

3.1

3.8

3.4

2.6

3.8

3.6

2.7

3.5

3.3

Entered 2005

3.5

3.3

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

4.1

3.1

3.7

3.3

3.0

3.6

3.0

Low wage (Women, Haredim, Ethiopian) have generally higher relative mobility, except for the Arabs.

Arabs ended up at the lowest deciles in all decades.

Men enter and ended up with the highest deciles.

Immigrants enter in moderate levels but have the highest mobility (especially in the first decade).

Starting deciles and increments after 10 years, industries Entered 1991 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

4.2

3.5

4.6

3.4

3.8

Entered 1995

4.0

3.8

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

5.2

3.4

Entered 2000 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

6.1

4.3

4.3

2.7

3.8

4.5

3.3

3.7

3.8

3.5

Entered 2005

3.7

3.7

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

5.5 3.1

3.7

3.3

3.0

3.6

3.0

High-tech: in the first period -- medium initial decile level but high mobility. Later on - the highest decile but moderate relative mobility. Overall- - at the end they reach the same highest location (7-8) after a decade.

Health&educ branch – except for 1995 – low initial decile with relatively high mobility (outsourcing? Fellows?)

24

Results from logistic regression for upward relative mobility after 10 years, Odds ratios

2005 0.62*** 0.57*** 0.54*** 0.98 0.93 0.63*** 1.04

2000 0.60*** 0.48*** 0.50*** 1.23*** 0.95 0.48*** 0.97

1995 0.59*** 0.27*** 0.47*** 0.89** 0.61** 0.43*** 0.77***

Entered in: Basic decile Arab Women Immigrants Ethiopia Immig.

Ultra-Orthodox Married

0.99 3.05*** 1.76*** 1.32* 1.40*** 1.60*** 0.83** 1.06***

Economic branches (base: other services) Traditional Hi-Tech Public admin.

Electricity Health & Education Finance & Proffesional services Trade & Food services number of months employed

0.97 1.53*** 1.05 1.09 1.06 1.17** 0.84*** 1.04***

0.81*** 1.15 1.64*** 1.53 0.88 0.97 0.74*** 1.05***

2.41***

2.50***

up to 3 academin years

3.77*** 0.31***

3.37*** 0.16***

0.22***

above 3 acedemic years 0.22*** _cons

21161

18001

16363

17

17 3559.26

15 3211.62

25 3744.08

0.78*** 1.72*** 1.23 0.79 1.33*** 1.23** 0.71*** 1.06***

1991 0.58*** 0.28*** 0.49*** 1.32*** 0.61** 0.42*** 0.80***

20847

N

15 df_m 4328.79 chi2

In the 1990s and in 2000’s, controlling for the initial decile, low-wage populations (women, Arabs UO’s Ethiopians in women - and trade &food branch) have also a lower probability of upward mobility. Employees in the high-tech industry: their advantages are less sharp than in the 1990s – could be because education controlled for.

Each work month raises the probability by 3% -6% for all groups.

The advantage of immigrants disappears in 2005.

Education - chances for upward mobility are much higher for those with up to 3 years of academic education years and even more for those with more than 3 academic years.

Results from logistic regression for upward relative mobility after 10 years, Arabs and Women, Odds ratios 2005 Women 0.62***

Arabs 0.58***

0.89

Women 0.60***

1991

Entered in:

Arabs 0.52***

Basic decile

0.32*** 0.68***

Arab 0.33***

Women

1.06

1.36***

Immigrants

0.78*

0.36**

Ethiopia Immig.

0.70*** 1.23***

0.46*** 0.81***

0.77**

Ultra-Orthodox Married

1.46* 5.41 4.91*** 0.94 6.14*** 2.60** 1.42 1.04***

Economic branches (base: other services) Traditional Hi-Tech Public admin.

Electricity Health & Education Finance & Proffesional services Trade & Food services number of months employed

0.89

0.89 1.31* 1.03 0.55 1.04 1.28** 0.85** 1.04***

0.99 7.27* 1.99** 2.83 1.48** 1.28 0.85 1.03***

2.33***

3.62***

up to 3 academin years

3.51*** 0.16***

4.05*** 0.75

above 3 acedemic years _cons

26

0.94 2.13*** 1.55*** 1.03 1.18 1.43*** 0.81* 1.06***

0.15***

0.66

Conclusions •

The worsening position of young people is in line with public feelings, and could partly explain the outbreak of the social protest in 2011.

•

This trend has indeed begun to be somewhat restrained during the period 2010-2015.

•

Overall, the findings suggest a growing polarization consistent with the skillbias of the labor market with higher wages and faster wage growth (albeit moderated in the 2000’s) among the highly educated and especially in high tech industries.

•

These trends could explain in part the growing gender differences in mobility patterns (as women are less likely to enter technical occupations).

•

The skill divide might also explain the growing inequality between Jews and Arabs, as Arabs’ educational attainment is far below that of Jews.

27

Thank you!

28

Main indicators

29

2010

2005

2000

1995

1991

Indicator / Year

3.7

2.3

6.0

3.7

-0.1

Growth rate

71.8

68.3

67.6

68.3

63.6

Employees rate 25-64

7.2

9.6

9.1

6.2

10.7

Unemployment rate

20,000

23,000

61,000

77,000

175,000

Num. of Immigrants

6,334

Real Avg. wage (2015 prices)

8,800

8,771

8,957

7,736

43,000

36,000

30,000

17,000

12,000 Recipients of a first degree

22,000

22,000

19,000

13,000

10,473

From Universities

Deciles / pop. Group and industries 1991-2015 entered to LM in 1991 Men Women Jewish Arabs Ethiop immigrants Immigrants Ultra-orthodox Traditional branches Hi-Tech Public admin.

Electricity&water Health&Educ Finanace*proffesional Trade&food services Other services

1995 6.4 4.7 5.7 4.8 4.7 5.8 4.8 5.7 7.4 5.8 6.5 5.4 5.6 5.4 5.3

2000 7.0 5.3 6.3 4.9 5.3 6.5 5.3 5.8 8.0 6.4 7.0 6.3 6.8 6.0 6.2

2005 7.1 5.5 6.5 5.0 5.7 6.7 5.4 5.9 8.0 6.8 7.3 6.6 7.0 6.3 6.3

2010 7.2 5.9 6.7 5.1 6.0 7.0 5.5 6.1 8.0 7.1 7.5 7.0 7.2 6.4 6.6

2015 7.2 6.1 6.8 5.3 6.3 7.1 5.6 6.3 7.9 7.3 7.5 7.3 7.2 6.4 6.6

1995

2000

2005

2010

2015

Men

4.6

6.5

6.9

7.1

7.2

Women

3.0

4.6

4.9

5.4

5.7

Jewish

3.8

5.7

6.0

6.3

6.5

Arabs

3.6

4.5

4.7

5.0

5.2

Ethiop immigrants

2.7

4.1

4.6

5.3

5.7

Immigrants

3.5

5.5

5.9

6.3

6.5

Ultra-orthodox

3.3

4.8

5.2

5.3

5.5

Traditional branches

4.0

5.4

5.7

6.0

6.2

entered to LM in 1995

1991 4.4 3.0 3.7 3.7 3.0 3.5 3.4 3.7 4.2 3.5 4.6 3.4 3.8 4.0 3.8

Hi-Tech

5.2

7.2

7.3

7.4

7.6

Public admin.

3.4

5.5

5.9

6.3

6.7

Electricity&water

4.5

5.7

6.0

6.2

6.2

Health&Educ

3.3

5.6

6.0

6.4

6.7

30

entered to LM in 2005

2005

2010

2015

Men

4.1

6.1

6.7

Women

3.1

4.7

5.2

Jewish

3.7

5.6

6.2

Arabs

3.3

4.4

4.8

Ethiop immigrants

3.0

4.7

5.2

Immigrants

3.6

5.5

6.1

Ultra-orthodox

3.0

4.4

5.0

Traditional branches

3.8

5.3

5.8

Hi-Tech

5.5

7.4

7.9

Public admin.

5.7

6.4

7.2

Electricity&water

4.9

6.1

6.6

Health&Educ

2.6

4.9

5.6

Finanace*proffesio nal

3.9

6.0

6.6

Trade&food services

3.5

5.2

5.7

Other services

3.4

5.4

5.9

2010

2015

Men

3.8

6.1

Women

2.9

4.7

entered to LM in 2010

Summary • Young people who enter the labor market are worse off in the 2000’s compared with those entering in the 1990s, in terms of their initial wages.

• Aggregate indices of relative mobility (of 10 years) show that alongside the rise in inequality during the period 1990-2005, there was also a decline in mobility. This changed only during the last decade, 2005-2015.

• Mobility of highly educated young people is much higher that that of the low. However returns for education in term of initial wage declined over time (this was also one of the claims in the 2011 protest).

• Different groups of young people have different wage trajectories: • (1) Despite a slight increase in the salaries of young women, their wages at entrance point in 2010 is still 30% lower than that of young men, and their wage mobility is lower.

• (2) In certain industries there are small returns to seniority or work stability. The high-tech industry has very high mobility during the first 10 years, but afterwards we observe a decrease in mobility compared to other industries.

31