Winners and Losers of the Income Dynamics in Germany

between 1998 and 2005

The Impact of Social Class on Inequality

Johannes Giesecke

Roland Verwiebe

University of Mannheim

University Hamburg

Structure of Presentation

1. Conceptional Background on Income Inequality, Hypothesis

2. Data Base and Methods

3. Empirical Results

4. Conclusion

I. Conceptional Background

• Main Assumption I: After a period of relatively high stability social inequality is rising in Germany

• Window of observation: 1998-2005, first social-democratic led government in Germany since the 1970’s– main political goals: modernization of welfare

state; reduction of labour costs, taxes, and bureaucracy for corporations; reduction of unemployment

… Conceptional Background

• Outcome of societal modernization: higher social inequality

• Underlying mode of rising overall social inequality: average real labour income (of dependently employed persons) is decreasing; profits and rents are increasing

• Main Assumption II: the dynamics of decreasing labour income and increasing profits/rents strengthens the importance of one’s individual class position

Hypothesis• Social class is an important explanatory factor for

the increase in the inequality of labour income. Specifically, higher social classes were overly able to increase their income in the recent past, which corresponds to lower classes having experienced a relative decrease of their incomes and wages.– Argument I: technological change and the resulting

demand for highly qualified labour increases income inequality (Atkinson 2000; DiPrete et al. 2002; Esping-Andersen 1993; Goldthorpe 2002)

– Argument II: based on efficiency-theory one could argue (e.g. Breen 1997) that the growing labour market risks are not equally distributed across social classes (re-commodification)

II. Data Structure, Methods• Data from the German Socio-Economic Panel (GSOEP),

longitudinal data structure 1998 – 2005• Two time slots considered: 1998-2001 and 2002-2005• Dependent variable: real (log) hourly wages• Results are based on growth curve models, specifying

random effects for both the intercept and the rate of growth

• Models allow to estimate average wage growth for different labour market groups as well as individual wage growth profiles

• Models focus on EGP-class and control for age, family size, citizenship, former unemployment experience, education and various labour market variables (e.g. firm- size, industry, type of contract, firm tenure)

• Separate analyses for men and women

III. Results: Distribution of Earnings for Men and Women between 1998-2005

01

00

02

00

03

00

04

00

05

00

0

Eu

ro

1998 1999 2000 2001 2002 2003 2004 2005

01

00

02

00

03

00

04

00

05

00

0

Eu

ro

1998 1999 2000 2001 2002 2003 2004 2005

01

00

02

00

03

00

04

00

05

00

0

Eu

ro

1998 1999 2000 2001 2002 2003 2004 2005

01

00

02

00

03

00

04

00

05

00

0

Eu

ro

1998 1999 2000 2001 2002 2003 2004 2005

Men West-Germany Women West-Germany

Men East-Germany Women East-Germany

Source: GSOEP 1998-2005

III. Results: Structures of Income Inequality in Germany (1998-2005)

Men Women 1998-2001 2002-2005 1998-2001 2002-2005 Reference: Semi/Unskilled Manual Higher Service Class 0.308*** 0.364*** 0.335*** 0.362***

Increase per year 0.0280*** 0.0249*** 0.0039 0.0108 Lower Service Class 0.198*** 0.255*** 0.236*** 0.243***

Increase per year 0.0178** 0.0158*** -0.001 0.0061 Routine Non-Manual 0.211*** 0.284*** 0.272*** 0.253***

Increase per year 0.0168 0.0208** -0.0042 0.0064 Routine Service-Sales 0.0730** 0.0413* 0.0368 0.0535**

Increase per year 0.0195** 0.0179** 0.0147 0.0062 Skilled Manual 0.0698*** 0.0631*** 0.0250 -0.0072

Increase per year 0.0078 0.0017 0.0023 0.0092 Farm Labour -0.0683 0.0696 -0.0428 0.0125

Increase per year 0.0270 -0.0041 0.0448 -0.0120 N 11297 14621 8320 12138 *p<0.05;**p<0.01;***p<0.01; Source: GSOEP 1998-2005, coefficients from growth-curve-models, dependent variable: log. hourly wages, various individual and labour market variables are controlled for

III. Results: Structures of Income Inequality in Germany (1998-2005)

Men Women 1998 2002 1998 2002 Higher Service Class 0.31 0.32 0.32 0.33 Lower Service Class 0.32 0.34 0.33 0.33 Routine Non-Manual 0.27 0.32 0.28 0.30 Routine Service-Sales 0.34 0.30 0.35 0.35 Skilled Manual 0.25 0.28 0.34 0.35 Semi/Unskilled Manual 0.29 0.30 0.31 0.35 Farm Labour 0.28 0.25 0.34 0.29 Source: GSOEP 1998-2005, class-specific standard deviations of the composite error (random intercept plus ideosyncratic error) at begin of each observational period, growth-curve-models, various individual and labour market variables are controlled for

IV. Conclusion• Societal modernisation of social-democratic led

government produced increase in labour income inequality between 1998 and 2005

• Social class is a major force driving this increase• Polarisation between classes is increasing: “winner”:

people from higher classes; ”loser”: lower classes• Heterogeneity within classes is by and large

increasing (for men more than for women)• As to whether these results can be generalized for

other European countries must be shown by future analyses