Causes of German Income Inequality across Time and Space

Causes of German Income Inequality across Time and Space Franziska K. Deutschmann Graduate School of Decision Sciences, University of Konstanz February 29, 2016 Abstract Similar to most industrialized countries, income inequality has increased in Germany since the 1970s. I study in a non-parametric fashion how demography, education, and employment trigger income inequality across households through their impact on disposable income. West German inequality from 2011 is compared to the 1970s. A comparison to 2011 East Germany is added because demography, education, and employment still differ substantially between East and West Germany. For the comparison across time, I find that the prevalence of singlehood, the increase in education as well as the change in employment disequalize income across households. Compared to East Germany, the low West German unemployment rate has a strong equalizing effect which is partly offset by more diversity in formal education. Significant effects of educational homogamy in marriages or the aging society cannot be found. JEL classification: D31, E25, I24, J11, J12, J21. Keywords: inequality, demography, household structure, assortative mating, education, employment, Germany Contact Inforamtion: franziska.deutschmann@uni-konstanz.de 1

Extended Abstract Introduction and Research Question Similar to most industrialized countries (Gottschalk & Smeeding, 2000, OECD, 2008, 2011), German income inequality has increased since the 1970s. 1,2 In particular, when utilizing the Gini coefficient for measuring income inequality of the working age households, inequality has increased from 0.230 in 1976 to 0.301 in 2011. Prior studies found that changes in employment (Biewen & Juhasz, 2011, Peichl, Pestel & Schneider, 2012), labor income, the tax system, and in the transfer system(biewen& Juhasz, 2012), capital income(rehm, Schmid& Wang, 2014, Schmid & Stein, 2013), as well as changes in the household structure (Biewen & Juhasz, 2012, Peichl, Pestel & Schneider, 2011, Peichl et al., 2012), and household characteristics (Biewen & Juhasz, 2012) are associated with the rise in German income inequality. Recently, Pestel (forthcoming) found that marital sorting increases earnings inequality. Faik (2012) forecasts that the aging society will have a strong disequalizing impact on inequality. 3 The first aim of the paper is to better understand which demographic factors are associated with the rise in income inequality across West German households since the 1970s. The effect of an older working age population, the prevalence of singlehood, particularly the prevalence of single parents, and the effect of educational homogamy in marriages are of interest. When considering educational homogamy, the pure effect of an increase in positive assortative mating needs to be separated from the effect of the rise in (female) education (see Grave & Schmidt, 2012). 4 I isolate the change in assortative mating from the increase in male and female education by a procedure described by Sinkhorn & Knopp (1967). Assuming that higher female education and marital sorting have a indirect effects on females labor supply, the overall effects of marital sorting and (female) education through their changes in female labor supply are additionally taken into account. Greenwood, Guner, Kocharkov & Santos (2014) have shown that whether marital sorting affects inequality depends on the females labor supply. Therefore, I also compare the effect of gender-specific changes in labor supply to the effects of demographic and educational changes. Second, besides the comparison of West German inequality in time, one can look at inequality differences across space. In particular, the German reunification in the early 1990s gave rise to a country which is composed of two regions which differ substantially in their demography. Although facing the same political system, East and West Germany differ in their age structure (Grünheid, 2009), family development (Schneider, Naderi & Ruppenthal, 2012) including the degree of positive assortative mating (Grave & Schmidt, 2012), the prevalence of single and single parent households (Ebert & Fuchs, 2012, Grünheid, 2009), as well as employment (Ebert & Fuchs, 2012) and education (Grave & Schmidt, 2012). However, income inequality across households is similar in both regions in 2011: The Gini coefficient is 0.301, 0.300 1 For details see (Bach, Adam, Niehues, Schröder, Frey, Schaltegger, Berthold & Gründler, 2014, Becker, 2012, Biewen, 2000, Hauser & Becker, 1998) 2 Here, and in the following, Germany refers to West Germany before 1990 and to the reunified Germany since 1990. 3 For a detailed overview of economic inequality and how it is linked to political systems, education, demography, and health, see Salverda, Nolan & Smeeding (2011). 4 Controlling for changes in education is necessary, as an increase in female or male education may drive the growing likelihood that males and females with the same educational level marry each other. Suppose, for instance, that female education is increasing in such a way that the fraction of high, middle, and low educated females becomes as large as the fraction of high, middle, and low educated males, respectively. Then, educational homogamy rises by definition. 2

and 0.301 in West, East and overall Germany, respectively. 5 Why is West and East German income inequality similar although the demographic, educational and employment situation vary essentially? What forces account for the observed inequality in each region? Prior research suggested, that the household structure, the labor force participation, as well as the German tax and transfer system play a major role for East and West German income inequality. So did Peichl et al. (2012) find that the declining household size and the employment behavior strengthened income inequality growth in the 1990s and early 2000s. In particular for the beginning of the 1990s, Biewen (2001) relates the diminishing female labor market participation and rising unemployment to the East German inequality increase. For both East and West Germany, Peichl et al. (2012) and Fuchs-Schündeln, Krueger & Sommer (2010) provide evidence that the welfare state moderates earnings inequality. The second part of the paper focuses on how the region specific demographic, educational and labor market circumstances affect East and West German income inequality. Questions like, how would income inequality differ if female labor force participation in West Germany is as high as in East Germany, will be answered. Method Utilizing the rich household data of the German microcensus, allows me to apply a purely non-parametric version of the reweighting procedure of Di Nardo, Fortin & Lemieux (1996). Household types are subdivided by age groups, the cohabitation status (i. e. single or couple), and the number of children, the adults gender, their educational degree (i. e. university, vocational training, or school) and their working status (i. e. full-time, part-time or not working). The rich microcensus data allows for a finer subdivision in comparison to the previous studies (e.g. Biewen & Juhasz, 2012, Peichl et al., 2012). I construct counterfactual income distributions by adjusting the composition of the population with respect to a specific attribute like age, cohabitation or education to the composition of another year or region. For instance, suppose I would like to derive a counterfactual income distribution for West Germany in 2011 if educational levels stayed unchanged since the 1970s. The fraction of low educated females would increase in the counterfactual case. While adjusting the educational levels, the population is fixed with respect to other attributes. That is, the overall age distribution and the fraction of female and male single households are held constant at 2011 levels. Additionally, I assume that in 2011 lower educated females make different labor supply and fertility decisions than lower educated females in the 1970s. Therefore, the working status and the number of children are conditioned on education and are set to 2011 levels. By this approach, I incorporate indirect effects of a change in education on the income distribution through the number of kids and the working status. Another advantage of the non-parametric framework is that any inequality measure can be used. In this study, the Gini coefficient, the Theil index as well as Atkinson indexes are utilized. 5 Here, the Gini coefficient measures inequality between households with a household head of working age, i. e. 25-59 years old. 3

References Results For the West German comparison, I find that the prevalence of singlehood, the increase in education as well as the change in employment significantly disequalize income across households. Incorporating the indirect effect of singlehood on the number of children shows that the prevalence of singlehood drives the effect the increased fraction of single parents have. Assuming in 2011 the same fraction of single males and single females in the population as in the 1970s decreases the Gini coefficient significantly from 0.301 to 0.285, i. e. by 1.6 Gini points. Following the interpretation of Blackburn (1989), this decline in inequality is equivalent to a transfer of 3.2 percent of the mean household income (i. e. EUR 58.03) from each above median income household to each household below median income. Furthermore adjusting the number of children to the 1970s level strengthens the effect only slightly. Inequality would decrease additionally by roughly half a Gini point. If in 2011 the households workings status (formal education) had been that of 1976, income inequality would be lower by 1.03 (1.46) Gini points. High unemployment in East Germany has a strong influence. Its disequalizing impact on income inequality is partly offset by a less diverse distribution in formal education. Demographic aspects such as the older age of the East German population and the stronger prevalence of singlehood do not alter inequality significantly. The larger fraction of single parent households has only a small statistically significant effect of around 0.7 Gini points. Assuming the same education levels in East Germany as in West Germany, increases income inequality significantly by 1.5 Gini points whereas the counterfactual West German working status has a strong equalizing effect. If East German unemployment is as low as in West Germany and married females work less often full-time, inequality would strongly decrease by 2.0 Gini points. An increase in female education affected income inequality neither in the across time nor the across space comparison. Moreover, I could not confirm the hypothesis that marital sorting in terms of education augments income inequality. At first appearance, this finding seems to contradict the results of Pestel (forthcoming). He finds that marital sorting in terms of earnings potentials, i. e. wages, affects earnings inequality substantially in East Germany. However, the results of Pestel refer to pre tax inequality so that firstly the substantial effects could be diminished by the equalizing German tax and transfer system (see Fuchs-Schündeln et al., 2010). Secondly, Pestel considers marital sorting in terms of potential earnings. As formal education is only one aspect of potential earnings, my findings suggest that the impact of marital sorting on inequality is driven by something else than educational homogamy. References Bach, S., Adam, H., Niehues, J., Schröder, C., Frey, C., Schaltegger, C. A., Berthold, N., & Gründler, K. (2014). Einkommens- und Vermögensverteilung zu ungleich? Wirtschaftsdienst, 94(10), 691 712. Becker, I. (2012). Personelle Einkommensverteilung. In P. Bartelheimer, S. Fromm, & J. Kädtler (Eds.), Berichterstattung zur sozioökonomischen Entwicklung in Deutschland (pp. 597 632). Wiesbaden: VS Verlag für Sozialwissenschaften. Biewen, M. (2000). Income Inequality in Germany During the 1980s and 1990s. Review of Income and Wealth, 46(1), 1 19. 4

References Biewen, M. (2001). Measuring the Effects of Socio-Economic Variables on the Income Distribution: An Application to the East German Transition Process. Review of Economics and Statistics, 83(1), 185 190. Biewen, M. & Juhasz, A. (2011). Can Employment Changes Explain Rising Income Inequality in Germany? Schmollers Jahrbuch, 131(2), 349 357. Biewen, M. & Juhasz, A. (2012). Understanding Rising Income Inequality in Germany, 1999/2000-2005/2006. Review of Income and Wealth, 58(4), 622 647. Di Nardo, J., Fortin, N. M., & Lemieux, T. (1996). Labor Market Institutions and the Distribution of Wages, 1973-1992: A Semiparametric Approach. Econometrica, 64(5), 1001 1044. Ebert, A. & Fuchs, T. (2012). Haushalt, Familie und soziale Nahbeziehungen. In P. Bartelheimer, S. Fromm, & J. Kädtler (Eds.), Berichterstattung zur sozioökonomischen Entwicklung in Deutschland: Teilhabe im Umbruch Zweiter Bericht (pp. 565 595). Wiesbaden: VS Verlag für Sozialwissenschaften. Faik, J. (2012). Income Inequality and Poverty in Front of and during the Economic Crisis: An Empirical Investigation for Germany 2002-2010. Fuchs-Schündeln, N., Krueger, D., & Sommer, M. (2010). Inequality trends for Germany in the last two decades: A tale of two countries. Review of Economic Dynamics, 13(1), 103 132. Gottschalk, P. & Smeeding, T. M. (2000). Empirical evidence on income inequality in industrialized countries. In A. B. Atkinson & F. Bourguignon (Eds.), Handbook of Income Distribution, volume 1 (pp. 261 307). Elsevier. Grave, B. S. & Schmidt, C. M. (2012). The Dynamics of Assortative Mating in Germany. Ruhr Economic Papers. Essen: Rheinisch-Westfälisches Institut für Wirtschaftsforschung. Greenwood, J., Guner, N., Kocharkov, G., & Santos, C. (2014). Marry Your Like: Assortative Mating and Income Inequality. American Economic Review, 104(5), 348 353. Grünheid, E.(2009). ÜberblicküberdiedemografischeEntwicklunginWest-undOstdeutschland von 1990 bis 2004. In I. Cassens, M. Luy, & R. Scholz (Eds.), Die Bevölkerung in Ostund Westdeutschland (pp. 12 47). Wiesbaden: VS Verlag für Sozialwissenschaften. Hauser, R. & Becker, I. (1998). Die langfristige Entwicklung der personellen Einkommensverteilung in der Bundesrepublik Deutschland. Empirische Forschung und wirtschaftspolitische Beratung : Festschrift für Hans-Jürgen Krupp zum 65. Geburtstag, 119 134. OECD (2008). Growing unequal? Income distribution and poverty in OECD countries. Paris: OECD, Organisation for Economic Cooperation and Development. OECD (2011). Divided we stand: Why inequality keeps rising. Paris: OECD. Peichl, A., Pestel, N., & Schneider, H. (2011). Mehr Ungleichheit durch kleinere Haushalte? Der Zusammenhang zwischen Veränderungen der Haushaltsstruktur und der Einkommensverteilung in Deutschland. Zeitschrift für Arbeitsmarktforschung, 43(4), 327 338. 5

References Peichl, A., Pestel, N., & Schneider, H. (2012). Does size matter? the impact of changes in household structure on income distribution in germany. Review of Income and Wealth, 58(1), 118 141. Pestel, N. (forthcoming, 2016). Marital Sorting, Inequality and the Role of Female Labor Supply: Evidence from East and West Germany. Economica. Rehm, M., Schmid, K. D., & Wang, D. (2014). Why Has Inequality in Germany Not Risen Further after 2005?, volume 690 of SOEPpapers on Multidisciplinary Panel Data Research at DIW Berlin. Berlin: DIW Berlin. Salverda, W., Nolan, B., & Smeeding, T. M. (2011). The Oxford Handbook of Economic Inequality. OUP Oxford. Schmid, K. D. & Stein, U. (2013). Explaining Rising Income Inequality in Germany, 1991-2010. Schneider, N. F., Naderi, R., & Ruppenthal, S. (2012). Familie in Deutschland nach dem gesellschaftlichen Umbruch. In J. Huinink, M. Kreyenfeld, & H. Trappe (Eds.), Familie und Partnerschaft in Ost- und Westdeutschland, volume 9 of Zeitschrift für Familienforschung : Sonderheft (pp. 29 53). Opladen [u.a.]: Budrich. Sinkhorn, R. & Knopp, P. (1967). Concerning nonnegative matrices and doubly stochastic matrices. Pacific Journal of Mathematics, 21(2), 343 348. 6