Why has Income Inequality in Germany Increased From 2002 to 2011? A Behavioral Microsimulation Decomposition

• Published date: 01 September 2019
• Author: Robin Jessen
• Date: 01 September 2019
• DOI: http://doi.org/10.1111/roiw.12397

© 2018 Internationa l Association for Res earch in Incom e and Wealth

540

WHY HAS INCOME INEQUALITY IN GERMANY I NCREASED FROM

2002 TO 2011? A BEHAVIORAL MICROSIMULATION DECOMPOSITION

by Robin Jessen*

RWI, Freie Uni versität Berlin

This paper proposes a method to decompose changes in income inequality into the contributions of

policy changes, wage rate changes, and population changes while considering labor supply reactions.

Using data from the Socio-Economic Panel (SOEP), this method is applied to decompose the increase

in income inequality in Germany from 2002 to 2011, a period that saw tax reductions and a controver-

sial overhaul of the transfer system. The simulations show that tax and transfer reforms have had an

inequality-reducing effect as measured by the mean log deviation and the Gini coefficient. For the Gini,

these effects are offset by labor supply reactions. In contrast, policy changes explain part of the increase

in the ratio between the 90th and the 50th income percentiles. Changes in wage rates have led to a

decrease in income inequality. Thus, the increase in inequality was due to changes in the population.

JEL Codes: D31, I38, J31

Keywords: inequality, decomposition, labor supply, microsimulation, policy reform

1. intRoduction

Income inequality increased considerably in Germany from 2002 to 2011. The Gini

coefficient of equivalized net household income increased from 28.5 to 29.5 (own cal-

culation). From a policy perspective, it is important to learn about the determinants of

increasing income inequality, in order to take appropriate countermeasures; for exam-

ple, if policy reforms have had an inadvertent inequality-increasing effect. The time span

from 2002 to 2011 is particularly interesting regarding the interaction of inequality and

tax and transfer policy, as it witnessed a strong increase in inequality as well as major

reforms to the tax and transfer system: the controversial Hartz IV reforms of the transfer

system, as well as part of the phasing in of major tax reforms, started in 2001. Increasing

wage dispersion is another potential explanation for the increase in income inequality.

These potential factors in increasing inequality are described in detail in Section 3.

Correspondence to: Robin Jessen, RWI, Invalidenstrasse 112, 10115 Berlin, Germany (robin.

jessen@rwi-essen.de).

Review of Inc ome and Wealth

Series 65, Numb er 3, September 2019

DOI : 10.1111 /roi w.123 97

Note: The author thanks Olivier Bargain, Martin Biewen, Giacomo Corneo, Jonas Jessen, Johannes

König, Andreas Peichl, Davud Rostam-Afschar, Sebastian Schmitz, Viktor Steiner, Maximilian

Stockhausen, two anonymous referees, seminar participants at Freie Universität Berlin, the Institute for

Fiscal Studies in London, the annual meeting of the Society of Labor Economists in Seattle, the SOEP

User Conference and the conference Crises and the Distribution in Berlin, the Tagung des Vereins für

Socialpolitik in Hamburg, the Annual Meeting of the Austrian Economic Association in Vienna, and the

editor, Conchita D’Ambrosio, for helpful comments and suggestions. The author is grateful for financial

support from the Hans-Böckler-Stiftung. The data used in this paper were made available by the German

Socio-Economic Panel Study (SOEP) at the German Institute for Economic Research (DIW), Berlin.

bs_bs_banner

Review of Income and Wealth, Series 65, Number 3, September 2019

541

© 2018 Internationa l Association for Res earch in Incom e and Wealth

The aim of this study is to quantify the impact of policy reforms, changes

in conditional wage rates, and remaining changes of the population on income

inequality. To allow for the joint analysis of these factors, the decomposition

framework by Bargain (2012a, b) is extended to explicitly account for the effect

of changes in conditional wage rates in the spirit of Bourguignon et al. (2008).

The decomposition method combines microsimulation, a structural labor supply

model, and the construction of counterfactual wage rates using Mincer-style wage

regressions. The decomposition is done in an entirely disaggregated way that is not

limited to a specific class of inequality indices. It allows for the graphical repre-

sentation of counterfactual distributions. Marginal effects of particular factors on

inequality are calculated by comparing actual and counterfactual distributions and

thus can be interpreted as ceteris paribus changes unconfounded by demographic

or business cycle changes. The decomposition method is explained in Section 3.

This study contributes to the literature on the decomposition of differences

between two income distributions and in particular to the literature using microsimu-

lation techniques. Bargain and Callan (2010), Bargain (2012b), Liégeois and Dekkers

(2014), and Bargain et al. (2017) simulate counterfactual net incomes by applying the

tax and transfer system of a given period to the population of another period using a

detailed tax and transfer calculator to obtain intermediate distributions. Creedy and

Herault (2011) and Bargain (2012a) expand the microsimulation approach by simulat-

ing counterfactual labor supply decisions. Bargain et al. (2015) simulate responses of

taxable income. Herault and Azpitarte (2016) allow for the simulation of a wide range

of additional determinants. The study at hand combines the simulation of counterfac-

tual labor supply with the prediction of counterfactual wages following Bourguignon

et al. (2008), akin to the decomposition method introduced by Blinder (1973) and

Oaxaca (1973). As pointed out by Bourguignon et al. (2008), the combination of

strictly parametric techniques offers the advantage of a straightforward economic

interpretation (see also Brewer and Wren-Lewis 2015; Herault and Azpitarte 2016).

Apart from the method, the analysis conducted in this study adds to a devel-

oping literature on the causes of increases in income inequality in Germany in

recent years (Arntz et al. 2007; Biewen and Juhasz 2012; Peichl et al. 2012; Biewen

et al. 2016; Bargain et al. 2017). These studies are summarized in the next section.

The results are presented in Section 5. The decomposition shows that changes

of the tax and transfer system have slightly alleviated inequality as measured

through the Gini index and the mean log deviation (MLD). The negative effect of

policy changes on the Gini is offset by labour supply reactions. In contrast, policy

changes have led to an increase in the ratio between the 90th and the 50th income

percentiles (Q90/50). The overall effect of changes in wage rates on inequality is

found to be negative. Thus, the overall increase in income inequality was caused

by changes in characteristics of the population that are not explicitly modeled; for

example, in the household structure.

2. PRevious studies on GeRmany

A few papers deco mpose the overall chang e in income inequ ality in Ger many

between two periods into different factors. Table 1 summar izes the methods