Workplace Heterogeneity and the Rise of West German Wage Inequality

TLDR: In this paper, the role of establishment-speci c wage premiums in generating recent increases in West German wage inequality was studied and it was shown that the increasing variability of West German wages has arisen from a combination of rising heterogeneity between workers, rising variability in the wage premiums at dierent establishments, and increasing assortativeness in the as-signment of workers to plants.

Abstract: We study the role of establishment-speci…c wage premiums in generating recent increases in West German wage inequality. Models with additive …xed eects for work- ers and establishments are …t in four sub-intervals spanning the period from 1985 to 2009. We show that these models provide a good approximation to the wage struc- ture and can explain nearly all of the dramatic rise in West German wage inequality. Our estimates suggest that the increasing variability of West German wages has arisen from a combination of rising heterogeneity between workers, rising variability in the wage premiums at dierent establishments, and increasing assortativeness in the as- signment of workers to plants. In contrast, the idiosyncratic job-match component of wage variation is small and stable over time. Decomposing changes in mean wages between dierent education groups, occupations, and industries, we …nd that increas- ing plant-level heterogeneity and rising assortativeness in the assignment of workers to establishments explain a large share of the rise in inequality along all three dimensions.

Figures

Figure 3b: Trends in Residual Wage Inequality

Figure 11: Decomposition of Variance of Log Wages

Figure 7: Relative Explanatory Power of AKM Model in Different Periods

Figure 12: Assignment of Workers to Establishments by Education Group: 1985-91 vs. 2002-09

Figure 13: Mundlak Decomposition of Return to Education

Table 1: Summary Statistics for Overall Sample and Individuals in Largest Connected Set

Full text

IAB Discussion Paper
Articles on labour market issues
26/2012
David Card
Jörg Heining
Patrick Kline
ISSN 2195-2663
Workplace heterogeneity and the rise
of West German wage inequality

Workplace Heterogeneity and the Rise of West
German Wage Inequality
David Card (University of California at Berkeley/NBER)
Jörg Heining (IAB)
Patrick Kline (University of California at Berkeley/NBER)
Mit der Reihe „IAB-Discussion Paper“ will das Forschungsinstitut der Bundesagentur für Arbeit den
Dialog mit der externen Wissenschaft intensivieren. Durch die rasche Verbreitung von Forschungs-
ergebnissen über das Internet soll noch vor Drucklegung Kritik angeregt und Qualität gesichert
werden.
The “IAB Discussion Paper” is published by the research institute of the German Federal Employ-
ment Agency in order to intensify the dialogue with the scientific community. The prompt publication
of the latest research results via the internet intends to stimulate criticism and to ensure research
quality at an early stage before printing.
IAB-Discussion Paper 26/2012
2

Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Zusammenfassung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Background - Macro Trends and Institutional Changes . . . . . . . . . . . . . . 8
3 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4 Overview of Trends in Wage Inequality . . . . . . . . . . . . . . . . . . . . . . 13
5 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
7 Decomposing Between-Group Wage Differentials . . . . . . . . . . . . . . . . 26
8 Rising Establishment Heterogeneity: Cohort Effects and Collective Bargaining
Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
9 Establishment Effects and the Survival of Establishments and Jobs . . . . . . . 32
10 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
IAB-Discussion Paper 26/2012
3

Abstract
We study the role of establishment-specific wage premiums in generating recent increases
in West German wage inequality. Models with additive fixed effects for workers and es-
tablishments are fit in four distinct time intervals spanning the period 1985-2009. Unlike
standard wage models, specifications with both worker and plant-level heterogeneity com-
ponents can explain the vast majority of the rise in wage inequality. Our estimates sug-
gest that the increasing variability of West German wages results from a combination of
rising heterogeneity between workers, rising variability in the wage premiums at different
establishments, and increasing assortativeness in the matching of workers to plants. We
use the models to decompose changes in wage gaps between different education levels,
occupations, and industries, and in all three cases find a growing contribution of plant
heterogeneity and rising assortativeness between workers and establishments.
Zusammenfassung
Wir untersuchen den Beitrag von betriebsspezifischen Lohnaufschlägen, die in jüngerer
Zeit zur Erhöhung der Lohnungleichheit in Westdeutschland beigetragen haben. Es wer-
den Modelle mit additiv fixen Effekten für Arbeitnehmer und Betriebe für vier verschiedene
Zeitintervalle über den Zeitraum 1985-2009 geschätzt. Anders als gewöhnliche Lohnmo-
delle, können Spezifikationen, welche sowohl Heterogenität bei den Arbeitnehmern als
auch auf Betriebsebene berücksichtigen, einen großen Teil des Anstiegs der Lohnungleich-
heit erklären. Unsere Schätzungen deuten auf eine zunehmende Variabilität der westdeut-
schen Löhne aufgrund einer Kombination aus steigender Heterogenität der Arbeitnehmer,
einer stärkeren Variabilität der Lohnaufschläge über die Betrieben und aus einem verbes-
serten Matching von Arbeitnehmern und Betrieben resultiert. Wir verwenden diese Mo-
delle, um die Veränderungen der Lohnlücken hinsichtlich verschiedener Bildungsstufen,
Berufe und Wirtschaftszweige zerlegen zu können. In allen drei Fällen finden wir eine zu-
nehmende Bedeutung der Betriebsheterogenität und ein verbessertes Matching von Ar-
beitnehmern und Betrieben.
JEL classification: J01,J3,J4
Keywords: Wage Inequality, Assortative Matching
Acknowledgements: We are grateful to the IAB and to Stefan Bender for assistance
in making this project possible, to Karl Schmidt, Robert Jentsch, Cerstin Erler, and Ali
Athmani for invaluable programming help, and to the Berkeley Center for Equitable Growth
for funding support. We also thank Christian Dustmann, Bernd Fitzenberger, and seminar
participants at EUI Florence, Yale, and the NBER Labor Studies meetings for many helpful
comments and suggestions. Robert Heimbach, Hedvig Horvath, and Michele Weynandt
provided excellent research assistance.
IAB-Discussion Paper 26/2012
4

1 Introduction
Many of the world’s developed economies have experienced dramatic increases in wage
inequality in recent decades (see Katz and Autor, 1999; and Acemoglu and Autor, 2011
for detailed reviews). While the timing varies from country to country, in each case there
was a widening of wage differentials between observable skill groups (e.g., age and edu-
cation groups), coupled with a rise in within-group inequality. The standard framework for
analyzing these changes focuses on shifts in supply and demand for different skill groups
(e.g., Katz and Murphy, 1992; Bound and Johnson, 1992; Juhn, Murphy and Pierce, 1993;
Goldin and Katz, 2008), mediated by institutional factors like minimum wages and unions
(e.g., DiNardo, Fortin, and Lemieux, 1996). Using a variety of data sources, researchers
have concluded that supply, demand, and institutions have all contributed to trends in wage
inequality, albeit with different magnitudes in different countries and time periods (Freeman
and Katz, 1994; Katz and Autor, 1999).
Although most studies focus on the causes of rising wage inequality between workers
with different skills, economists have long recognized that employer heterogeneity may
generate wage variation among workers with identical skills, even within narrowly defined
economic sectors (Slichter, 1950; Reynolds, 1951; Lester, 1967; Rees and Schultz, 1970).
There are large differences in productivity across firms and plants (Bernard, Eaton, Jensen,
Kortum, 2003; Bloom and Van Reenen, 2007; Foster, Haltiwanger, and Syverson, 2008;
Hseih and Klenow, 2009). Modern search-theoretic models with realistic frictions suggest
that such differences will lead to an employer-specific component of wage inequality among
otherwise identical workers (see e.g., Lentz and Mortensen, 2010). Likewise, an earlier
generation of studies argued that employers can offer efficiency wages to reduce shirking
and boost morale (Akerlof, 1982; Shapiro and Stiglitz, 1984; Krueger and Summers, 1988;
Katz and Summers, 1989). And, in a tradition dating back to Adam Smith, economists have
argued that workplace-specific disamenities like longer hours or more strenuous effort will
lead to compensating wage differences across jobs (e.g., Brown, 1980).
The magnitude of this workplace component of wage inequality is explored in a number
of recent papers that utilize matched employer-employee datasets (Abowd, Kramarz, and
Margolis, 1999; Goux and Maurin, 1999; Abowd, Creecy, and Kramarz, 2002; Gruetter and
Lalive, 2009).
1
Virtually all these studies find a significant employer-specific wage compo-
nent. To date, however, little evidence exists on whether this component has become larger
over time, and if so whether such changes can explain some of the trend in cross-sectional
wage inequality.
2
A priori, workplace heterogeneity could have become more or less important in recent
1
An earlier generation of studies (e.g., Davis and Haltiwanger, 1991; Groshen, 1991; Bernard and Jensen,
1995) documented substantial between-plant variation in wages but was unable to deal fully with nonrandom
assignment of workers to firms.
2
Barth et al. (2011) find that between plant inequality has grown over time in the U.S. but are unable to fully
account for changes in the pattern of sorting of workers to firms due to limitations in their data. Cardoso
(1997, 1999) provides evidence that between plant wage variation grew in Portugal but is again unable to
fully account for selection of workers into firms based upon unobservables. Skans et al. (2009) provide
similar between-plant evidence for Sweden.
IAB-Discussion Paper 26/2012
5

Frequently asked questions

Q1. How much of the variance of wages would have increased under this scenario?

Under this scenario, the variance of wages would have risen by 0.077, or about 70% of the actual rise, suggesting that the increase in sorting can account for about 30% of the rise in variance. 

Q2. What is the reason for the difference in wages between workers in different collective bargaining arrangements?

the trend in coverage seems48 Gürtzgen (2007) suggests that a substantial fraction of the observed differences in mean wages between workers in different collective bargaining arrangements is due to selectivity. 

Q3. What are the recent contributions by Andersson et al. (2012)?

8 Recent contributions by Andersson et al. (2012) and Bagger, Sorensen, and Vejlin (2012) also document increases in assortative matching between workers and employers. 

Q4. How many different establishment effects are assigned to workers born relatively early in the study period?

Note that establishments born relatively early in their study period are assigned up to four different establishment effects, depending on how long they survive. 

Q5. What is the potential presence of a unit root component in the worker-firm model?

The potential presence of such a component motivates in part their decision to estimate worker-firm models over relatively short time intervals, during which the variance of ζit is limited. 

Q6. How do the authors avoid selecting on the job-changer’s own wage?

To avoid selecting on the job-changer’s own wage the authors classify jobs based on the mean wages of co-workers, omitting the wage of the individual being studied (i.e., using the “leave-out mean” wage). 

Q7. How much of the growth in inequality is attributable to rising assortativeness between workers?

The authors find that rising assortativeness between workers and establishments also explains an important share (around 40%) of the growth in inequality across occupations and industries. 

Q8. How much of the variance in wages is due to rising dispersion in worker quality?

As shown in column 5 of the table, rising dispersion in worker quality explains a sizeable share (about 44%) of the rise in the variance of wages across industries, while rising dispersion in establishment pay premiums contributes another 19%. 

Q9. How much has the fraction of establishments with some form of collective bargaining fallen since the mid?

This fraction has fallen from 50-55% for establishments born from 1986 to 1996, to 30% or less for establishments born after 2007.