Occ asiOnal PaPer series
nO 90 / July 2008
WaGe GrOWTH DisPersiOn
acrOss THe eurO
area cOunTries
sOMe sTyliseD FacTs
by Malin Andersson, Arne Gieseck,
Beatrice Pierluigi and Nick Vidalis

---

O C C A S I O N A L P A P E R S E R I E S
N O 9 0 / J U L Y 2 0 0 8
WAGE GROWTH DISPERSION ACROSS
THE EURO AREA COUNTRIES
SOME STYLISED FACTS 1
by Malin Andersson, Arne Gieseck,
Beatrice Pierluigi and
Nick Vidalis
In 2008 all ECB
publications
This paper can be downloaded without charge from
feature a motif
http://www.ecb.europa.eu or from the Social Science Research Network
taken from the
€10 banknote.
electronic librar y at http://ssrn.com/abstract_id= 1084929.
1 Comments from Hans-Joachim Klöckers, Gerard Korteweg, Klaus Masuch, Ad van Riet and an anonymous referee as well as editing
by Claire Vaudry are gratefully acknowledged.

---

© European Central Bank, 2008
Address
Kaiserstrasse 29
60311 Frankfurt am Main
Germany
Postal address
Postfach 16 03 19
60066 Frankfurt am Main
Germany
Telephone
+49 69 1344 0
Website
http://www.ecb.europa.eu
Fax
+49 69 1344 6000
All rights reserved. Any reproduction,
publication or reprint in the form of a
different publication, whether printed or
produced electronically, in whole or in
part, is permitted only with the explicit
written authorisation of the ECB or the
author(s).
The views expressed in this paper do not
necessarily refl ect those of the European
Central Bank.
ISSN 1607-1484 (print)
ISSN 1725-6534 (online)

---

CONTENTS
C O N T E N T S
ABSTRACT 4
ANNEX 4
REAL WAGE LEVELS
4 8
EXECUTIVE SUMMARY
5
ANNEX 5
1 INTRODUCTION
7
SYNCHRONISATION OF WAGE GROWTH IN
THE SIX MACRO SECTORS
5 0
2 DISPERSION IN WAGE GROWTH RATES
ACROSS EURO AREA COUNTRIES OVER TIME 1 0
EUROPEAN CENTRAL BANK OCCASIONAL
PAPER SERIES SINCE 2007
5 2
3 WAGE GROWTH DISPERSION IN THE
EURO AREA COMPARED TO THAT IN
SELECTED BENCHMARK AREAS
1 4
4 COUNTRY DEVELOPMENTS BEHIND
EURO AREA WAGE GROWTH DISPERSION
1 8
5 FACTORS BEHIND WAGE GROWTH
DIFFERENTIALS AMONG THE EURO AREA
COUNTRIES 2 1
6 DISPERSION IN NOMINAL WAGE LEVELS
ACROSS EURO AREA COUNTRIES
2 7
6.1 Some stylised facts
2 7
6.2 How much wage level dispersion
can be explained by dispersion in
labour productivity levels?
3 1
6.3 Can wage growth dispersion be
seen empirically as a longer-term
wage level convergence process
across euro area countries?
3 4
7
CO-MOVEMENT OF COMPENSATION
GROWTH RATES ACROSS EURO AREA
COUNTRIES AND SECTORS
3 6
7.1 Some stylised facts
3 6
7.2 Business cyles and wage growth
synchronisation
3 9
REFERENCES 4 1
ANNEX 1
DATA SOURCES AND COVERAGE
4 4
ANNEX 2
THE INCLUSION OF CYPRUS, MALTA AND
SLOVENIA 4 5
ANNEX 3
MEASURES OF DISPERSION
4 6
ECB
Occasional Paper No 90
July 2008
3

---

ABSTRACT
The study also shows a limited co-movement
of wage growth across countries, even in the
This study presents some stylised facts on context of a high degree of business cycle
wage growth differentials across the euro area synchronisation seen in the last few years.
countries in the years before and in the fi rst This suggests that the impact on wage growth
eight years after the introduction of Economic of country-specifi c developments across euro
and Monetary Union (EMU) in 1999. The study area countries has been larger than the impact
shows that wage growth dispersion, i.e. the of common cyclical developments and external
degree of difference in wage growth at a given shocks. This could refl ect the normal and
point in time, has been on a clear downward desirable working of adjustment mechanisms,
trend since the early 1980s. However, wage which – in an optimally functioning currency
growth dispersion across the euro area countries union with synchronised business cycles –
still appears to be higher than the degree of would take place via price and cost and wage
wage growth dispersion within West Germany, developments. On the other hand, structural
the United States, Italy and Spain.
impediments, for example a relatively low degree
of openness in domestically-oriented sectors in
Differences in wage growth rates between some countries, might prevent a stronger link
individual euro area countries and the euro area between the degree of synchronisation of wage
in the years before and in the fi rst eight years after
growth rates and business cycles.
the introduction of EMU appear to be positively
related to the respective differences between Key words: cross-country wage dispersion,
their Harmonised Index of Consumer Prices wage and productivity levels across countries
(HICP) infl ation and average HICP infl ation in and sectors. JEL: E24, E31, C10.
the euro area. Conversely, relative wage growth
differentials across euro area countries have
been somewhat unrelated to relative productivity
growth differentials. Some countries combine
positive wage growth differentials and negative
productivity growth differentials vis-à-vis the
euro area average over an extended period –
and hence positive unit labour cost growth
differentials. These countries run the risk of
accumulating competitiveness losses and it is
therefore a challenge to ensure that the necessary
adjustment mechanisms operate fully, in the
sense that wage developments are suffi ciently
fl exible and refl ect productivity developments.
Wage growth persistence within individual
euro area countries – largely refl ecting infl ation
persistence and certain institutional factors
– might also have contributed somewhat to
wage growth differentials across the euro area
countries. Moreover, wage level convergence
has also played a role in explaining wage
growth patterns in the 1980s and the 1990s.
However, since 1999, the link between the
initial compensation level and the subsequent
growth rate of compensation per employee
appears barely signifi cant.
ECB
4
Occasional Paper No 90
July 2008

---

E X E C U T I V E
S U M M A R Y
EXECUTIVE SUMMARY
• The current degree of wage growth dispersion
across the euro area countries appears to
Wage growth differentials are a desirable
be higher than the degree of wage growth
feature of a well-functioning economy. Such
dispersion within West Germany, the United
differentials are necessary in order to refl ect
States, Italy and Spain. The lower degree of
differences in local labour market conditions,
dispersion of wage developments within the
“catching-up” factors, and diverse productivity
benchmark economies might be attributable
developments across sectors and regions. In the
to a much more advanced convergence
context of a monetary union and in the absence
process fostered by long histories of a shared
of adjustment via exchange rate changes,
currency and a higher degree of economic
nominal wages also serve as an important
integration, especially labour mobility,
vehicle for adjustment via the competitiveness
inside the four benchmark areas than within
channel.
the euro area, as well as lower infl ation
dispersion in the benchmark areas.
The objective of this study is to look at some
important aspects of nominal wage dynamics • The decline in wage growth dispersion
in the euro area, as nominal wages appear to
in the euro area does not stem from lower
have contributed signifi cantly to differences in
wage growth differences for some outliers
unit labour cost developments across countries.
with respect to the euro area average but can
The study analyses wage growth differentials
be related to declining differentials in most
by considering both the dispersion of wage
euro area countries. Taking into account the
growth rates across the euro area countries,
adjustment process in Germany, following
i.e. the degree of difference in wage growth
unifi cation, the persistence of wage growth
rates at a given point in time, and by assessing
differentials across euro area countries
the degree of synchronisation of wage growth
appears to be comparable to that in the
across the euro area countries, i.e. the degree
benchmark areas.
of co-movement of wage growth rates over a
certain period of time. Moreover, the study • Certain longer-term factors appear to be
presents evidence on wage level developments
behind wage growth differentials among
across the euro area countries, as wage growth
the euro area countries. Differences in wage
differentials might be attributable to catching-
growth rates between individual euro area
up processes in some countries.
countries and the euro area appear to be
positively related to the respective differences
The key fi ndings of this study are the following:
between their HICP infl ation and average
HICP infl ation in the euro area. Conversely,
• The dispersion of wage growth per
relative wage growth differentials across euro
employee and per hour across the euro
area countries have largely been unrelated
area countries was on a clear downward
to and are generally higher than relative
trend during the 1980s and the early 1990s.
productivity growth differentials. Although
Since about 1993, however, most measures
for a number of countries the relative wage
suggest that wage growth dispersion has
and productivity growth differentials appear
fl
uctuated within a relatively narrow
to be small, countries that combine positive
range. Wage growth dispersion exhibited a
wage growth differentials and negative
similar picture in all main sectors (except
productivity growth differentials vis-à-
agriculture) across the euro area countries,
vis the euro area average over an extended
but cross-country wage growth dispersion
period – and hence positive unit labour
at the sectoral level was larger and its
cost growth differentials – run the risk of
evolution less stable than wage growth
accumulating losses in competitiveness. It
dispersion in the total economy.
is therefore a challenge for those countries
ECB
Occasional Paper No 90
July 2008
5

---

in which relative wage developments • Turning to the analysis of wage growth
exceed relative productivity developments
synchronisation, the cross-correlation analysis
to ensure that the necessary adjustment
suggests that there are no signifi cant
mechanisms operate fully, in the sense that
co-movements in wage developments
wage developments are suffi ciently fl exible
within the euro area. In particular, the rather
and refl ect productivity developments. Wage
high and stable degree of business cycle
growth persistence within individual euro
synchronisation seen in recent years does
area countries – largely refl ecting infl ation
not seem to have coincided with a similar
persistence and certain institutional factors,
degree of wage growth synchronisation.
such as indexation and multi-year contracts –
This suggests that, while the presence of
might also have contributed, to some extent,
common shocks might have played a role in
to wage growth differentials across the euro
the synchronisation of business cycles across
area countries.
euro area countries, wage growth remains
dominated by country-specifi c factors. A low
• Wage level convergence, albeit still far
degree of wage growth synchronisation and
from being complete, has played a role in
a loose link to more synchronised business
explaining wage growth patterns during the
cycles might, in fact, be desirable, as in an
1980s and the 1990s. In this period, growth
optimal currency union with synchronised
in compensation per employee had indeed
business cycles, adjustment to shocks would
been visibly slower in countries with high
take place via relative price and cost changes.
initial levels of compensation, while a higher
On the other hand, structural reasons, related
rate of wage growth had been observed in
for example to a low degree of competition
countries with initially low compensation
and a relatively low degree of openness
per employee levels. However, in the
in domestically-oriented sectors in some
period after the inception of EMU, the link
countries, might also prevent a stronger
between the initial compensation level in
link between the degree of synchronisation
1999 and the subsequent average growth
of wage growth rates and business cycle
rate of compensation per employee appears
synchronisation in these sectors. This
barely signifi cant. This seems to imply that
might be considered a potential source
some convergence of wage levels took place
for concern, in particular to the extent that
before 1999, bolstered by the completion of
such factors prevent relative wage growth
the Single Market, while it came broadly to
developments to follow relative productivity
a halt after the inception of Stage 3 of EMU.
developments.
As wage level convergence is far from
complete, it may continue to be a factor
behind wage growth dispersion within the
euro area for the foreseeable future.
• The modest decline in the dispersion of wage
levels was in line with a modest decline in
the dispersion of productivity levels between
1993 and 2006. However, for certain
individual euro area countries – both for the
total economy as well as for their individual
sectors – the developments in relative
wage levels have not always followed the
developments in relative productivity levels.
ECB
6
Occasional Paper No 90
July 2008

---

I I N T R O D U C T I O N
1 INTRODUCTION
in productivity growth rates across countries,
persistent nominal wage growth differences have
Wage growth differentials are a desirable led to considerably diverse cumulated unit labour
feature of a well-functioning economy. Such cost developments. These differences have, in
differentials are necessary in order to refl ect turn, contributed to signifi cant differentials in
differences in local labour market conditions, competitiveness developments and in infl ation 3,
“catching-up” factors, and diverse productivity with repercussions for economic activity and
developments across regions. In the context of a employment over time.
monetary union and in the absence of adjustment
via nominal exchange rate changes, nominal Monetary policy is conducted by the Governing
wages also serve as an important vehicle for Council of the ECB with the primary objective
adjustment via the competitiveness channel.
of maintaining price stability in the euro area as a
whole. Monetary policy cannot therefore address
In the run up to EMU, there was widespread differences in wage growth or other country-
concern amongst policymakers that signifi cant specifi
c economic developments. However,
rigidities and other shortcomings in the wage it is necessary for the European Central Bank
formation processes across the euro area (ECB) to assess the underlying causes of such
countries could seriously impair the effi cient wage differentials, as this is key to better
working of EMU.1 One reason for concern was understanding euro area wage developments
that excessive nominal wage increases in some and it facilitates the identifi cation of structural
euro area countries, particularly in the larger barriers that may hamper macroeconomic
euro area countries, could provoke a tightening adjustments in the euro area.
of monetary conditions with possibly adverse
effects on growth and employment in the entire Against this general background, this study
monetary union. Another concern was that presents some stylised facts on nominal wage
large and persistent positive nominal wage differentials across the euro area countries. The
growth differentials, which do not refl ect objective is to quantify the heterogeneity of wage
productivity differentials across countries, growth and wage levels from a cross-country
could lead to substantial increases in unit standpoint using standard measures for dispersion
labour costs in some euro area countries. With and synchronisation and to discuss the factors
nominal exchange rate devaluation no longer which may be behind these differentials from a
being an option, substantial and persistent unit cross-country point of view. The study takes
labour cost growth would cause severe losses a strictly factual approach and does not aim
in competitiveness with adverse repercussions at a normative assessment of wage growth
for economic activity and employment in some heterogeneity across euro area countries. Moreover,
euro area countries. In particular, it was feared the study is entirely based on a cross-country
that substantial and persistent above-average approach, and it does not consider the working of
unit labour cost growth would ultimately adjustment processes of individual countries via
translate into deteriorating labour market nominal wages. The study starts with an analysis
conditions in these euro area countries, of the dispersion of wage growth, i.e. the degree
requiring painful adjustment thereafter.
of difference in wage growth at a given point in
time, across the euro area countries both overall
Several years after the start of the third stage of and at the sectoral level. The study then moves
EMU, the debate on wage growth differentials on to assess the dispersion of wage levels across
within EMU is still ongoing, in the light of the euro area countries, since differences in wage
relatively high and persistent wage growth in levels could be seen as one major driving factor
some euro area countries and more modest wage
1 See European Commission (1990).
developments in some other countries.2 Indeed, 2 See European Commission (2006a), pp. 79-108.
in an environment of relatively small differences
3 See ECB (2005), pp. 61-77.
ECB
Occasional Paper No 90
July 2008
7

---

behind wage growth dispersion. Finally, the study starting point when considering unit labour cost
presents evidence on the degree of synchronisation and real wage developments.
of wage growth within the euro area, i.e. the degree
of co-movement of wage growth across countries With respect to the question of whether to
over a certain period of time.
consider wages per hour or per person, economic
theory would suggest a focus on wages per hour
The wage concept used in this study is that of worked as the most accurate measure of labour
nominal compensation per employee, i.e. overall
costs. However, empirical work has shown that
compensation paid by employers. Apart from both measures provide useful information on
negotiated wages, the concept of compensation wage developments. While in past decades,
per employee also includes wage drift and wages in terms of persons could have been
social security contributions. The concept of considered as a rather good approximation of
compensation per employee is a convenient wages per hour worked, in more recent times
choice for international comparisons owing to this might not be the case. In fact, one of the key
the availability of related data and its similar stylised facts of the euro area’s labour markets
defi nition across countries, as opposed to other is that the annual average working time per
“narrower” wage concepts where the defi nitions worker has declined substantially across the
may differ substantially from one country to euro area countries over the last 25 years.4 This
the other. In what follows, the words “wage” is attributable to the increased use of part-time
and “compensation per employee” will be used working arrangements, which is often related to
interchangeably.
the greater number of women entering the labour
market, to institutional factors such as tax
Nominal wages play an important role in the wedges which create disincentives to work, or
assessment of macroeconomic developments to specifi c policy measures including changes in
in a number of ways. From a business cycle working time regulations, such as the
perspective, nominal wages are a key factor introduction of the 35 hour week in France and
driving income growth and distribution. recent labour market reforms in Germany and
Nominal wage developments also play an Italy.5 As was the case in France, the decline in
important role in shaping the path of overall working time has often been accompanied by
economic activity, infl ation and employment. increasing hourly compensation in order to
While the latter variables may also exert an broadly maintain monthly compensation levels.6
important impact on nominal wage developments,
Against this background, results will be
nominal wages can be regarded as a relatively presented as far as possible on the basis of data
“exogenous” variable as they are mainly for both compensation per employee and
determined in wage negotiations which might compensation per hour.
or might not take into account other economic
relationships. From a cross-country point of The following analysis is backward looking and
view and compared with other benchmark is built as far as possible on data covering the
areas, nominal wages are the main driving 12 countries that joined EMU before 2006.7
force behind the dispersion of unit labour cost Our data requirements in terms of frequency,
developments in an environment of relatively
modest divergence in productivity growth. In 4 For a detailed discussion see Leiner-Killinger et al. (2005).
5 Specifi cally, the introduction of very fl exible employment
the context of monetary union and the absence
contracts in Italy had the effect of triggering a strong increase
of exchange rates as the traditional “bailout”,
of “part-time” workers, thus creating a signifi cant gap between
nominal wages serve as a key instrument for
the growth rate of employees measured in heads and that of total
hours worked.
adjustment via the competitiveness channel 6 See ECB (2006), pp. 43-44.
and play an important role in the analysis of 7 Due to limited availability of data, Malta, Cyprus and Slovenia
are not included in the analysis. However, Annex 2 shows that
competitiveness developments across the euro
from 1996 onwards, the inclusion of these countries does not
area countries. Thus, nominal wages are the
affect the main conclusions of the analysis.
ECB
8
Occasional Paper No 90
July 2008

---

I I N T R O D U C T I O N
sample length and degree of sectoral irrevocable conversion rates. This means that
disaggregation required the use of various exchange rate movements across euro area
databases. National account data (ESA2000) of countries do not have any impact on the growth
compensation per employee are available for all rates or levels shown in this study.
euro area countries at an annual frequency.
However, when the analysis requires the use of
quarterly information, data limitations
necessitate the use of a euro area aggregate
(EA8) on the basis of only eight countries,
encompassing Germany, France, Italy, Spain,
the Netherlands, Belgium, Austria and Finland.
The national account data cover the total
economy and the six largest sectors (agriculture,
industry excluding construction, construction,
trade and transport, fi nancial intermediation, and
other services). However, national statistical
offi ces in most euro area countries have not or
have only recently started to collect data on the
number of hours worked, in annual and quarterly
terms. To overcome this limitation of national
accounts, the EU KLEMS database compiled by
the Groningen Growth and Development Centre
(GGDB) has been utilised to enlarge the
available information set.8 In particular, these
alternative sources allow for an analysis of wage
dispersion in terms of hours worked across the
various euro area countries and across a large
number of sectors. This data currently extend
only up to 2004. Given the diffi culties in
evaluating the reliability of hours worked, which
are not included in the “offi cial”, i.e. national
statistical offi ce databases, this paper will, in
general, put more emphasis on the results found
in terms of persons.
With respect to the starting date of the data
sample, while the data are available from 1980
onwards, the following analysis is mainly
focused on the period from 1993 to 2006.
From 1993, bilateral exchange rates of the
12 countries which had adopted the single
currency before 2006 were either factually fi xed
or fl uctuated within a relatively narrow band,
roughly allowing for the hypothesis that nominal
wages – and not exchange rate movements –
was the key variable determining relative unit
labour cost developments across euro area
countries. Finally, all data used in this study 8 For details on the databases and the sectoral disaggregation,
are denominated in euro using the respective
see Annex 1.
ECB
Occasional Paper No 90
July 2008
9

---

Document Outline

• WAGE GROWTH DISPERSION ACROSS THE EURO AREA COUNTRIES - SOME STYLISED FACTS, JULY 2008
• CONTENTS
• ABSTRACT
• EXECUTIVE SUMMARY
• 1 INTRODUCTION
• 2 DISPERSION IN WAGE GROWTH RATES ACROSS EURO AREA COUNTRIES OVER TIME
• 3 WAGE GROWTH DISPERSION IN THE EURO AREA COMPARED TO THAT IN SELECTED BENCHMARK AREAS
• 4 COUNTRY DEVELOPMENTS BEHIND EURO AREA WAGE GROWTH DISPERSION
• 5 FACTORS BEHIND WAGE GROWTH DIFFERENTIALS AMONG THE EURO AREA COUNTRIES
• 6 DISPERSION IN NOMINAL WAGE LEVELS ACROSS EURO AREA COUNTRIES
  • 6.1 SOME STYLISED FACTS
  • 6.2 HOW MUCH WAGE LEVEL DISPERSION CAN BE EXPLAINED BY DISPERSION IN LABOUR PRODUCTIVITY LEVELS?
  • 6.3 CAN WAGE GROWTH DISPERSION BE SEEN EMPIRICALLY AS A LONGER-TERM WAGE LEVEL CONVERGENCE PROCESS ACROSS EURO AREA COUNTRIES?
• 7 CO-MOVEMENT OF COMPENSATION GROWTH RATES ACROSS EURO AREA COUNTRIES AND SECTORS
  • 7.1 SOME STYLISED FACTS
  • 7.2 BUSINESS CYLES AND WAGE GROWTH SYNCHRONISATION
• REFERENCES
• ANNEXES
  • 1 DATA SOURCES AND COVERAGE
  • 2 THE INCLUSION OF CYPRUS, MALTA AND SLOVENIA
  • 3 MEASURES OF DISPERSION
  • 4 REAL WAGE LEVELS
  • 5 SYNCHRONISATION OF WAGE GROWTH IN THE SIX MACRO SECTORS
• EUROPEAN CENTRAL BANK OCCASIONAL PAPER SERIES SINCE 2007

---