EurWORK European Observatory of Working Life

Wages and working conditions in the crisis

About

Author: 
Sem Vandekerckhove, Jan Van Peteghem and Guy Van Gyes
Institution: 
Higher Institute of Labour Studies, HIVA-KU LEUVEN

  [ pdf version size 2055 kb ]  - see also the Executive Summary  [ pdf version size 39.07 kb ]

The economic and financial crisis of 2008–2010 has impacted on pay in most EU Member States leading to wage deceleration, pay freezes and sometimes pay cuts. The crisis hit vulnerable groups (low-skilled, young, migrants) particularly hard. Data from five key sectors (manufacturing, construction, accommodation and food services, financial services, public administration) reveal more crisis effects on employment than on wages. Cuts in low-paid and temporary jobs, or reductions in their hours, tended to be the first measure adopted while the ‘wage cushion’ often seen in higher-ranking jobs allowed cost savings through cuts in bonuses and other rewards. Cutting wages is also seen as detrimental to worker motivation and retention. Most responses taken were temporary with few trade-offs at company level between wages and other elements of the employment relationship.

The study was compiled on the basis of individual national reports submitted by the EWCO correspondents. The text of each of these national reports is available below. The reports have not been edited or approved by the European Foundation for the Improvement of Living and Working Conditions. The national reports were drawn up in response to a questionnaire and should be read in conjunction with it.


Introduction

This comparative analytical report (CAR) examines the relationship between the recent evolution of earnings and working conditions in Europe, with the current crisis as the economic background. Did the wage freezes and cuts translate into an improvement of other working conditions such as job security guarantees, enhanced access to training and career development or reduced working hours/overtime? Or did working conditions deteriorate at the same time and in the same occupations/sectors as those where there were wage cuts and freezes?

Context

The current financial and economic crisis has put increased pressure on workers’ wages in most countries of the European Union (TN1109060S). This pressure has been highlighted in recent negotiations and agreements at national, sector and company level in which wage moderation, pay freezes or cuts have been agreed. However, forms of wage moderation, pay freezes or pay cuts have been agreed in some cases in an attempt to limit the negative impact of the crisis on other working conditions such as job security, employment guarantees and better access to training. Furthermore, limited wage increases have been observed in some economic sectors while other sectors and companies have experienced pay freezes or pay cuts. As highlighted in the note Assessing the links between wage setting, competitiveness and imbalances (511Kb PDF) from the European Commission Directorate General of Economic and Financial Affairs (DG ECFIN) to the Economic Policy Committee in July 2011, these wage developments have also come to the forefront of discussions on how to tackle the current macro-imbalances.

In response to the problems posed by the financial and economic crisis, the European Commission has put forward a series of new policies better known as ‘European economic governance’. As emphasised in the Euro Plus Pact (166Kb PDF), wages and collective bargaining systems are seen as one of the main instruments for the European coordination of economic policy and wage developments should be in line with developments in labour productivity. At the same time, the commitments to the Europe 2020 strategy are reconfirmed. Making work more attractive or enhancing job quality are among the Europe 2020 commitments. The report examines how these two policy goals have been combined in Member States in the current period of economic crisis.

Methodology

The report covers the 27 EU Member States and Norway, and is based on the answers of the national European Working Conditions Observatory (EWCO) correspondents to a structured questionnaire about the issue.

The report covers mainly the first phase of the crisis, that is, the financial or credit crisis followed by the first abrupt economic downturn. In 2008, a worldwide crisis set in following the crisis in the housing sector in the USA (the credit crisis). At about the same time, EU Member States such as Ireland and Spain were trapped by a stimulating monetary policy in the euro zone, which caused overheating of their economies (also from the housing sector). The crisis in the international financial sector followed that in construction and real estate as many banks now held worthless shares in funds derived from the housing market. Countries intervened through nationalisation of banks or state bank guarantees to prevent a collapse of the financial system and to limit the effect of the financial crisis on other sectors.

Some countries (for example, Belgium and Germany) recovered quickly from the first crisis mainly as a result of a growing demand from emerging economies, although most countries did not succeed in dealing fully with the impact of the crisis in the following years. Initially Greece appeared to mask its real economic state, which put pressure on the euro. As a consequence, other Mediterranean countries and Ireland became targeted by investors in government bonds. As interest rates rose, the economic outlook of the whole of Europe worsened – causing what is termed the ‘debt crisis’. However, this 2011 crisis is not included in this report as there is generally a delay of at least a year after events in the availability of data from administrative sources and academic publications on wages. Therefore, this report focuses on the 2008–2009 crisis (the Great Recession) and the aftermath up until 2010, although anecdotal evidence from 2011 is included in the policy discussion.

Because the crisis has affected different economic sectors differently, the report takes a sectoral approach at different points by adopting the broad statistical classification of economic activities in the European Community, NACE Rev. 2 (5.63Mb PDF), set out in Table 1.

Table 1: Sector codes used in this report
Code Sector
C

Manufacturing

F

Construction

I

Accommodation and food services (hotels, restaurants and catering, further referenced to as Horeca)

K

Financial services, banking sector

O

Public administration

Structure of the report

The report has four sections.

Section 1 describes of the main wage trends in EU countries in the period 2008–2010. Attention is paid to the wage evolution and, to a lesser extent, wage inequality and the use of variable pay. When available, national data are presented according to individual characteristics such as gender, age, type of employment contract and occupation.

Section 2 presents recent national studies that investigated the relationship between wages (and their recent evolution) and aspects of working conditions. It also considers what empirical data are available on how working conditions factors such as job security and job satisfaction have been evolving alongside existing wage trends.

Section 3 contains examples of key company and/or sectoral arrangements where a trade-off between wages and other features of the employment relationship was obtained including:

  • wage moderation;
  • pay freezes or pay cuts in exchange for improvements such as job security, other forms of rewards, working time reduction, and more and better training.

Section 4 provides an overview of recently implemented policies to support groups of workers who have been affected by recent negative wage trends. Examples of these policies include minimum wage policies, training financing, and social benefits such as transport and housing support.


Wage trends 2008–2010

General country trends

The reports from the national correspondents document the key official sources on wage information for their country. These sources contain the most appropriate information on workers’ wages, in contrast with the commonly gathered information on wage costs or wage information derived from national accounts that also include company payments for social benefits. It is important to compare the information in the national reports and EU statistics from DG ECFIN’s annual macroeconomic database (AMECO). This study looks specifically at what happened in 2009 and 2010, the two main years following the beginning of the credit crisis.

Table 2 sets out the country trends for these years (precise data are given in Annex 1 to this report) in relation to wage deceleration, that is, an average actual nominal wage growth lower than the year before. The table shows that wage deceleration took place in almost all EU Member States in 2009, with the notable exceptions of Denmark, France and Greece. In 2010, the pace of wage growth continued to decrease in about one third of the countries (Bulgaria, the Czech Republic, Denmark, Greece, Finland, Hungary, Ireland, Luxembourg and Poland, Romania, Slovakia, Sweden and the UK). This figure is influenced by the strength of wage growth during the upward business cycle of 2008.

Table 2: Wage deceleration by country, 2009 and 2010

Country

2009

2010

Austria

yes

n.d.

Belgium

yes

n.d.

Bulgaria

yes

yes

Cyprus

yes

n.d.

Czech Republic

yes

yes

Denmark

no

yes

Estonia

n.d.

n.d.

Germany

yes

no

Greece

no

yes

Finland

yes

yes

France

no

n.d.

Hungary

yes

yes

Ireland

yes

yes

Italy

yes

no

Latvia

yes

no

Lithuania

yes

no

Luxembourg

yes

yes

Malta

yes

no

Netherlands

yes

n.d.

Poland

yes

yes

Portugal

yes

no

Romania

yes

yes

Slovakia

yes

yes

Slovenia

yes

no

Spain

yes

n.d.

Sweden

yes

yes

United Kingdom

yes

yes

Norway

yes

no

Notes: n.d. = no data

Austria and UK figures are based on median wage data.

Source: National reports based on national wage statistics

The analysis further examined wage cuts, that is, a lower average wage in comparison with the year before. Nominal figures were taken from the national reports and real figures were derived from these by deflating them (accounting for country-specific inflation) by the consumer price index (CPI) used in the AMECO dataset for calculating national account based real wages. This allowed both sources to be compared.

Finally, wage freezes are the special case where wages show zero growth. In this case, a band of ±0.3% has been taken to allow for the fact that most of the gathered data come from surveys. If possible, wage freezes are preferred to cuts, either in nominal or real terms. Therefore, some spiking around zero wage growth and wage growth equal to inflation (coined ‘rigidity’ by some) could be expected. As highlighted in a Eurofound report on wage indexation in the European Union, adjustments will of course be mainly in real terms in countries with a strong system of (automatic) wage indexation such as found in Belgium, Cyprus, Luxembourg and Malta.

Country classification in terms of wage growth

The maps in Figures 1–3 summarise the data for 2009 and 2010, and classify the countries in terms of the amount of wage growth.

  • Negative figures represent a wage cut. These countries are coloured red and orange.
  • A country with a wage freeze on average is presented in yellow.
  • Blue represents countries with an average positive wage growth.

Data are presented in nominal and real terms. Figures in real terms come from two sources: AMECO and the national contributions for this report. Detailed data are given in the tables in Annex 1 to this report.

Figure 1: Nominal average wage growth by country, 2009 and 2010 (%)

tn1203015s.tmp00.jpg

Note: dark blue = (0.1, 0.3); mid-blue = (0.05, 0.1); light blue = (0.03, 0.05); light orange = (-0.03, 0.03); dark orange = (-0.05, -0.03); red = (-0.2, -0.05); white = no data.

Source: National reports

Figure 2: Real average wage growth by country, 2009 and 2010

tn1203015s.tmp01.jpg

Note: dark blue = (0.1, 0.3); mid-blue = (0.05, 0.1); light blue = (0.03, 0.05); light orange = (-0.03, 0.03); dark orange = (-0.05, -0.03); red = (-0.2, -0.05); white = no data.

Source: National reports

Figure 3: Real average wage growth by country, 2009 and 2010 (AMECO)

tn1203015s.tmp02.jpg

Note: dark blue = (0.1, 0.3); mid-blue = (0.05, 0.1); light blue = (0.03, 0.05); light orange = (-0.03, 0.03); dark orange = (-0.05, -0.03); red = (-0.2, -0.05); white = no data

Source: AMECO

The data on average nominal wage growth in Figure 1 show that the freezing of wages is the dominant pattern. Only two countries (Latvia and Lithuania) had average wage cuts in 2009, being joined in 2010 by Greece and Ireland. Poland, Slovakia, Slovenia and Norway are the only countries with a persistent wage growth on average.

Findings differ for real wage growth according to the data source. In 2009, the figures from the national reports (Figure 2) are more positive than the AMECO calculations (Figure 3) but, in 2010, it is the opposite. AMECO sketches a dominant picture of freezing of the average wage in both years, with the exception of real cuts in Hungary, Latvia, Lithuania and Romania in 2009 and Greece, Hungary, Latvia and Romania in 2010. The data from national reports still show a picture of an average real positive wage growth in 2009 with the exception of Latvia and Lithuania. By this time the west European countries belonging to the euro zone already show a trend of freezing wages. The 2010 data from the national reports show a consistent pattern of average wage cuts in the majority of countries for which data are available.

Information from European Central Bank’s company survey

A company survey carried out by the Wage Dynamics Network (WDN) of the European Central Bank (ECB) included questions on the company’s capacity to deal with an adverse ‘demand shock’; a demand shock is a sudden surprise event that temporarily increases or decreases demand for goods or services. The survey was held in 12 EU countries covering a total of 11,918 companies in 2007–2008 and repeated in mid-2009 in nine countries. Table 3 presents data from those nine countries where both waves of the survey took place.

Table 3: Incidence of wage cuts and freezes
  Nominal wage cuts Nominal wage freezes
% Companies % Employees % Companies % Employees
2007 2009* 2009** 2007 2009 (past) 2007 2009* 2009** 2007 2009 (past)
AT

1.54

1.72

1.51

0.06

1.23

9.30

1.76

8.43

5.71

1.07

BE

2.87

1.04

1.76

0.20

0.27

15.89

23.72

4.41

2.39

14.58

CZ

9.32

8.95

3.24

1.13

3.71

31.39

54.63

11.72

11.95

49.13

EE

3.68

44.08

38.61

0.14

30.35

21.27

61.54

64.61

9.60

56.94

ES

0.14

2.55

0.52

0.03

1.35

1.45

26.68

3.73

0.80

22.21

FR

2.54

1.92

4.73

0.86

1.21

7.68

85.98

83.77

5.27

82.48

IT

0.68

2.03

4.29

0.06

1.14

3.81

31.71

62.77

1.25

30.86

NL

1.58

2.55

3.78

0.17

1.18

25.80

15.22

8.67

15.86

12.58

PL

5.70

4.20

1.58

3.79

2.60

9.72

17.98

8.07

7.79

16.59

Euro zone

1.27

2.07

3.29

0.23

1.14

7.64

37.09

43.12

3.89

34.38

Non-euro zone

6.43

6.48

2.68

3.10

3.70

14.80

27.37

10.25

8.76

24.99

Total

2.63

3.22

3.13

0.99

1.81

9.53

34.51

34.46

5.18

31.88

Notes: *So far in 2009; **Expected for 2009. See Annex 2 for a full list of country codes.

Source: Fabiani et al (2010)

To a large extent, these company figures confirm the average wage trends. Wage cuts remained an exception, increasing from only 2.6% of companies in the pre-crisis survey to 3.2% in the follow-up survey, and affecting only 1.8% of all employees (Table 3). Introduction of wage freezes, nevertheless, increased from 9.5% to 34.5% of companies between the two surveys.

Although recent data for Estonia are missing in the national reports, the figures here do seem to fit with the image of the seriously affected Baltic States in the previous analyses, that is, wage freezes are common and wage cuts are found much more than in any other country in this table. Austria, Belgium, the Netherlands and Poland seem to be better off in general.

Affected groups

Although wage cuts may not be the general outcome of the crisis, specific groups may be affected depending on their gender, working time regime or occupational status.

Gender pay gap

Between 2008 and 2009, the gender pay gap was shrinking in most countries (represented by the minus signs in Table 4) with Bulgaria, Hungary, Latvia, Romania, Slovakia and Spain the sole exceptions.

Table 4: Gender pay gap trends, 2008–2010
 

2008

2009

2010

Austria

+

n.d.

Belgium

n.d.

Bulgaria

+

+

Cyprus

n.d.

Czech Republic

+

Denmark

+

+

Estonia

n.d.

n.d.

Finland

France

+

n.d.

Germany

+

+

Greece

n.d.

Hungary

+

+

Ireland

+

n.d.

Italy

n.d.

Latvia

+

+

Lithuania

n.d.

+

Luxembourg

Malta

Netherlands

n.d.

Poland

n.d.

n.d.

n.d.

Portugal

+

Romania

+

n.d.

Slovakia

+

Slovenia

+

Spain

+

n.d.

Sweden

UK

Norway

+

-

-

Notes: (-) decreasing pay gap; (+) increasing pay gap

n.d. = no data

Source: National reports, own classification

Working time regime

Very few data are available on working time regimes. In Belgium, France, Latvia and Spain, full-time workers earned more relative to part-time workers in 2009 compared with 2008. However, this pay gap diminished in Austria, Finland, Germany, Malta, the Netherlands, Slovakia, the UK and Norway.

Occupation category

Wage changes may also differ according to occupation category. National correspondents were asked to provide wage figures for International Standard Classification of Occupations (ISCO) groups at the one-digit level, or similar classifications if the one-digit level is not available (this was the case in Finland, France, Germany, Italy and Spain).

The pay gaps computed based on these figures present the difference between:

  • the top salary occupation category and the median occupation category (high/mid category pay gap);
  • the median group and the low salary occupation category (mid/low category pay gap).

The plus (increasing) and minus (decreasing) signs in Table 5 show the direction of the trend in 2008, 2009 and 2010.

In 2009, the pay gap in the high/mid and the mid/low categories rose in about half the countries and shrunk in the others. Comparison of the wage inequality trend between the mid/low-paid occupations with the high/mid trend shows a mirroring pattern in many countries; for example, the low-paid occupations earned relatively less in Belgium in 2009 while the top-paid occupations moved towards the median wage occupation category, implying less wage inequality in the upper part of the wage distribution. The same pattern is found in Cyprus, Germany, Ireland, Romania and Norway. The opposite pattern (a stretched wage distribution from the top incomes) is found in Bulgaria, the Czech Republic, Finland, Slovakia and the UK.

The report deals with the more robust aggregate wage inequality figures later, but concludes for now that a stretch of the wage distribution in both directions (for example, in Austria, Hungary, Spain and Sweden) was uncommon in the immediate post-crisis period. There is too little information available to make any general conclusion about 2010.

Table 5: Pay gap trends by occupational category, 2008–2010
 

High/mid category

Mid/low category

 

2008

2009

2010

2008

2009

2010

AT

+

n.d.

+

n.d.

BE

+

n.d.

+

n.d.

BG

+

+

+

+

CY

+

n.d.

+

n.d.

CZ

+

+

+

+

DE

+

+

+

DK

+

+

+

+

EE

+

n.d.

n.d.

n.d.

n.d.

EL

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

ES

+

n.d.

+

n.d.

FI

+

+

+

FR

n.d.

n.d.

HU

+

+

+

+

IE

+

n.d.

+

+

n.d.

IT

+

n.d.

n.d.

LT

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

LU

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

LV

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

MT

+

+

NL

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

PL

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

PT

+

n.d.

n.d.

RO

n.d.

+

n.d.

SE

+

+

+

+

+

SI

+

+

SK

+

+

UK

+

+

+

NO

+

+

+

+

Notes: (-) decreasing pay gap; (+) increasing pay gap

n.d. = no data

See each national report separately for how occupations are classified in different categories (mostly based on ISCO). See Annex 2 for a full list of country codes.

Source: National reports

Impact of the crisis on sectoral employment and wages

Information from the national reports can be used to compare sectors with respect to the impact of the crisis. This can either have a negative effect on employment (L), on wages (W), on both (LW), or have no negative effect (O), as represented in Table 6.

Table 6: Classification scheme for wage or employment effect of crisis by sector

Labour/wages

Increase/status quo

Decrease

Increase/status quo

O

W

Decrease

L

LW

The series of maps in Figures 4–8 depict the main effects of the crisis in each Member State, for which data are available, on the five different sectors studied in the two years following 2008 in terms of:

  • employment decrease (L);
  • wage decrease (W);
  • stability/growth, that is, no negative effect (O);
  • employment and wage decrease (LW).

Figure 4: Effects of the crisis in manufacturing sector, 2009 and 2010

tn1203015s.tmp03.jpg

Note: yellow = employment decrease; orange = wages decrease; green = stability/growth; red = employment and wages decrease; grey = no data

Source: National reports

Figure 5: Effects of the crisis in construction sector, 2009 and 2010

tn1203015s.tmp04.jpg

Note: yellow = employment decrease; orange = wages decrease; green = stability/growth; red = employment and wages decrease; grey = no data

Source: National reports

Figure 6: Effects of the crisis in Horeca sector, 2009 and 2010

tn1203015s.tmp05.jpg

Note: yellow = employment decrease; orange = wages decrease; green = stability/growth; red = employment and wages decrease; grey = no data

Source: National reports

Figure 7: Effects of the crisis in financial services sector, 2009 and 2010

tn1203015s.tmp06.jpg

Note: yellow = employment decrease; orange = wages decrease; green = stability/growth; red = employment and wages decrease; grey = no data

Source: National reports

Figure 8: Effects of the crisis in public administration sector, 2009 and 2010

tn1203015s.tmp07.jpg

Note: yellow = employment decrease; orange = wages decrease; green = stability/growth; red = employment and wages decrease; grey = no data

Source: National reports

In 2009, manufacturing was hit by restructuring and downsizing in almost the entire EU. In Figure 4, employment decreases (type L) are evident in 17 countries (see also Table 7). In six countries, the crisis also involved nominal wage decreases (LW).

The picture for construction is similar (Figure 5), although there is no change in either nominal wages or employment in four countries in 2009 (the Czech Republic, Poland, Slovakia and Sweden).

Accommodation and food services (Figure 6) shows a more scattered pattern, with 16 countries still showing an employment decrease (type L or LW). Again, in a small number of countries, the impact of the crisis did not hit this sector.

Financial services, for obvious reasons, was affected primarily by wage decreases (type W or LW), found in 11 countries (Figure 7).

Finally, public administration (Figure 8) was a relatively safe haven in this period; there was no effect of the crisis in this sector in 13 countries.

Country effects are in general dominant, meaning that most sectors are hit in the same way in the same country (see Figures 4–8). Clear cases are Latvia and Lithuania, where both wages and employment were affected in all sectors. Slovenia is an exception here as it has four different types of effects in the five sectors examined.

Table 7: Wage or employment effect of crisis by sector and Member State, 2009
 

Manufacturing

Construction

HORECA

Financial services

Public administration

Total O

1

4

5

5

13

Total L

17

14

12

10

4

Total W

0

2

4

3

3

Total LW

6

6

4

8

3

Source: National reports, own classification

Although there is less information for 2010, a similar pattern is apparent (see Figures 4–8 and Table 8). The number of countries for which data are available ranges from 14 to 18 depending on the sector of activity.

Again, in manufacturing, shrinkage of the labour force is common.

In construction, a limited number of countries escape the crisis, though these are different from the ones in 2009 apart from Sweden.

In Horeca, as in 2009, a variety of patterns emerge, although an employment decrease is most frequently observed. This sector did not seem to suffer much in this period in Denmark, Germany, Hungary and Norway.

In financial services, the decrease in employment continued in 2010, but with fewer cases of wage decline (the Czech Republic, Latvia, Lithuania and Romania only).

Finally, public administration seems to have been hit with some delay in many countries, as the total of countries showing stability/growth (type O) has shrunk from 13 out of 23 where data are available in 2009 to four out of 15 cases in 2010.

In summary, the impact of the crisis seems much more sector-specific in 2010 than in 2009. Nevertheless, Lithuanian employment and wages are again affected in all sectors and, in each sector, more than half the countries show a decrease in employment.

Table 8: Wage or employment effect of crisis by sector and Member State, 2010
 

Manufacturing

Construction

HORECA

Financial services

Public administration

Total O

1

5

4

1

4

Total L

11

7

7

12

6

Total W

0

1

3

1

1

Total LW

4

5

3

3

4

Source: National reports, own classification

Wage dispersion

The subject of wage inequality or wage dispersion was touched on during the discussion in the previous section on vulnerable groups who were particularly affected by the crisis. Recognised as a trending topic in social sciences (Wilkinson and Pickett, 2009), it is of general importance to well-being, welfare and development. Nevertheless, when a crisis hits the economy, there is the fear that inequalities might arise. It is important to consider whether this has been the case in the aftermath of the credit crisis. Again, measurement is often lacking, but some trends can be identified in various national sources. Wage inequality either remained stable or increased, thereby postponing or reversing any trend towards more wage equality that may be discernible in the long run.

One indicator is the ‘d9/d1 ratio’, that is, the ratio between the income at 10% from the highest value in the total income distribution and the income at 10% from the lowest value in the distribution of wage earners.

In Belgium, the d9/d1 ratio was 2.3 in 2009, which is not fundamentally different from the years before. The same stable inequality appears in France at a slightly higher level of about 2.9 and in Luxembourg at around 4.2. In Italy, wage dispersion, as measured by the Gini coefficient (indicating the variation in the distribution of incomes from a perfectly equal distribution), remains stable at a level of 25%. In this case, however, there is information on wage inequality within occupational groups; recently there has been decreasing wage dispersion among managers and clerks and increasing wage dispersion among blue-collar workers.

In Germany, the d9/d1 ratio was stable at around 3.3 until 2008 but is estimated to have increased lately. In Estonia, no figures after 2008 have been published but there seems to be a trend of rising wage inequality due to more people returning to education and an increasing percentage of low-paid workers.

In Ireland, more recent data are available which show an increase in the Gini coefficient from 30.7 in 2008 to 33.9 in 2010, and a similar rise in the income quintile share ratio, that is, the ratio between the total income received by the 20% of the population with the highest income and that received by the 20% of the population with the lowest income (also known as S80/S20).

In Latvia, even though dispersion is hampered by downward pressure on wages because of the crisis, the d9/d1 ratio in the public sector rose from 8.7 to 10.7 between 2008 and 2010, one explanation being an increasing share of low-paid work. A similar trend is found in Lithuania. An expansion of low-paid work in Poland is a trend that preceded the crisis but which continues to push inequality. In Portugal, however, the rising inequality is said to stem from a sharp rise in the income of the upper quintile.

In summary, there is stable inequality in most of the ‘old’ European countries for which data are available and increasing wage inequality in newer Member States and Germany. However, this conclusion should be treated with caution as it is heavily constrained due to the lack of information.

Explaining the trends

Company-level evidence for cost-cutting strategies

Every aggregate figure is the result of decisions made by micro-level units. To obtain an insight into the company-level decision-making process, the ECB’s Wage Dynamics Network undertook a company survey in 2008–2009 addressed at managers. Although mainly based on pre-crisis data, the survey’s findings indicate that when companies cut costs in reaction to an economic shock, cutting wages is an exceptional practice. Even then, the cuts will mainly affect the variable part of wages (for example, compensation for overtime or a change in shift assignments) and the reduction or elimination of certain wage components outside laws or collective agreements (for example, bonus payments and fringe benefits). More often, labour costs are reduced through changes in employment levels. Stopping recruitment overcomes the issue of training entrants and cutting temporary employment can be a response to lower demand. Other options are to increase prices, decrease margins or cut non-labour costs.

For a limited number of countries, the survey was repeated in 2009, taking into account the first phase of the crisis. Only 1.2% of the managers surveyed in the nine countries and three sectors covered in this study (manufacturing, trade and market services) suggested that cutting base wages was a common strategy used during the Great Recession of 2008–2009 (Table 9).

Table 9: Main cost-cutting strategy adopted by companies, 2009 (%)
 

Base wages

Flexible wages

Permanent employment

Temporary employment

Hours worked

Non-labour costs

Austria

0.3

12.2

12.2

11.1

36.2

28

Belgium

0.9

3.1

16.8

29.6

24.9

24.6

Czech Republic

0.0

10.4

27.9

16.4

5.3

40.1

Estonia

14.3

25.1

24.2

3.7

9.3

23.5

France

0.1

9.9

17.1

33.9

12.4

26.2

Italy

1.3

8.9

16.6

21.1

18.4

33.7

Netherlands

1.4

5.0

8.1

40.5

6.2

38.8

Poland

1.9

15.9

16.7

9.1

7.6

48.7

Spain

1.0

5.5

23.2

41.6

5.9

22.8

Weak demand shock

0.8

9.5

13.2

21.6

11.6

42.5

Strong demand/ weak credit shock

1.6

11.9

17.5

29.8

16.1

22.6

Strong demand/ strong credit shock

2.4

7.0

31.2

24.0

14.6

20.8

Total

1.2

9.8

16.9

24.3

13.6

33.9

Source: Fabiani et al (2010, Table 7)

In their analysis (436Kb PDF) of the WDN survey data, Fabiani et al (2010) found important heterogeneity in the behaviour of European companies across different countries depending on their labour market institutions (employment protection legislation and union bargaining power). Companies in countries where employment protection legislation is more stringent have tended to adjust the number of temporary workers and their hours, rather than the number of workers with permanent contracts. Nevertheless, the latter strategy was found in companies reporting a strong demand shock (lower part of Table 9). Union bargaining power, as measured by the share of workers covered by collective agreements, was found to be positively correlated with variations in employment (hours worked and temporary employment as a flexible instrument) and negatively correlated with wage adjustments (base and flexible wages). The low collective agreement coverage in Estonia partly explains why this country deviates from the trend mentioned above by having a high rate of base wage cuts.

Previous research by Babecký et al (2010) on the data obtained from the 2007–2008 wave of the survey investigated the reasons why companies do not opt for a strategy of wage cuts and put forward the following arguments.

  • Labour regulation or collective agreements prevent wages from being cut.
  • A strategy of wage cuts would reduce employees’ effort or have a negative impact on employees’ morale, resulting in less output or poorer service.
  • It would damage the company’s reputation as an employer, making it more difficult to hire workers in the future.
  • The imposition of a wage cut might make the most productive employees leave the company.
  • It would increase the number of employees leaving, increasing the cost of hiring and training new workers.
  • It would create difficulties in attracting new workers.
  • Workers dislike unpredictable reductions in income. Therefore, workers and companies reach an implicit contract that wages will neither fall in recessions nor rise in expansions.
  • Workers compare their wages to those of similarly qualified workers in other companies in the same market.

The results presented in Tables 10 and 11 show that companies avoid cutting wages in the first place because it would reduce effort and the best employees may leave. Labour regulation also has some impact.

  • Regulation is important for companies that employ a higher proportion of blue-collar and low-skilled white-collar workers. For these companies, arguments related to effort seem irrelevant.
  • Companies employing a higher proportion of temporary workers pay more attention to turnover arguments rather than regulation.
  • Companies with high union coverage and any type of agreements pay more attention to labour regulation.
  • Companies with high union coverage and outside agreements pay more attention to reputation and implicit contracts.
Table 10: Reasons for not applying wage cuts by country (% of total) 
  AT BE CZ EL EE ES FR HU IE IT LT NL PL PT
Regulation/ agreement

80

89

58

82

62

75

93

44

39

91

51

68

36

82

Reduced effort/ morale

93

92

91

95

97

93

75

85

87

88

91

80

76

91

Reputation

66

58

71

53

89

79

46

56

69

60

73

66

62

61

Best employees leave

86

84

97

82

98

92

73

72

83

92

98

79

91

88

Hiring/ training costs

78

69

89

43

96

76

57

48

59

88

95

64

69

59

Hiring difficulty

50

75

84

72

92

81

62

46

72

73

87

81

79

60

Implicit contracts

47

84

49

26

67

80

76

81

77

35

70

80

74

88

Employees compare

73

72

79

53

90

81

53

75

78

79

90

71

54

69

Note: Proportion of companies seeing the argument as ‘relevant’ or ‘very relevant’. See Annex 2 for a full list of country codes.

Source: Babecký et al (2010)

Table 11: Reasons for not applying wage cuts by area (% of total)
  Euro zone Non-euro zone All countries
Regulation/agreement

85

43

74

Reduced effort/morale

87

82

86

Reputation

58

64

60

Best employees leave

84

90

86

Hiring/training costs

69

72

70

Hiring difficulty

70

76

72

Implicit contracts

55

70

59

Employees compare

68

65

67

Note: Proportion of companies seeing the argument as ‘relevant’ or ‘very relevant’.

Source: Babecký et al (2010)

Employment shifts

As described above, the first reaction to the crisis was to cut labour costs through employment and only to a far lesser extent through wages. However, an artificial effect means that both terms are related in average data. National wage evolutions are most commonly presented through country averages, as done above both in the general discussion and in terms of key background variables. The analysis has looked at what way these trends show elements of wage deceleration, freezes or cuts.

Yet changes in these averages may not be exactly what happens at the individual level. The aim here is to compare the average wage of the entire workforce for a particular year with the average wage of the workforce for a consecutive year. Nevertheless the workforce may have changed during this period of time. This is shown below with a simple, fictitious example (Table 12). A group of three people earn €2, €4 and €6 per hour, respectively. There are two scenarios for the next year. In scenario 1, the person with the lowest paid job loses their job and the other people’s wages are frozen. Nobody has a higher wage in the second year, yet the average increases from €4 to €5. In scenario 2, the wages of persons 1 and 2 are cut and person 3, who has the highest wage, receives a pay rise. The average is unchanged, although the majority of people have been confronted with a wage cut.

Table 12: Fictitious example of composition effects on the average wage trend
 

Start wage

Scenario 1

Scenario 2

Person 1

€2

Unemployed

€1

Person 2

€4

€4

€3

Person 3

€6

€6

€8

Average wage

€4

€5

€4

This example illustrates the limitations of an aggregate economic analysis. This study shows that employment shocks are the main consequence in most countries and sectors. Thus it is necessary to consider the effects that can be expected from these shifts in employment.

In the 2011 Eurofound report Shifts in the job structure in Europe during the Great Recession, composition changes are measured through a so-called job distribution. This is an ordering of occupations in five groups according to the wage quintile to which the occupations belong in the base year. The next step is to examine the changes in employment levels for these clusters of occupations.

Five patterns of changes in the job distribution are distinguished during the credit crisis:

  • Polarisation (stable levels in the middle and job growth in the best and worst-paid jobs): Bulgaria, Cyprus, France, Greece, Ireland, Latvia, Portugal, Slovenia, Spain and the UK;
  • Hybrid polarisation/upgrading (as for polarisation, but with more growth in the best-paid jobs): Austria, Belgium, Finland, the Netherlands and Poland;
  • Upgrading (no particular trend, except from growth in the best-paid jobs): Germany, Luxembourg, Slovakia and Sweden;
  • Hybrid upgrading/growing middle (growth in the middle and best paid jobs): Romania;
  • Downgrading (growth in the worst-paid jobs): the Czech Republic, Denmark, Hungary, Italy and Lithuania;
  • Not defined: Estonia, Malta and Norway (not included in the study).

The question arises as to how well the shifts in employment fit with the evolution of wages. Figure 9 shows the clustering of the five patterns of change along the dimensions of nominal and real growth during the Great Recession, 2009–2010. As might be expected, the wage growth figures of the downgrading pattern of change are partially situated below the zero line, and on average, lower than all the other types. The upgrading and the hybrid upgrading/polarisation patterns of change generally stay above a real wage freeze, but the effect is not universal and not definitive. In addition, the polarisation pattern of change has the largest variation and so, in order to interpret the wage evolution in terms of employment shifts, there should be differently skewed income distributions over all occupations.

The relationship between the different patterns of growth and wage growth may not cover the whole story but, in general, the patterns of job growth seem to cause or, at least, reinforce the wage evolution.

Figure 9: Nominal and real growth by pattern of job change during 2009–2010

tn1203015s.tmp08.jpg

Note: A prime symbol after the country code indicates 2010 data. See Annex 2 for a full list of country codes.

Source: Wages and employment data from the national contributions, patterns from the 2011 Eurofound report Shifts in the job structure in Europe during the Great Recession

Compression of the wage cushion

Whenever wages are adjusted, the target is almost always the flexible part of wages. In addition, the presence and power of a trade union(s) seems to act as a restraint on wage cuts.

Although strictly speaking, wage drift is an indicator of the degree of collective bargaining coordination, it is also an indirect indicator of wage flexibility. Wage drift is defined as the difference between actual wages and collectively agreed wages that define the minimum wage by sector or occupation. Actual wages will vary through flexible wages and composition effects as discussed above, while the collectively agreed wages reflect what happens in a fixed set of sectors or occupations. Unexpected business trend movements may also drive wage drift. A decreasing wage drift can thus be used as proof of the compression of this ‘flexible’ part of the wages or ‘wage cushion’ in periods of crisis.

Only a few of the country reports address this issue, mainly because of the poor availability of collectively agreed wage information or a rather limited practice of collective bargaining on pay (see tables in Annex 1).

For example, Belgium is traditionally a highly unionised country and does monitor the evolution of collectively agreed wages. Looking at wage trends during the crisis, a decrease in wage drift can be seen, indicating that a safety margin for wages was built up in an upward business cycle to be compressed during a downturn. Together with composition shifts, this caused wage drift to be negative at the aggregate level in 2008 and 2009, and limited in 2010. Germany follows the same line of reasoning, but anticipates a steep increase in wage drift in 2011. In the Czech Republic, a different trend is found: collective bargaining safeguards wage growth in those companies where negotiation takes place. Here wage drift did not decrease during the crisis, although this is an artificial effect as fewer agreements were reached because of the crisis.

Conclusions on trends

Based on the provided national statistics and in comparison with the AMECO data, the following conclusions can be drawn on European wage trends in the first phase of the current crisis.

  • The overall picture shows wage deceleration and, to a lesser extent, freezing of wages. Wage cuts seem to have been rather scarce.
  • The wage trends were particularly negative in the Baltic States.
  • Mixed results can be reported for the affected groups. In some countries the crisis led to widening wage gaps for vulnerable groups, while in others, the opposite trend can be observed. Similarly, wage inequality did not increase universally.
  • The sectoral data for 2009–2010 confirm that the crisis had a stronger impact on employment than on wages.
  • In this first phase of the crisis, public sector employment was relatively less affected by these shocks.

These conclusions are based on looking at the trends in aggregate averages. Taking the employment effects of the crisis into account, still more positive wage growth in these country averages can be related to an ‘upgrading’ of the workforce and a more negative wage growth to a ‘downgrading’ of the workforce. These composition effects reinforce or counterbalance the trend of wage growth slowdown.

Other results on wage drift (reported in the national contributions) show that a margin of flexible wages seems to be used strategically in times of crisis to save on labour costs without touching basic wages. This confirms the finding that, most of the time, wage cuts are only the third option when dealing with an economic shock. Cutting (mainly non-core) employment and reducing working hours or overtime are the first types of measures to be implemented, though again the chosen strategy depends on economic sector, occupational group and importance of collective bargaining.


Link with other working conditions

This section reviews the relationship between trends in wages and quality of work. It considers whether a decline in wages coincides with deteriorating quality of work at the individual level, or whether wage cuts are an uncorrelated measure used in times of crisis.

The national contributions to this report concluded that formal studies or statistically significant findings showing a relationship between wage changes and other working conditions are rare. Most of the limited number of studies presented in the various national reports focus on the relationship between pay level and job satisfaction, a topic that has been the subject of a whole range of surveys.

However, the European Working Conditions Survey (EWCS) contains some questions that may shed some light on this relationship. These findings are reported in the second half of this section. As highlighted in the previous section, companies reacted to the crisis mainly with changes in employment (job cuts or working time reduction) and what can be generally called wage moderation. Strong intervention in wages was not a dominant practice. Data from the fifth EWCS in 2010 allow for a more detailed investigation of the types of employees who are more or less confronted with the challenge of income reductions and increased job insecurity (or both).

Nationally reported findings

Factors affecting job satisfaction

A Eurofound study from 2006 on Mesuring job satisfaction in surveys demonstrated that a myriad of factors can influence job satisfaction.

Some theorists (for example, Sousa-Poza and Sousa-Poza, 2000) postulate that job satisfaction depends on the balance between work–role inputs (such as education, working time and effort) and work–role outputs (wages, fringe benefits, status, working conditions and intrinsic aspects of the job). If the work–role outputs (‘pleasures’) increase relative to work–role inputs (‘pains’), then job satisfaction will increase.

Others (for example, Rose, 2001) regard job satisfaction as a bi-dimensional concept consisting of intrinsic and extrinsic satisfaction dimensions. Intrinsic sources of satisfaction are related to individual characteristics such as the ability to use initiative, relations with supervisors and the work that the person actually performs, whereas extrinsic sources of satisfaction are situational and depend on the environment (for example, pay, promotion and job security) and are essentially the financial and other material rewards or advantages of a job. Common to these views is the fact that pay may not be the most important but is certainly one of the factors explaining job satisfaction.

Nonetheless, Herzberg’s motivation–hygiene theory suggests a poor relationship between wages and job satisfaction (Herzberg, 2003): being underpaid results in dissatisfaction, while being overpaid does not result necessarily in more satisfaction. Herzberg therefore considered wages as a ‘hygiene motivational factor’.

Criticism of Herzberg’s theory emerged initially because it was based on a survey among employees with higher levels of educational attainment. A similar survey among unskilled workers (let alone the working poor) yielded other results, as surveys cited in several of the country reports clearly demonstrate. A 2004 study on job satisfaction in Finland (202Kb PDF) by the Research Institute of the Finnish Economy (ETLA) examined determinants of job satisfaction using Finnish data from the European Community Household Panel (ECHP). According to these results, higher wages increased job satisfaction even when controlling for a host of other factors. The ETLA study also found that a loss of income due to involuntarily having to work part time had a strong negative effect on job satisfaction.

The International Social Survey Programme (ISSP) survey on Quality of Working Life was carried out in 2006 by two Czech research centres on a sample of 2,043 employed people aged over 15 years (CZ0901019D). Data from the survey imply that at least two-thirds of respondents regarded their income level as ‘definitely important’ and a further 28.7% of respondents as ‘rather important’. It can therefore be concluded that income level seems to have a significant direct influence on job satisfaction in the Czech Republic.

A completely opposite situation is found in Greece, one of the Member States hit particularly badly by the economic crisis. A recent survey by Adecco Greece found that, in the middle of the crisis, the feeling of security and stability constituted the most important parameter desired from work for 32% of the participants while only 4% stated that pay was an important parameter (GR1109019I). The overall majority (93%) of the respondents were willing to show flexibility on pay during the crisis provided other factors such as job content, employment relationship and career perspective were satisfactory.

In Austria, information on wages and job satisfaction during the crisis is available through the Work Climate Index (Arbeitsklima Index), a measure of economic and social change from the point of view of employees compiled regularly by the Chamber of Labour (AK). The index explores assessments concerned with society, companies, work and future expectations, and thus records subjective perspectives. It is representative of all Austrian employees and is based on quarterly surveys. The index shows that Austrian employees were less satisfied with their income in 2010 than they were in 2000: 65% reported being ‘(very) satisfied’ with their income in 2000 while 58% were ‘(very) satisfied’ in 2010. However, between 2006 and 2010 the share of employees being ‘(very) satisfied’ rose from 59% in 2006 to 61% in 2007, falling to 59% in 2008 before rising again to 60% in 2009. As noted in the September 2010 issue of the Arbeitsklima Index newsletter (in German), these figures indicate that the crisis did not have a negative influence on employees’ satisfaction with their income. However, the crisis did have an influence on employees’ optimism about future economic developments. The share of employees looking toward to the future with optimism amounted to 83% in 2007, decreased to 72% in 2008 and was only 61% in 2010. This somewhat pessimistic view is confirmed by the responses described in the March 2011 issue of the Arbeitsklima Index newsletter (in German) to a question about how easy it would be to find a new and acceptable job: 52% in 2008 believed that it would be ‘(very) easy’ to find a new job but this share had fallen to only 45% at the end of 2010.

The Bulgarian Working Climate Index study, a national representative survey carried out by Institute for Social and Trade Union Research (ISTUR)/the Confederation of Independent Trade Unions in Bulgaria (CITUB) in November–December 2010, revealed a moderately strong relationship between the dynamics of wages in the last year and job security: 59.5% of respondents experienced no pay increase in 2010 compared to 2009, whereas 23.4% received lower pay and 17.1% higher pay. Of the respondents with increased wages, 59.5% were positive about keeping their job, 22.9% had negative expectations and 17.6% were not sure. At the opposite end, only 41.4% of those declaring a reduction in wages were optimistic about keeping their job, 43.2% declared their job was at risk and 15.4% were not sure.

In Spain, some interesting information on the issue refers to the evolution of the general average satisfaction of Spanish workers with their pay conditions. According to data from the annual Spanish survey on quality of life at work (in Spanish), there is an increasing dissatisfaction with pay among Spanish workers: on a scale from one to 10, the average level of satisfaction was 6.2 in 2006 and 5.8 in 2010. In particular, 2010 recorded a remarkable downward trend in comparison with the previous two years. This decreasing level of satisfaction applied to both men and women; it extended to all professional situations but was especially relevant among employers (both with and without employees). Employees in the private sector also had a more negative perception of their salary than those in the public sector. It is difficult to establish a link between decreasing levels of satisfaction with wages and the economic crisis. From a tentative, common sense perspective, it could be argued that workers (especially salaried ones) currently have two opposing perceptions about their salary. On one hand, they are increasingly dissatisfied about the wage moderation experienced in the last two years but, on the other hand, (many) workers may feel that, compared with unemployed people, they still have a job and receive an associated wage (though not in all cases) and so there is limited scope to complain.

In France, there is no systematic analysis of the relationship between job satisfaction and the evolution of wages and salaries. A survey on working conditions (in French) by a polling company conducted among 903 French employees in March 2009 showed that a majority (89%) were happy with their working conditions despite the upcoming crisis. Some 68% replied that their working conditions had not deteriorated over the previous six months. Only around a quarter of the surveyed employees (24% of the managers and 21% of the employees) reported a shift in the opposite direction – in particular an increase in stress levels and work pressure, and pay freezes, and to a lesser extent, a decline in the social climate at work (rising tensions between colleagues) and customer relations. A quarter (25%) feared a worsening of their working conditions in the months to come, the most prominent fear being to be laid off (61%).

Dutch workers have been hardly touched by the economic crisis, with wages still showing some upward movement and employment staying at the high levels seen before the crisis. According to statistics (in Dutch) from the Netherlands Institute for Social Research (SCP), working conditions were not an issue in the studied period. The percentage of employees satisfied with their working conditions was 77% in 2010 compared with 73% in 2007. Experiences of work-related stress, the number of accidents at work and physical demands remained quite unaffected by the economic crisis.

A report on employee attitudes to pay in 2011 by the Chartered Institute for Personnel and Development (CIPD) in the UK found that the net satisfaction score among those who received a pay rise in 2011 was +56 (net scores refer to the proportion of people agreeing with a statement minus those disagreeing). The net satisfaction score among those who received a pay freeze was -41, down from -26 in the 2010 survey and -23 in the 2009 survey. The main reasons given for dissatisfaction were that:

  • ‘pay did not keep pace with increases in the cost of living’ (63%, up from 53% in 2010);
  • ‘pay did not reflect how well I performed at work’ (22%);
  • ‘pay is below what I could get elsewhere for doing the same job’ (18%).

Not only employees but also employers dread a negative impact of pay limitations on the motivation and loyalty of employees. A 2011 working paper on how firms adjust in a crisis (417Kb PDF) by the Central Bank of Luxembourg based on WDN surveys indirectly makes reference to the relationship between wage changes and employee satisfaction. The survey asked respondents to identify obstacles to possible wage cuts. Some of the major obstacles identified by companies were the fear of reducing workers’ level of effort and morale (different from the notion of job satisfaction), hence lowering the quantity and quality of output.

Pay variations and the number of hours worked

As seen above, working time reduction is one of the major strategies used by employers to cope with the crisis. Several of the national contributions reported a link between working time and wages.

Data from the fifth EWCS report show that 60% of the individuals in Spain who had suffered a reduction in working time in the course of 2010 had to cope with a corresponding salary decrease; for 27% of them salary was maintained and the remaining 13% even managed to acquire a pay increase notwithstanding the fact that they actually worked less.

The same phenomenon was observed in Slovakia, a country where 26% of the workers enjoyed a growing income but 16% had to cope with declining pay. Data from the EWCS on the number of hours worked in that country suggest that, in 2010, the wages decline in Slovakia correlated strongly with the decrease in the number of hours worked.

The Bulgarian Working Climate Index reveals a relationship between the dynamics of wages and satisfaction with the daily hours of work. In the group of respondents who experienced a wage increase in 2010, 85.7% were satisfied with their working time arrangements, 7.1% were negative on the issue and 7.2% undecided. The shares obtained for the cohort of workers whose pay decreased in 2010 were 69.9%, 19.1% and 11.0%, respectively.

As such, these survey micro-data confirm the general observation that the crisis-related changes in wages and working time are very often strongly related.

Pay variations and job stability

Some national contributions reported specific insights into how the wage effects related to aspects of job stability. Conflicting findings were presented in this regard.

In some countries, the general trend of lower job security and lower (relative) earnings since the crisis leads to higher job loyalty (keeping what one has) or to higher job mobility (changing what one has for something new).

Although the research project ‘Our society’ of the Public Opinion Research Centre (CVVM) found that job satisfaction in the Czech Republic has declined in recent years, the growing insecurity in the labour market was reflected in employees’ unwillingness to leave their current job. While the share of workers who felt significantly undervalued in terms of pay and were considering a job change grew significantly between 2006 and 2008, their number fell substantially in the period to 2011. The only group of employees who were considering leaving their job more often than in previous times were the ones who considered their remuneration adequate. The national report assumes that such people are primarily in higher and specialised posts and do not feel threatened by the current labour market situation (Figure 10).

Figure 10: Views on existing wage by those considering leaving their job in the Czech Republic

Description: TN1203015S_Figure10

Source: Czech national report

In Norway, the financial crisis and economic uncertainty has led more than a third of employees to be more loyal to their employers than ever before. This finding is revealed in the 2010 edition of a global workplace survey called the Kelly Global Workforce Index (316Kb PDF) which obtained the views of around 134,000 employees worldwide, including about 2,000 in Norway and was conducted between October 2009 and the end of January 2010. The survey revealed that 35% of Norwegian workers believed the financial crisis had made them more loyal to their employers, 4% responded that they were less loyal and 61% said the crisis had not made any difference. The impact of economic uncertainty was greatest among older workers (aged 48–65 years), among whom 38% said they had become more loyal compared with 36% of those aged 18–19 years and 33% of those aged 30–47 years. Surprisingly, pay was not the main reason for employees feeling more loyal than before. In response to the question, ‘what factor would make you more committed or engaged in work,’ only 14% stated that higher wages or better benefits were critical, with 35% % giving ‘more interesting / challenging work’ and 31% ‘more or improved job training’ as the two main reasons.

In Ireland, the economic hardships had a completely opposite effect. The Mercer ‘What’s Working’ survey, conducted between the final quarter of 2010 and the second quarter of 2011, attempted to measure the impact of the current crisis on company performance and employee morale. The survey found that far fewer Irish workers today:

  • believe they are paid fairly given their performance and contributions to the organisation (53%, down from 62% in 2004);
  • understand how their pay is determined (67%, down from 75% in 2004);
  • believe the pay in their organisation is as good as, or better than, that of other organisations in their industry (49%, down from 54%).

Slightly more than one in three (35%) of Irish workers was reported to be seriously considering leaving their organisation in 2011, up sharply from 22% in 2004.

In Malta, this coincides with the general feeling of employers who believe that the implementation of wage cuts would damage their company as they would risk losing their most productive workers to their competitors; for details, see the Wage dynamics report Malta (723Kb PDF) commissioned by the Central Bank of Malta.

Findings from the EWCS

Micro-level data from the fifth EWCS in 2010 makes it possible to investigate different types of consequences depending on the impact of the crisis on job security and wages. The comparative exercise below shows which scenario is most common in which countries. This is followed by a discussion of vulnerable groups and their relative risk of facing a wage cut, job insecurity or both simultaneously at aggregate, EU27 level.

Methodology

Dependent variable

For the self-declared evolution of wages and job security, the consequences of the crisis have been operationalised into four basic types of consequences. The scheme set out in Table 13 shows these possibilities. In the upper right square, the only measure taken is a reduction of pay (W). In the upper left square, no changes happened during the crisis (O). In the lower left square, only job security decreased (L) and, finally, in the lower right square, there is a combination of pay reduction and reduced job security (LW).

Table 13: Matrix of crisis consequences
  Stable or improved pay Reduced pay
Stable or improved job security

O

W

Reduced job security

L

LW

The question about changes in pay in the EWCS is as follows: ‘If you compare your current situation with that of January 2009, have you experienced a change in your salary or income?’. The possible answers are: a decrease, an increase or no change. The latter two options were interpreted for this analysis as stable or improved pay, while the first option was taken to refer to pay reduction. Note that this can reflect a reduction in the basic pay, other components of pay, a reduction in working hours, or combinations of these items.

The EWCS examined job security with a Likert type question (‘strongly agree’ to ‘strongly disagree’) based on the statement: ‘I might lose my job in the next six months’. Respondents who ‘agreed’ or ‘strongly agreed’ were considered to have reduced job security while those who ‘disagreed’, ‘strongly disagreed’ or expressed neither agreement nor disagreement were considered to have stable or improved job security.

The appropriate method by which to analyse categorical dependent variables is multinomial logistic regression. This technique shows the relative risk of ending up in one category rather than in a reference category. In this case, the reference category is stable or improved job security and pay (type O). Three equations are obtained for the other categories. The value of the intercept in each equation shows the odds for that category. For example, if the intercept for ‘reduced pay only’ is 2, it means that it is twice as likely to experience a pay reduction (and no increase in job insecurity) than to experience a stable or improved situation. Further, if the equation for the situation of ‘reduced pay and reduced job security’ (type LW) has an intercept of 1, it is equally likely to end up in this situation as it is to experience no worsening at all. By simple logic, the reduced pay only measure will be twice as likely as this category too.

In order to be able to answer the first research question, the equation is arranged so that the country dummies (variables that take the value 0 or 1 to indicate the absence or presence of each country) each capture the country-specific intercept. Adding independent variables to the equation, allows for adjustments of this base risk. All these variables have a reference value of either 0 (for the metric variables) or a reference category (for the non-metric variables, that is, categorical). These values point to an outcome that is equally likely as the (country) intercept. In case of an increase with one point or a respondent belonging to another category, the regression coefficient changes this likelihood. Note that, despite having country intercepts, the regression coefficients are based on the average intercept, which may be slightly biased from a mixed model point of view.

Independent variables

Two sets of independent variables are used.

One refers to vulnerable groups and includes age (three categories), country of origin (‘Were you and both of your parents born in this country?’), income (10 deciles) and gender.

The other group consists of two control variables:

  • sector divisions:
    • primary sector (that is, agriculture, fishery and mining) and construction;
    • manufacturing and energy resource exploitation;
    • services and entertainment;
    • public administration, education and healthcare (PEH);
  • Erikson–Goldthorpe–Portocarero (EGP) classification (excluding the self-employed), which clusters occupation by skill requirements, autonomy and degree of authority (Erikson et al, 1979).

Results

The presence of one of the types of crisis consequences may be due to country particularities or worker characteristics. The focus here is the risks for vulnerable groups. The analysis consists of three steps.

  1. Check the country intercepts.
  2. Examine the dimensions on which the vulnerable groups are situated (age, gender, origin and income).
  3. Decide whether industry and occupation are explaining variables.

Figure 11 shows the distribution of crisis consequences in the EU27 and Norway from the EWCS data. The modal category is clearly the no-change type (O). The ordering by share of this type appears to largely follow the rank of gross domestic product (GDP) per capita. As a result, this category does not cover a majority of respondents in four Member States (Estonia, Ireland, Latvia and Lithuania); more specifically type W (wage cut only) appears more frequently in Latvia and Lithuania. With the exception of the Baltic States, type LW is least common. These country effects largely remain even after controlling for the presence of vulnerable groups (see tables in Annex 1).

Figure 11: Labour conditions consequences of the crisis (employees only), 2010

Description: TN1203015S_Figure11

Note: See Table 13 for an explanation of categories.

Source: EWCS 2010, own calculations

In Tables 14–16, which present the regression results without the country dummies, the age effects present a striking distribution:

  • older workers are more frequently faced with wage cuts (type W);
  • younger workers risk job insecurity (type L);
  • type LW is found equally in all three age groups so, other things being equal, each group has a risk equal to the country mean.

These highly significant results do not change when the industry and occupation variables are introduced.

Another remarkable finding is the gender effect. It seems that men form the vulnerable group, with a greater chance of being subject to each of the three impact types. Controlling for industry and occupation, however, type W risk becomes insignificant and L and LW risks move closer to one (that is, equality between men and women).

With respect to origin, non-natives have a slightly greater chance of being subject to wage cuts (type W) and a much greater chance of facing job insecurity (type L) than natives. These effects stay robust after controlling for industry and occupation. No differences are found for type LW.

Table 14: Multinomial regression of crisis consequences on vulnerable groups and sector and occupation variables – type W

Model step

2

 

3

 
Age:        
Under 34

1.00

 

1.00

 
35–49

1.42

***

1.41

***

50+

1.60

***

1.60

***

Gender:        
Men

1.00

 

1.00

 
Women

0.79

***

0.91

 
Origin:        
Native

1.00

 

1.00

 
Non-native

1.17

*

1.17

*

EGP:        
Higher controllers    

1.10

 
Lower controllers    

1.15

 
Routine non-manual work    

1.00

 
Manual work supervisor    

1.53

*

Skilled manual work    

1.17

 
Semi-unskilled manual work    

1.24

**

Farm worker    

1.15

 
Sector        
Primary sector + construction    

1.28

**

Manufacturing + energy resource exploitation    

1.19

*

Services + entertainment    

1.00

 
PEH    

0.95

 
Income:        
Decile 1

2.53

***

2.60

***

Decile 2

1.78

***

1.84

***

Decile 3

1.55

***

1.53

***

Decile 4

1.30

*

1.33

*

Decile 5

1.00

 

1.00

 
Decile 6

1.12

 

1.11

 
Decile 7

0.82

 

0.86

 
Decile 8

0.93

 

0.97

 
Decile 9

0.89

 

0.91

 
Decile 10

0.77

*

0.81

 

Notes: * p < 0.05, ** p < 0.01, *** p < 0.001

Controlled for country effects (see Table A1 in Annex 1).

Source: EWCS 2010, own calculations

Table 15: Multinomial regression of crisis consequences on vulnerable groups and sector and occupation variables – type L

Model step

2

 

3

 
Age:        
Under 34

1.00

 

1.00

 
35–49

0.86

***

0.86

***

50+

0.85

**

0.86

**

Gender:        
Men

1.00

 

1.00

 
Women

0.75

***

0.88

**

Origin:        
Native

1.00

 

1.00

 
Non-native

1.41

***

1.40

***

EGP:        
Higher controllers    

0.99

 
Lower controllers    

0.98

 
Routine non-manual work    

1.00

 
Manual work supervisor    

0.70

*

Skilled manual work    

0.92

 
Semi-unskilled manual work    

1.15

*

Farm worker    

1.08

 
Sector        
Primary sector + construction    

1.38

***

Manufacturing + energy resource exploitation    

1.27

***

Services + entertainment    

1.00

 
PEH    

0.72

***

Income:        
Decile 1

1.41

***

1.44

***

Decile 2

1.39

***

1.35

***

Decile 3

1.19

*

1.18

*

Decile 4

1.13

 

1.10

 
Decile 5

1.00

 

1.00

 
Decile 6

0.77

***

0.79

**

Decile 7

0.61

***

0.64

***

Decile 8

0.64

***

0.67

***

Decile 9

0.52

***

0.56

***

Decile 10

0.51

***

0.56

***

Notes: * p < 0.05, ** p < 0.01, *** p < 0.001

Controlled for country effects (see Table A2 in Annex 1).

Source: EWCS 2010, own calculations

Table 16: Multinomial regression of crisis consequences on vulnerable groups and sector and occupation variables – type LW

Model step

2

 

3

 
Age:        
Under 34

1.00

 

1.00

 
35–49

1.10

 

1.07

 
50+

1.04

 

1.05

 
Gender:        
Men

1.00

 

1.00

 
Women

0.57

***

0.75

***

Origin:        
Native

1.00

 

1.00

 
Non-native

1.22

 

1.18

 
EGP:        
Higher controllers    

0.78

 
Lower controllers    

1.05

 
Routine non-manual work    

1.00

 
Manual work supervisor    

1.81

*

Skilled manual work    

1.29

*

Semi-unskilled manual work    

1.39

**

Farm worker    

1.04

 
Sector        
Primary sector + construction    

1.49

***

Manufacturing + energy resource exploitation    

1.57

***

Services + entertainment    

1.00

 
PEH    

0.61

***

Income:        
Decile 1

3.09

***

3.27

***

Decile 2

2.66

***

2.68

***

Decile 3

2.02

***

2.03

***

Decile 4

1.16

 

1.16

 
Decile 5

1.00

 

1.00

 
Decile 6

1.02

 

1.03

 
Decile 7

0.70

*

0.76

 
Decile 8

0.62

**

0.70

*

Decile 9

0.34

***

0.36

***

Decile 10

0.35

***

0.46

***

Notes: * p < 0.05, ** p < 0.01, *** p < 0.001

Controlled for country effects (see Table A3 in Annex 1).

Source: EWCS 2010, own calculations

The income groups, divided by decile, are all compared with the median group.

For type W (Table 14), there is an almost linear effect below the median; the lower the income group, the higher the odds become, at an almost constant rate. The upper and lower deciles show another bump in the extreme directions. The effects below the median remain significant after introducing the control variables.

For type L (Table 15), the linear effect is found at both sides of the median group and on both sides this remains in the third step (that is, when occupation and sectors are considered).

Type LW shows again a polarised pattern (Table 16), but less explicit around the median and stronger towards the upper and lower deciles. This effect is not fundamentally altered when the control variables are introduced. This means that the inequalities are not merely a result of, for example, non-natives doing unskilled work or women working in the PEH sectors.

Looking at the control variables, type W can be found more often within the occupational groups of manual work supervisors and semi-unskilled manual workers, and in the primary sector (including construction) and the manufacturing industry (including the energy sector). It could be that the wage cushion explains this particular vulnerability, as it is easiest for these groups to set targets and to reduce shift and overtime work.

Type L follows almost the same pattern, but with less job insecurity for the manual work supervisors.

Finally, type LW reflects type W, but the differences are much more explicit and skilled manual workers also have a greater chance of suffering both a wage cut and increased job insecurity compared with the reference category of routine non-manual workers.

In conclusion, the choice between income and employment security in response to the crisis is not spread evenly throughout the working population in the European countries. It seems that richer countries suffered less than, for instance, many new Member States or Mediterranean economies. Within countries, additional inequalities can be found: lower-skilled manual labour and migrant workers were more hit by adverse effects.

The analysis also showed that:

  • men were more affected by the consequences of the crisis in terms of wages and job security than women;
  • the higher-level manual workers were confronted more by wage measures than employment insecurity.


Trade-offs within policy strategies

The final section of the report turns to social dialogue and policymaking.

National correspondents were asked to identify and describe examples of key company and/or sectoral arrangements where a trade-off between wages and other features of the employment relationship had been obtained, that is:

  • wage moderation;
  • wage freezes or wage cuts in exchange for improvements such as job security, other forms of rewards, working time reduction, more and better training.

The goal was to determine where and how the trade-off idea is part of the current crises bargaining. Certain questions were asked.

  • Are the wage concessions of the crisis period used to obtain (longer-term) improvement in other aspects of work quality?
  • Are there innovative approaches that deserve further attention by social partners in other European countries?

National correspondents were also asked to collect an overview of recently implemented policies aiming to support groups of workers who might have been particularly affected by the crisis and the recent negative wage trends. Examples of such policies include:

  • minimum wage policies;
  • funding training;
  • social benefits such as transport or housing support.

After all, the EU2020 strategy proclaims ‘inclusive growth’ (that is, ‘fostering a high-employment economy delivering economic, social and territorial cohesion’).

Trade-offs in bargaining?

The reported examples of trade-offs in bargaining are summarised below by level of social dialogue (national, sectoral and company).

National programmes

National correspondents reported a range of national actions that focus on coordinating wage moderation and organising in a regulated way derogations at company level to implement wage cuts. The (limited) trade-off comes in these cases from a procedural point of view. Most of the reported examples come not surprisingly from (west European) countries characterised by an organised and centralised system of wage bargaining.

Coordinated wage moderation

In some countries, in-company wage cuts are rare because wage moderations are already set at the national level by a general pay norm. In specific countries (for example, Spain), cross-sectoral consultation leads to general wage moderation measures so as to reinforce enterprises’ competitiveness and to avoid further employment destruction. Belgium is a case in point: pay norms result from cross-sectoral bargaining and are imposed by a general bargaining or even by law, and aim to stimulate competiveness and employment.

Other countries have a mechanism that is meant to keep pay levels in line with cost of living (automatic wage indexation), but have recently modified it (Italy) or are in the middle of a discussion on the system (for example, Belgium and Cyprus).

Working hours’ measures

Some Member States have implemented a sort of self-regulating system.

A good example is Finland, which has a tradition of paying out so-called holiday bonuses (typically 50% of daily salary per each day of holiday). As a result of the financial crisis, exchanging holiday bonuses for days of holiday has become more common, thus saving the employer money (actual figures are not available).

In Slovenia, an act was adopted promoting a significant reduction in working hours to counteract lay-offs; companies were eligible for subsidies granted for every employee included in the short-time working scheme.

General measures, but more rigid?

The main advantage of these national measures is the general, economy-wide impact. However, this national coordination of wage-setting leads to more rules and seems to prevent real wage cuts.

In France, for example, pay reductions can only be agreed on by the individual employee, and trade-offs between wages and other features of the employment contractual arrangements are logically uncommon. The three cases highlighted in the French national contribution all pertain to multinational companies belonging to the automotive sector.

Push for decentralisation

Some governments deliberately aim at decentralising the collective bargaining process (for example, the current cases of Italy and Spain) in the hope that sectoral or in-company negotiations will lead to working conditions that are more in line with the sector’s or the company’s financial capacity.

Derogation clauses

In other countries, the social partners introduce alternative solutions.

After joint consultation on the national level, the Romanian government adopted an ‘emergency ordinance’, providing enterprises with the option to suspend business temporarily without terminating the employment relationship with their staff and paying out a minimum base salary.

The Austrian electronics sector is another example. In 2009, the collective agreement included a system of hardship clauses (Konjunkturklauseln) that offered companies in extremely difficult economic situations a reduction in wage increase from 2.2% to 1.4%. To apply these hardship clauses, a company had to:

  • prove that turnover had fallen by at least 15% in the first quarter of 2009;
  • obtain works council and trade union approval.

According to the 2011 working paper ‘Collective bargaining and balanced recovery: The case of Austria (547Kb PDF)’ from the International Labour Office (ILO), a total of 60 companies encompassing more than a quarter of all employees in the electronics sector qualified for the clauses and were allowed to increase wages by only 1.4%.

Minimum wage

When discussing pay interventions at a national level, special attention should be given to minimum wage policies.

EU Member States have very different opinions on the issue. In some countries, the argument to implement a minimum wage has won attraction (Germany), while other countries have frozen it or limited increases to the increased cost of living (Belgium, the Netherlands, Romania). A number of countries have even reduced the level of the minimum wage in order to boost employment for those with a low level of educational attainment. However, some countries have increased it as a measure to counteract the ‘working poor’ phenomenon (see also the section further below on measures to support vulnerable groups).

Examples of national programmes

In order to make employment more financially attractive for low-wage earners and to increase the gap between wages and unemployment benefit, the fiscal and para-fiscal pressure on low wages has been further reduced in Belgium. For a minimum wage of €1,415 a month, employees are granted a 100% cut in their personal social security contributions. These contributions increase gradually thereafter to achieve a level equal to the regular rate of 13.07% for monthly wages in excess of €2,247. The Belgian government has also decided on a direct increase of €120 per year in the net wage to the minimum wage level of €1,415 per month. This increase is being implemented via a new tax credit equal to a fixed percentage of the actual reduction in personal social security contributions. The tax credit is steadily reduced until a monthly wage of €2,247 is achieved. With a view to achieving a direct net effect on monthly remuneration, an impact on the payroll tax is anticipated, starting in April 2011. A budget of €30 million in 2011 and €40 million in 2012 has been earmarked for the purpose.

Following a motion from the social partners (nationally representative trade unions and employers’ confederations) who, following negotiations, agreed to act to reduce the risk of job loss, the Romanian government adopted an emergency ordinance (No. 28/2009) which gave employers the possibility of suspending business temporarily without terminating the employment relationship with their workers. The ordinance permits a stoppage of operations for a maximum period of three months, with the minimum payment of 75% of the base salary, and a tax holiday for social insurance contributions and income tax for both employers and employees. This ordinance has produced effects for companies in the manufacturing, construction and Horeca sectors. The evidence so far suggests that this ordinance has contributed to the conservation of jobs in Romania, thereby easing the effects of the economic crisis.

Sectoral agreements

In many countries with a tradition of sectoral bargaining, sector agreements sometimes contain so-called employment security clauses. These stipulate that no worker can be made redundant for a certain period, or that a specific consultation procedure with the trade unions must be followed when lay-offs are intended. Many of these agreements have been prolonged in the crisis period and are part of the broader sectoral agreement.

A total of 22 specific sectoral agreements were reported by the national correspondents in 13 countries (Belgium, Bulgaria, Denmark, Germany, Ireland, Italy, Lithuania, Luxembourg, the Netherlands, Portugal, Slovakia, Slovenia and Spain). All the agreements sought to significantly reduce labour costs and concerned public administration (nine cases), manufacturing and construction (including metalwork, brewing, energy and chemicals, eight cases) and the services sector (including banking and Horeca, with five cases).

All but one of these sectoral agreements aimed to reduce salary costs, mostly via temporary pay freezes or pay cuts, but in three cases, by temporarily reducing working hours, which represented an actual loss to the employees affected. Only a collective agreement in the Latvian energy sector did not aim to lower personnel-related expenses, but instead focused on establishing a reintegration clause for dismissed workers who would be the first to be rehired in better times.

Examples of sectoral programmes

After three consecutive years in crisis, the Bulgarian government froze wages in all budget agencies (state and municipal offices) and proposed the introduction of a new pay system for Bulgarian state employees from the beginning of 2012. Wages will be maintained at their existing levels but any increases will depend on the specific performance of employees, which will be evaluated periodically. The model of pay in Bulgaria’s public administration sector is such that there will be a basic monthly remuneration that will be a fixed rate (70% of fixed gross wage and 30% of bonuses). Employees will be evaluated annually for wage increases and quarterly for additional bonuses or extra incentives. This new model of payment in the state administration was fully supported by employers’ organisations in the National Council for Tripartite Cooperation (NCTC), but trade unions expressed concerns about the emergence of subjectivity in the assessment of employee performance. They therefore suggested that the new system should be introduced initially as a pilot scheme in only a few departments.

A collective labour agreement was concluded on 1 April 2010 in the Dutch Horeca sector. This was four months later than planned because negotiations proved quite difficult as the employers’ side could not come to an agreement. In the Netherlands, wages in the Horeca sector slumped during the past decade. Between 2001 and 2009, wages have risen by only 21% compared with 27% in all other sectors. The trade unions’ demand was for a pay rise in 2010 and 2011 but, during the negotiations, it became clear that the employers were not prepared to award any pay rise at all. In the end, an agreement was concluded whereby no rise was foreseen in 2010. Initially, a 1% rise was agreed for 2010 but this was changed to a different proposal, that is, 0.75% in January 2011 and 1% on 1 July 2011. The trade unions were prepared to agree this wage moderation because they wanted to give some more breathing space to those employers facing financial difficulties. The same scenario is happening again for the new agreement (2012–2013). The trade unions are prepared to show some moderation but want a guarantee that wages will follow inflation. The employers want to change the whole method of wage increases by linking any rises (even when corrected for inflation) to a performance assessment whereby, if an employer does not carry out the assessment in the allotted period, the worker should receive a 3% pay rise. The precise content of the 2012–2013 proposals is not yet known.

Company cases

National correspondents identified a total of 41 company cases spread over 20 Member States. As they are the most visible in local media, nearly all the case studies feature major organisations, implying that (sectoral or in-company) trade union organisations were actively involved in the trade-off negotiations.

Table 17 gives an overview of the sectors involved based on the two to three examples given by each national correspondent and, as such, should be regarded as illustrative only. Nevertheless, the manufacturing and automotive sectors, as well as banking and public transport seem to lead in terms of prevalence of agreements setting out trade-offs between wages and other working conditions.

Table 17: Trade-off company agreements identified by sector

Sector

Number

Agriculture

1

Automotive

11

Other manufacturing (plastics, electronics, metallurgy, energy, etc.)

10

Construction

1

Banking

6

Other services (car rental, hotel, retail sales, insurances, public sector organisation, etc.)

6

Public transport

6

Total

41

Source: National reports

Figure 12 shows the different types of labour cost-saving measures implemented by the management of the case study organisations.

Most of the cost-saving measures had only a temporary effect. This was primarily because employer and employee representatives were of the opinion that the economic recession would not have a lasting effect on the company’s perspectives.

Over two-thirds of the identified measures led to pay cuts. Only a minority of these reported cuts related to basic pay. In the other cases, the cuts were obtained by reducing working time or by saving on the remuneration on top of basic pay. The latter may be involved with either a variable pay system or with a reduction in extra paid working hours (overtime or weekend work). These are called interventions in the wage cushion.

The majority of the companies involved limited themselves to negotiating one single cost-saving measure.

Only one employer, a UK-based automotive company, implemented at least five cost-saving measures. In return for a clear and unambiguous guarantee for employment stability, workers had to accept:

  • a 12-month pay freeze;
  • a two-hour increase in weekly working time for blue-collar workers for up to two years with one hour of non-paid labour;
  • temporary suspension of shift payments over the same period;
  • an increase in the working week from 37 to 40 hours for white-collar workers;
  • an increase in employee contributions to the company pension scheme.

Figure 12: Negotiated cost-saving actions by companies

Description: TN1203015S_Figure12

Notes: The figures shown provide only a relative approximation of the most commonly used actions.

The totals exceed 41 because some companies adopted more than one measure.

Source: National reports, own classification

Figure 13 shows the types of compensation agreed as a result of negotiations between the organisation’s management and sector or in-company workers’ representatives.

In five out of the 41 cases, no specific types of compensation were granted in return for (sometimes substantial) income losses for the individuals involved. However, most of the aspects guaranteed by enterprise’s management touched on either the conservation of employment or the limitation of planned lay-offs. Other types of compensation involved other ways of income savings or guarantees, or state interventions to help the company’s economic recovery.

In all three Maltese case studies, for example, government action played a significant role in the negotiations. A number of cost-saving actions were accepted by workers under the condition of the company receiving substantial governmental subsidies, which among other things were used to finance intensive employee training sessions.

Figure 13: Types of negotiated compensation in company agreements

Description: TN1203015S_Figure13

Notes: Five cases reported no specific compensation.

The totals exceed 41 because, in one case, two types of compensation were granted.

Source: National reports, own classification

Examples of company cases

The Czech national air carrier, Czech Airlines (ČSA), has suffered long-lasting financial difficulties. In 2009, the company planned to dismiss up to 200 people if employees did not accept a wage cut. However, a direct reduction of wages was blocked by an existing three-year collective agreement. This set out a wage increase of one percentage point above inflation in 2009, which stood at 1.2% in June 2009. The trade unions and ČSA management discussed the issue of a wage cut for a number of weeks and the pilots finally agreed to a 6.5% wage cut in order to save jobs.

A Danish manufacturing enterprise put in place a number of arrangements involving its employees in order to manage the economic situation during the crisis. The enterprise’s first proposal was a nominal reduction in pay of approximately €6.70 (DKK 50) per hour, which was rejected by the employees and the trade union. After lengthy negotiations, the employees accepted a savings package in 2010 which consisted of a wage freeze until 2013, a reduction in bonuses and variable pay, and an end to employer-paid health insurance, physiotherapist, chiropractor and massage therapist. The previously full-time health and safety representative was given other job functions and the position of health and safety representative was deemed secondary.

Case studies

Although the cases reported by the national correspondents as part of this study are at different levels (national, sectoral, company), a number of elements can be stressed.

  • Most of the reported company cases are driven by a ‘negative’ curative strategy: pay cuts are a kind of last-resort strategy and as such are not a matter of trade-off. Jobs are directly in danger and have to be saved by drastic reductions in labour costs.
  • Most measures tend to also have a temporary basis. Companies suddenly hit hard by the crisis reacted by seeking immediate savings.
  • It is not surprising that the reported wage cuts are almost never accompanied by deals to enhance workers’ employability or the workability of jobs. The short-term perspective of trading labour cost savings with job security/guarantees is perhaps the main feature of these ultimate cases of reaction to the crisis.
  • A more ‘balanced’ approach can only be detected more strongly at supra-company levels of social dialogue. Wage moderations are combined with derogation possibilities. Reduction in working time is compensated by benefits and/or training possibilities. Job savings achieved by companies are accompanied by state investment or tax subsidies.

Overall, it can be concluded that only limited attention is paid in the current social dialogue to balancing the current crisis-induced wage measures with investment in measures to reach longer-term goals of quality of work and employability.

German company-level pacts for employment and competitiveness

Since the 1990s, pacts for employment and competitiveness (PECs) have increasingly become an integral part of the German system of industrial relations. These pacts are characterised by concessions from both sides. Employees agree to company-specific deviations from a sectoral collective agreement that includes reduced wages or prolonged working time in exchange for employment guarantees or investment programmes. Direct threats of job losses, internal restructuring, threat of plant relocation and insufficient profits are the main reasons for agreeing such pacts.

Recent information from the Institute for Employment Research (IAB) panel study reveals that about 1%–2% of all companies in Germany adopted this kind of agreement in 2009; the proportion rose to 15% in companies employing 100–250 people and to 30% in companies employing more than 500 people. The number of companies with a PEC increased by 50% between 2006 and 2009, that is, from 1.2% to 1.7% (Bellmann and Gerner, 2012a).

Evaluations of the pre-crisis period showed mixed results in terms of employment savings, that is, positive results in the short term, negative results in the medium term and positive results again in the longer term (Hübler, 2005). ‘Internal’ flexibility (in pay, teamwork or job tasks) seemed to be the key characteristic to avoid ‘external’ flexibility (Massa-Wirth and Seifert, 2005). The companies with PECs were also not providing more training (Bellmann and Gerner, 2012b). In other words, PECs seemed to be primarily a bargaining concession designed to contribute to a direct and short-term company stabilisation (of jobs). Therefore, it is perhaps no coincidence that, in their recent evaluation of the PECs in the 2008–2009 crisis, Bellmann and Gerner (2012a) demonstrate a clear employment effect, concluding that the agreements clearly reduced the negative employment effects of the crisis in those companies affected by it.

Measures to support vulnerable groups

Nearly all countries have put in place policies to support vulnerable groups of workers. Some (for example, France, Luxemburg and Poland) have even gone to great lengths to establish temporary or permanent protective measures to enhance the welfare of vulnerable workers or the unemployed. Only Denmark does not appear to have put additional policies in place probably because, like other Nordic countries, it was not hit so hard by the economic recession. The UK, nevertheless, reversed a number of its policy measures meant to support the living and working conditions of vulnerable groups of workers as part of its drive to drastically cut public spending.

The UK is the only country that limited the minimum wage for younger workers in order to boost employment for this category of the unemployed. Finland discontinued a temporary financial support system for enterprises hiring low-income and older workers on the basis of a study showing a negative assessment of the results. Portugal had put in place a series of measures aiming at maintaining employment in sectors and companies at risk (namely, reducing social security contributions for older workers, subsidising training initiatives in specific sectors), but it ended these initiatives by the end of December 2010 in an effort also intended to further cut public spending.

The following overview discusses only additional (new) or temporary interventions. Some countries (for example, Belgium and the Netherlands) periodically update their minimum wage or subsistence minima in line with the cost of living or the increase in general prosperity: these measures will not be considered here.

The national correspondents reported a total of 68 measures to promote better working and living conditions of vulnerable groups. Table 18 provides an overview classified by country.

Table 18: Support measures implemented since the crisis

Target

Policy measure

Countries

Employers

Tax allowances for companies or reduction of social security contributions, specifically oriented towards low wages

BE(2), EE, ES(2), FI, LT(2), NL, PL, SE

 

Increased access to training initiatives, mostly via subsidies

AT, BE, CY, FR, LV(2), NO, PL, SE, SK

 

Direct subsidies to companies in financial distress

HU, MT, SK

 

Various measures to facilitate hiring of specific groups of vulnerable workers

CY, FR(3), HU(2), LU(2), NL, SE

 

Limiting the minimum wage, sectoral, general or for some worker categories to facilitate hiring

EL, PL, RO, UK

 

Increasing redundancy payments or unemployment benefits for certain categories of workers

EL, BE, LU

Workers

Establishing or increasing the minimum wage (sectoral or general)

BG, DE, IE, LT

 

Provisions allowing a combination of welfare benefits and low-paid work; weakening of the criteria for entitlement to unemployment benefits, increasing unemployment benefits for certain categories of workers

CY, FR(2), LT, LU(2), LV, PL, SI

General or sectoral wage increase for low-paid workers

IE, LU, NO

Note: See Annex 2 for a full list of country codes.

Source: National reports, own classification

The most prevalent governmental measures are:

  • tax allowances;
  • subsidy systems for attendance at training and education initiatives;
  • weakening the criteria for entitlement to unemployment benefits.

The number of measures in favour of companies is roughly equal to the number of measures aimed at improving the living and working conditions of vulnerable groups of (potential) workers.

An issue of policy controversy is the minimum wage. In some countries, it has been increased to support income while in others it has been reduced as an incentive to create jobs.

The measures are mostly oriented towards the unemployed of various categories (17 measures) and low-paid workers (in total: 15 measures, among others implying an increase in the minimum wage). Norway oriented one intervention aimed at women (equal pay measures) and one aimed at migrant workers in the construction sector.

Examples of measures for vulnerable groups

In Austria, favourable conditions were implemented for short-time work that could be combined with training measures. Short-time work allowed workers to keep their jobs during the crisis and at the same time enabled companies to retain their employees and thus their company-related skills and knowledge. The short-time work provisions enable a temporary reduction of standard working hours based on a contractual agreement between social partners (company–works council) that aims to secure jobs and to compensate for employees’ wage losses by means of a short-time work allowance. The further training allowance aimed to encourage people on short-time work to participate in qualification measures.

In Spain, the most important recent policy initiative intended to support the situation of the most vulnerable groups of workers is the Labour Reform passed in June 2010 by the Spanish government. For more vulnerable groups of workers, one part of the reform refers to the promotion of employment of young and unemployed people. The most important issue raised under this heading is the reallocation of tax benefits for employers for the recruitment of individuals at high risk of unemployment. The benefits defined after the reform are per year for a period of three years, as follows:

a tax benefit of €800 for unemployed workers aged 16–31 years;

a tax benefit of €1,200 for unemployed workers over 45 years of age;

a tax benefit of €500 (€700 for women) for workers subject to training, handover and replacement contracts due to early retirement age.

These tax benefits can be complemented by other tax benefits aimed at encouraging the use of training contracts, which are generally given to young people.

Conclusions

In the course of 2008–2010, measures leading to pay cuts or pay freezes in one form or another on a national, regional, sectoral or in-company level were a common tool to counteract the effects of the economic crisis. Other interventions were also observed, such as reducing training opportunities, but these were rather an exception. A striking feature is that, by far, most measures were adopted on a temporary basis. This fits with the observation that the economic crisis in 2008–2009 was deep, unexpected and considered to be of a short duration. Governmental agencies, sectoral bodies and enterprises seem to have responded very quickly to the sudden economic downfall and to have focused on straightforward approaches.

A study of the individual country reports leads to the impression that, at the very beginning of the economic downfall, national governments were the first to respond by taking (rather unilaterally) some drastic measures. Contrary to the corrective measures defined via joint consultation at a somewhat later stage, many governmental interventions were aimed at having a lasting effect – tax reductions or other fiscal measures, modifications of the minimum wage, fine-tuning the social security system and, in many cases, in the first place, the unemployment benefits system.

There is, however, a clear difference between individual EU Member States. Some countries, having a major deficit, were limited in their capability to ‘grease’ social bargaining outcomes with extra financial means (for example, the Anglo-Saxon or southern European countries). Others have invested a considerable amount of public funding in an upswing of their socioeconomic fabric (for example, the Scandinavian countries and Malta).

This does not imply that collective bargaining institutions did not fully play their role – quite the contrary. However, the approaches taken vary considerably across the Member States. Industrial relations systems are deeply entrenched in national traditions, and there continues to be great diversity in how collective bargaining is organised and the number of workers covered by collective agreements (Hayter, 2011).

The general picture is, however, the same everywhere. From a managerial point of view, there has to be a balance in all cost-saving measures between protection measures with a longer-term perspective aimed at maintaining employment and high-skilled labour, and those measures aimed at immediate cost reductions, such as the temporary abolishment of some bonuses or the dismissal of a cohort of workers. In the harsh economic climate during 2008, running into 2009 and 2010, many concessions seem to have been made by the trade unions in favour of employment guarantees. Notwithstanding a few exceptions, these concessions were rather unsubstantiated and far from concrete.


General conclusions

The following general conclusions can be stated.

  • The main wage trends in the EU countries for the period 2008–2010 were deceleration and freezing. Sectoral data show that the crisis had more employment effects than wage impacts. These employment effects (upgrading or downgrading of the workforce) intrude in the average evolution of wages.
  • Wage cuts are thus not the first applied strategy in a crisis period. Lower-ranked jobs, especially the temporary ones, are cut or their working time is reduced. Higher-ranked jobs more often have a ‘wage cushion’ of variable pay or other rewards rather than basic pay that can be used to cut costs.
  • Cutting wages is seen as detrimental to motivation and job engagement in the workplace. The best workers would leave. These arguments are mentioned by employers and are confirmed in surveys of workers which indicate that ‘rewards’ are a key component in job satisfaction. Wages are the basic rewards. Labour regulation seems more important for blue-collar jobs to prevent pay interventions to cope with the crisis.
  • Specific EWCS analysis shows furthermore that the group of younger, low-skilled and low-paid workers (certainly those with a migrant background) are hit twice by the crisis. Their employment security and real income has decreased. Moreover, in those economies which showed a downgrading of employment, there is the strongest presence of wage freezing.
  • The presented case examples confirm the picture that concession bargaining on wages at company level is most of the time only applied when jobs are directly in danger. In these high-risk situations, there also seems to be no further room or attention given for a trade-off with other issues such as training. It is only at higher-level tripartite consultations that the question of trade-offs is tackled: wage cuts/freezing are ‘traded’ for state investments, income measures or new employment policies. A point of debate is the role and place of the minimum wage in this regard. The question arises regarding whether to lower it to incentivise the creation of jobs or increase it to guarantee a minimum for household incomes.
  • Nevertheless, the overall picture of these policies provides proof that the crisis came to a large extent as a surprise and that a lot of direct action was necessary. There was not much policy innovation but rather existing policies were applied and extended. Most strategies were also considered as temporary.

Wage issues have now come into the picture in the current EU debate on the economic governance of the crisis. Two main ideas emerge in this study.

  • It is important to look not only at average wage trends (based on aggregated national accounts) but also to carry out more detailed investigations, as those averages can hide important employment effects that influence the averages. Wages can be cut, but when the economy has lost a significant number of low-paid jobs, it is possible to get the impression that the average trend is one of growth.
  • There is much debate about second-round negative employment effects of wage growth during an economic downturn. However, the linking together of the various results of this study in a bottom–up approach shows that excessive wage freezing has also had a negative effect on job engagement and productivity, which can ultimately lead to macroeconomic imbalances.

Given the long-term EU2020 commitments to fight poverty and to get more people into quality jobs, the results from this study are worrying.

  • Particular groups in European societies (the low-skilled, young, migrants) have been hit hard by the crisis. Some have lost their jobs and, even for those still in a job, the job has become more precarious (lower income and higher job insecurity). These groups are already vulnerable in the labour market and struggling with social exclusion and poverty.
  • The crisis has also created more chaos in career development. The rising insecurity in employment and income has led, in a range of countries, to changes in perspective. In some (for example, Ireland), emigration is again an issue while in others (for example, the Czech Republic), the decision to look for a new job is being re-considered and ‘keeping what one has’ is the new attitude.
  • Based on the reported national findings, handling of the crisis through social dialogue seems not to have delivered many instruments and solutions. Most of the social dialogue, especially at company level, has focused on direct trading of wage concessions in return for saving jobs. This is the application of a direct and hard form of flexicurity, that is, flexibility on wages or working time in order to save jobs. Approaches that included employability issues or job quality questions were rarely reported. Only higher-level concertation seems to tackle these growing ‘imbalances’, albeit inconsistently.
  • Direct handling of the crisis seems to have taken up all the energy put into social dialogue in the period between 2008 and 2010. Policy innovations that combine ways to tackle the crisis with long-term strategies of inclusive growth appear, to a large extent, to be lacking.


References

Babecký, J., Du Caju, P., Kosma, T., Lawless, M., Messina, J. and Rõõm, T. (2010), ‘Downward nominal and real wage rigidity: Survey evidence from European Firms’, Scandinavian Journal of Economics, Vol. 112, No. 4, 2010, pp. 643–920.

Bellmann, L. and Gerner, H.-D. (2012a), ‘Company-level pacts for employment in the global crisis 2008/09’, The International Journal of Human Resource Management [in press].

Bellmann, L. and Gerner, H.-D. (2012b), ‘Further training and company-level pacts for employment in Germany’, Jahrbücher für Nationalökonomie und Statistik, Vol. 232, No. 2, 2012, pp. 98–115.

Erikson, R., Goldthorpe, J.H. and Portocarero, L. (1979), ‘Intergenerational class mobility in three Western European societies: England, France and Sweden’, British Journal of Sociology, Vol. 30, No. 4, 1979, pp. 341–415.

Fabiani, S., Lamo, A., Messina, J., and Rõõm, T. (2010), Firm adjustment during times of crisis (conference presentation), 7th ECB/CEPR Labour Market Workshop, Unemployment developments after the crisis, 16–17 December 2010, Frankfurt am Main, Germany.

Hayter, S., Fashoyin, T. and Kochan, T.A. (2011), ‘Collective bargaining for the 21st century’, The Journal of Industrial Relations, Vol. 53, No. 2, 2011, pp. 225–247.

Herzberg F. (2003), One more time: How do you motivate employees?, Harvard Business Review, Vol. 81, No. 1, 2003, pp. 87–96.

Hübler, O. (2005), ‘Sind betriebliche Bündnisse für Arbeit erfolgreich?’, Jahrbücher für Nationalökonomie und Statistik, Vol. 225, No. 6, 2005, pp. 630–652.

Massa-Wirth, H. and Seifert, H. (2005), ‘Pacts for employment and competitiveness in Germany’, Industrial Relations Journal, Vol. 36, No. 3, pp. 217–240.

Rose, M. (2001), Disparate measures in the workplace: Quantifying overall job satisfaction (6.77Kb PDF) (conference presentation), 5–7 July 2001, British Household Panel Survey (BHPS) Research Conference, Colchester.

Sousa-Poza, A. and Sousa-Poza, A.A. (2000), ‘Well-being at work: a cross-national analysis of the levels and determinants of job satisfaction’, Journal of Socio-Economics, Vol. 29, No. 6, 2000, pp. 517–538.

Wilkinson, R. and Pickett, K. (2009), The spirit level: why more equal societies almost always do better, Allen Lane, London.

Sem Vandekerckhove, Jan Van Peteghem and Guy Van Gyes, HIVA-KU LEUVEN


Annex 1: Country intercepts in multinomial model of types of crisis outcomes

Table A1: Estimates of country intercepts in multinomial model of types of crisis outcomes – type W
Model step 1   2   3  
Austria

0.08

***

0.06

***

0.05

***

Belgium

0.08

***

0.06

***

0.04

***

Bulgaria

0.14

***

0.11

***

0.08

***

Cyprus

0.07

***

0.05

***

0.04

***

Czech Republic

0.11

***

0.08

***

0.06

***

Denmark

0.16

***

0.11

***

0.08

***

Estonia

0.88

 

0.7

 

0.52

*

Finland

0.1

***

0.08

***

0.06

***

France

0.09

***

0.06

***

0.05

***

Germany

0.06

***

0.04

***

0.03

***

Greece

0.13

***

0.09

***

0.07

***

Hungary

0.19

***

0.15

***

0.11

***

Ireland

0.93

 

0.73

 

0.59

**

Italy

0.09

***

0.06

***

0.05

***

Latvia

2.2

***

1.81

*

1.38

 
Lithuania

1.58

*

1.35

 

0.98

 
Luxembourg

0.06

**

0.04

**

0.03

***

Malta

0.08

**

0.06

**

0.05

**

Netherlands

0.12

***

0.08

***

0.06

***

Poland

0.07

***

0.05

***

0.04

***

Portugal

0.07

***

0.04

***

0.03

***

Romania

0.23

***

0.18

***

0.14

***

Slovakia

0.21

***

0.15

***

0.11

***

Slovenia

0.2

***

0.14

***

0.1

***

Spain

0.11

***

0.08

***

0.06

***

Sweden

0.06

***

0.04

***

0.03

***

United Kingdom

0.11

***

0.08

***

0.07

***

Norway

0.08

***

0.06

***

0.05

***

Note: * p <0.05; ** p <0.01; *** p <0.001

Source: EWCS 2010, own calculations

Table A2: Estimates of country intercepts in multinomial model of types of crisis outcomes – type L

Model step

1

 

2

 

3

 
Austria

0.12

***

0.17

***

0.15

***

Belgium

0.2

***

0.28

***

0.26

***

Bulgaria

0.38

***

0.61

***

0.51

***

Cyprus

0.16

***

0.23

**

0.21

***

Czech Republic

0.52

***

0.79

*

0.75

*

Denmark

0.1

***

0.13

***

0.12

***

Estonia

0.39

**

0.62

 

0.54

 
Finland

0.18

***

0.27

***

0.24

***

France

0.12

***

0.17

***

0.15

***

Germany

0.12

***

0.17

***

0.16

***

Greece

0.26

***

0.38

***

0.35

***

Hungary

0.28

***

0.44

***

0.39

***

Ireland

0.38

***

0.6

*

0.57

*

Italy

0.14

***

0.21

***

0.19

***

Latvia

0.39

**

0.53

 

0.46

 
Lithuania

0.56

*

0.89

 

0.81

 
Luxembourg

0.1

***

0.13

**

0.13

**

Malta

0.22

*

0.29

 

0.28

 
Netherlands

0.16

***

0.21

***

0.2

***

Poland

0.24

***

0.36

***

0.31

***

Portugal

0.2

***

0.3

***

0.25

***

Romania

0.26

***

0.43

***

0.36

***

Slovakia

0.14

***

0.2

***

0.17

***

Slovenia

0.32

***

0.45

***

0.38

***

Spain

0.32

***

0.48

***

0.42

***

Sweden

0.25

***

0.31

***

0.3

***

United Kingdom

0.13

***

0.2

***

0.18

***

Norway

0.1

***

0.13

***

0.12

***

Note: * p <0.05; ** p <0.01; *** p <0.001

Source: EWCS 2010, own calculations

Table A3: Estimates of country intercepts in multinomial model of types of crisis outcomes – type LW

Model step

1

 

2

 

3

 
Austria

0.02

***

0.03

***

0.02

***

Belgium

0.03

***

0.03

***

0.02

***

Bulgaria

0.15

***

0.18

***

0.12

***

Cyprus

0.04

***

0.04

***

0.03

***

Czech Republic

0.12

***

0.15

***

0.1

***

Denmark

0.03

***

0.03

***

0.02

***

Estonia

0.69

 

0.95

 

0.64

 
Finland

0.03

***

0.03

***

0.02

***

France

0.04

***

0.04

***

0.03

***

Germany

0.02

***

0.02

***

0.02

***

Greece

0.08

***

0.1

***

0.07

***

Hungary

0.13

***

0.15

***

0.1

***

Ireland

0.33

***

0.47

**

0.37

***

Italy

0.05

***

0.05

***

0.03

***

Latvia

1.15

 

1.64

 

1.15

 
Lithuania

1.39

 

2.26

**

1.53

 
Luxembourg

0.02

**

0.02

*

0.02

*

Malta

0.04

*

0.03

*

0.03

*

Netherlands

0.03

***

0.03

***

0.02

***

Poland

0.05

***

0.06

***

0.04

***

Portugal

0.05

***

0.06

***

0.04

***

Romania

0.19

***

0.26

***

0.16

***

Slovakia

0.06

***

0.07

***

0.04

***

Slovenia

0.12

***

0.13

***

0.08

***

Spain

0.06

***

0.07

***

0.05

***

Sweden

0.04

***

0.04

***

0.03

***

United Kingdom

0.03

***

0.04

***

0.03

***

Norway

0.01

***

0.01

***

0.01

***

Note: * p <0.05; ** p <0.01; *** p <0.001

Source: EWCS 2010, own calculations


Annex 2: Country codes

Country code Country name
AT

Austria

BE

Belgium

BG

Bulgaria

CY

Cyprus

CZ

Czech Republic

DE

Germany

DK

Denmark

EE

Estonia

EL

Greece

ES

Spain

FI

Finland

FR

France

HU

Hungary

IE

Ireland

IT

Italy

LT

Lithuania

LU

Luxembourg

LV

Latvia

MT

Malta

NL

Netherlands

NO

Norway

PL

Poland

PT

Portugal

RO

Romania

SE

Sweden

SI

Slovenia

SK

Slovakia

UK

United Kingdom

EF/12/44/EN