Che materia stai cercando?

Centralization of Wage Bargaining Revisited - Driffill Appunti scolastici Premium

Materiale didattico per il corso di Politica comparata del prof. Marco Giuliani. Trattasi dell'articolo di John Driffill dal titolo "The Centralization of Wage Bargaining Revisited: What Have we Learnt?", riguardante gli effetti del corporativismo sui salari e sulla disoccupazione.

Esame di Politica comparata docente Prof. M. Giuliani




the firms then set a price for their products, the market determined demand

for the good and the numbers of workers employed. Each union (or group of

unions) set a money wage rate, taking as given the money wage rates set by

every other bargaining unit. The outcome for the economy as a whole was

modelled as a Nash equilibrium among the wage-setting units. Given the

decisions made by all the other decision-makers, no decision-maker wished

to alter her or his decision. The implicit assumption about monetary policy is

that it is non-accommodating. Fixing total nominal demand is equivalent, in a

simple economy in which the quantity theory of the demand for money holds,

to fixing the money supply.

The model showed that the non-monotonic relationship between centraliza-

tion and wage-setting emerged as a theoretical possibility under assumptions

that, we argued, were quite plausible. The key assumption was that at higher

levels of aggregation, goods became less and less good substitutes for each

other. While individual firms may have faced elastic demand, industries faced

less elastic demands for their products and at the level of the whole economy,

the demand for goods was less elastic still. Thus more centralized unions en-

joyed more monopoly power and would raise wages further. Offsetting this was

the factor that when unions raise wages for their members, they also raise the

prices of some of the goods they consume. The bigger the union, the stronger

this effect and the less incentive they have therefore to raise wages. The inter-

action of these two effects may, depending on the elasticities of demand, lead

to the non-monotonic relation between the centralization of bargaining on the

one hand and real wages and employment on the other.

Empirical Analysis

Our empirical analysis looked at the relationship between indices of

centralization and/or co-ordination of bargaining, on the one hand, and

unemployment or a misery index (a combination of unemployment and

inflation) on the other (see Tables 1 and 2, and Figure 1). For the most part

it was a simple bivariate relationship. Other factors that might have affected

macroeconomic performance were not considered.

How should corporatism/centralization/co-ordination of bargaining be

defined and measured? We took a narrowly defined measure, arguing that

‘corporatism’ had been defined in many and various different ways and could

be taken to include many different factors, some fairly objectively measurable,

others more subjective and that there was a danger of circularity. There has

been much reconsideration of the definitions and the data since then.

We took an index based on ‘the extent of inter-union and inter-employer

co-operation in wage bargaining with the other side’ (Calmfors and Driffill,

© 2006 The Author(s)

Journal compilation © 2006 Blackwell Publishing Ltd 735


Table 1: Centralization of Bargaining

1970–74 1975–79 1980–84 1985–89 1990–94 1995–2000

Australia 4.0 4.0 4.0 4.0 2.0 2.0

Austria 3.0 3.0 3.0 3.0 3.0 3.0

Belgium 4.0 3.5 3.0 3.0 3.0 3.0

Canada 1.0 1.0 1.0 1.0 1.0 1.0

Denmark 5.0 5.0 3.0 3.0 3.0 2.0

Finland 5.0 5.0 4.0 5.0 5.0 5.0

France 2.0 2.0 2.0 2.0 2.0 2.0

Germany 3.0 3.0 3.0 3.0 3.0 3.0



Ireland 4.0 4.0 1.0 2.5 4.0 4.0

Italy 2.0 2.0 3.5 2.0 2.0 2.0

Japan 1.0 1.0 1.0 1.0 1.0 1.0


Netherlands 3.0 3.0 3.0 3.0 3.0 3.0

New Zealand 3.0 3.0 3.0 3.0 1.0 1.0

Norway 4.5 4.5 3.5 4.5 4.5 4.5

Portugal 5.0 4.0 3.0 3.0 4.0 4.0

Spain 5.0 4.0 4.0 3.5 3.0 3.0

Sweden 5.0 5.0 4.5 3.0 3.0 3.0

Switzerland 3.0 3.0 3.0 3.0 2.0 2.0

United Kingdom 2.0 2.0 1.0 1.0 1.0 1.0

United States 1.0 1.0 1.0 1.0 1.0 1.0

OECD (2004a, Table 3.5).


1988, p. 17). The index was the sum of two elements: the level of co-ordination

within national trade union confederations and within national employer

organizations; and the number of existing central union federations and their

co-operation and the number of existing central employer federations and their

co-operation. Each index was scored 1, 2, or 3. The ranking that emerged is

described in Table 3.

II. Empirical Critique and Developments

The rankings and scores were disputed from the outset, in some cases because

they ignored co-ordination amongst players that was not represented in the

© 2006 The Author(s)

Journal compilation: © 2006 Blackwell Publishing Ltd


Table 2: Co-ordination Index

1970–74 1975–79 1980–84 1985–89 1990–94 1995–2000

Australia 4.0 4.0 4.5 4.0 2.0 2.0

Austria 5.0 5.0 4.5 4.0 4.0 4.0

Belgium 4.0 3.5 4.0 4.0 4.0 4.5

Canada 1.0 3.0 1.0 1.0 1.0 1.0

Denmark 5.0 5.0 3.0 4.0 3.0 4.0

Finland 5.0 5.0 4.0 5.0 5.0 5.0

France 2.0 2.0 2.0 2.0 2.0 2.0

4.0 4.0 4.0 4.0 4.0 4.0




Ireland 4.0 4.0 1.0 2.5 4.0 4.0

Italy 2.0 2.0 3.5 2.0 3.0 4.0

Japan 4.0 4.0 4.0 4.0 4.0 4.0


Netherlands 3.0 4.0 4.5 4.0 4.0 4.0

4.0 4.0 4.0 4.0 1.0 1.0

New Zealand 4.5 4.5 3.5 4.5 4.5 4.5

Norway 5.0 4.0 3.0 3.0 4.0 4.0


Spain 5.0 4.0 4.0 3.5 3.0 3.0

Sweden 4.0 4.0 3.5 3.0 3.0 3.0

Switzerland 4.0 4.0 4.0 4.0 4.0 4.0

United Kingdom 3.0 4.0 1.0 1.0 1.0 1.0

United States 1.0 1.0 1.0 1.0 1.0 1.0

OECD (2004a, Table 3.5).


visible structures and in some because they took no account of the influences

of other factors like minimum wages, as in France. Switzerland, it was argued,

was in fact a neo-corporatist system with heavy involvement of employers’

organizations in policy design and much more co-ordination than the centraliza-

tion indices indicated. (These are comments that followed the presentation of

the paper in 1988). David Soskice (1991) argued that Japan and Switzerland

had not been treated properly, and that when they were treated properly, a

linear relationship not a hump emerged. The thrust of his argument was that

© 2006 The Author(s)

Journal compilation © 2006 Blackwell Publishing Ltd 737


Figure 1: Trade Union Density and Coverage 2000

100 Austria

France Finland Sweden


90 Australia

Spain Netherlands Denmark

80 Italy

Portugal Norway

70 Germany

(%) 60 Luxembourg

Coverage 50 Switzerland

40 Canada

30 UK

New Zealand

20 USA Japan


0 0 20 40 60 80 100

Density (%)

OECD (2004a, Table 3.3).


Table 3: Calmfors and Driffill (1988) Scoring of Countries for Centralization of

Bargaining Co-ordination Level Existence of Total Score

Central Central Organizations


Organizations and their Co-operation

1 Austria 3 3 6

2 Norway 3 2 5

3 Sweden 3 2 5

4 Denmark 3– 2 5–

5 Finland 3– 2 5–

6 Germany 2– 3– 5–

7 Netherlands 2 2+ 4+

8 Belgium 2 2 4

9 New Zealand 1 3 4

10 Australia 1 3 4

11 France 1+ 2 3+

12 UK 0+ 3 3+

13 Italy 1+ 2 3+

14 Japan 1 2 3

15 Switzerland 1 2 3

16 US 1 1 2

17 Canada 1 1 2

Table A1 from Calmfors and Driffill (1988). The rankings of countries with equal scores are exp-


lained in the appendix of the paper.

© 2006 The Author(s)

Journal compilation: © 2006 Blackwell Publishing Ltd


Calmfors and Driffill had paid too little attention to the role of co-ordination

among employers’ organizations and that too much attention had been paid

to the level at which bargains were struck (firm/industry/economy) and too

little to the extent of co-ordination among participants to wage-setting across

units (again, firms or industries as the case may be). Thus he also revised the

relative positions of France, Italy and the UK. Robert Flanagan (1999) writes

that, after using a wide variety of indicators of economic performance and new

data on the centralization and co-ordination of collective bargaining, a 1997

OECD study concluded that there was no evidence to support the Calmfors

and Driffill hypothesis in the 1990s.

Nevertheless, Lars Calmfors (2001) reports a number of studies that have

found a hump-shaped relationship between centralization or co-ordination

and economic performance. He reports six studies that show a monotonic

relationship and five with a hump. The average unemployment rate for low

and medium relative to high co-ordination economies is 6.8 and 3.2 per

cent among the monotonic studies and 4.9 and 6.8 per cent among the non-

monotonic studies. (The studies he reports are these. Monotonic: Layard et

1991; Zetterburg, 1995; Scarpetta, 1996; Bleaney, 1996; Elmeskov

al., et al.,

1996; Nickell and Layard, 1999. Non-monotonic: Zetterburg, 1995; Bleaney,

1996, Scarpetta, 1996; Elmeskov 1996 [two entries]). However, all these

et al.,

studies appear to suggest that high co-ordination leads to lower unemployment

than does low co-ordination.

The more important way in which empirical research in this area has de-

veloped is in the direction of considering bargaining structure as not the only

but one of many factors that might affect unemployment and other measures

of economic performance and also taking account of dynamics of effects,

rather than simply looking at static models. Recent examples in this vein in-

clude Nickell (2005) and Blanchard (2006). Nickell consider the

et al. et al.

evolution of unemployment in the OECD from the 1960s to the 1990s and

allow for the effects of the unemployment benefit system, systems of wage

determination, employment protection legislation, labour taxes and barriers

to labour mobility. In addition to these structural variables, which evolve over

time, they also allow for the effects of shocks to interact with the structural

features. They find that a fall in union density causes a significant fall in un-

employment; higher co-ordination in bargaining reduces unemployment and

the effect of co-ordination is greater when union density is greater; the effect

of co-ordination is greater when the tax rate on employment is higher. Here

then the effect of bargaining comes principally through co-ordination and the

effect is monotonic. However, they do not look for a non-monotonic effect, so

this does not provide very strong evidence against such an effect. This study

shows that there is an important effect of bargaining structures when controlling

© 2006 The Author(s)

Journal compilation © 2006 Blackwell Publishing Ltd 739


for other influences on unemployment. Bargaining is not the only or principal

influence, but it has a clear effect. Employment protection legislation does not

appear to raise unemployment. But the replacement ratio and the duration of

benefits have a clear effect. Employment taxes have a positive effect, while

labour demand shocks and productivity (tfp) shocks reduce unemployment and

real import price shocks, money supply shocks and real interest rates increase


Blanchard’s (2006) account puts rather more weight on real and nominal

wage rigidity, on factors affecting flows in the labour market and on the

interaction of shocks and institutions. He in fact expresses a degree of doubt

about the amount of information that can be extracted from an exercise like

that of Nickell (2005) (‘It is clear however that the number of shocks,

et al.

institutions and interactions is sufficiently large that the ability of such panel

data regressions to tell us what exact combination of shocks and institutions

matter is limited. Such regressions allow us to check for simple and partial

correlations; they are unlikely to tell us about which combination of shocks

and institutions is responsible for unemployment’ (p. 26)). He actually seems

to place little importance on collective bargaining, density, co-ordination and

so on.

While the empirical status of the hump has been questioned by subsequent

research, one of Calmfors and Driffill’s theoretical predictions has been widely

confirmed: highly co-ordinated or centralized bargaining is associated with

lower unemployment and better economic performance. The many studies

referred to in the previous paragraph confirm this. Baker (2005) in their

et al.

cross-country study, find that co-ordination of bargaining is among the variables

that consistently has a significant negative effect on unemployment.

III. Theoretical Developments

Theoretical modelling of the relationship between centralization and unem-

ployment has developed in various directions. The issue of how labour markets

would interact with monetary policy was obviously a key question that Calmfors

and Driffill (1988) did not address. Its importance grew with the impending

arrival of the European Central Bank in 1999 when Economic and Monetary

Union was established in Europe. In the late 1980s and early 1990s monetary

policy shifted towards what is now the dominant model, consisting of an in-

dependent central bank operating by means of setting short-term interest rates,

with an inflation target. Torben Iversen and David Soskice (1998, 2000) and

Alex Cukierman and his collaborators, have taken up this issue and modelled

the interaction between union wage setting and monetary policy. Coricelli et

© 2006 The Author(s)

Journal compilation: © 2006 Blackwell Publishing Ltd


(2005) refer to the very many papers on this and related topics and cite a


good number of them.

Cukierman and Lippi (1999) consider the interaction of wage setting

(centralization of wage bargaining, or CWB) and central bank independence

(CBI). They take an economy with several unions and a completely union-

ized labour market, in which unions face less elastic labour demands as they

become more centralized. They set nominal wages in anticipation of the reac-

tion of the central bank. The unions are assumed to like higher real wages and

full employment of their members, but to dislike inflation. The central bank

(CB) is assumed to have conventional objectives. It has a higher employment

target than the unions and a zero inflation target. The unions set wages before

the central bank sets the price level. The central bank tends to accommodate

partially the wage increases set by the unions and it may be more or less averse

to inflation. Inflation is increasing in the excess of the union-set wage over the

competitive real wage. When unions are more centralized they have greater

market power, which they could use to raise real wages by more, but on the

other hand they take more account of the effect they have on inflation and thus

moderate their wage demands. The interaction of the two forces can give rise to

a hump-shaped relationship between centralization and real wages/unemploy-

ment for a given degree of central bank independence. A fully decentralized

labour market may give lower wages and unemployment and lower inflation

than a fully centralized one.

The key question they consider is: what happens if you make the central

bank more independent (conservative) or inflation averse? Under certain

circumstances it can lead to unions setting higher real wages and the outcome

can be higher unemployment. It is possible that inflation is higher also. With

a single union, the socially optimal outcome emerges when the central bank is

ultra-liberal which, as it were, scares the union into setting a wage consistent

with zero inflation. Thus this model gives some unexpected and counter-

intuitive results.

Compared with Calmfors and Driffill (1988), Cukierman and Lippi introduce

an accommodating central bank. The central bank in their analysis is a short-

term optimizer: it lacks commitment to an objective of policy. They also

ex ante

introduce another key ingredient into the analysis: the unions care about the rate

of inflation. This feature explains why an ultra-liberal central bank can induce

an inflation-averse centralized union to restrain wage increases. Calmfors and

Driffill omitted both of these elements, whose inclusion greatly increases the

range of outcomes the model can predict, depending on the strengths of the

preferences of the unions and central bank regarding inflation.

While Cukierman and Lippi’s result that, with highly centralized unions,

a more conservative central bank may induce higher unemployment and

© 2006 The Author(s)

Journal compilation © 2006 Blackwell Publishing Ltd 741


inflation is striking and suggests that the set-up of the European Central Bank

with its hawkish stance on inflation may have undesirable and unintended

consequences, one is bound to wonder about its empirical significance. Within

the context of a single European market, unions within any individual European

country cover only a small fraction of the labour market within any particular

industry and certainly they cover at most only a small fraction of the euro

area labour market as a whole. Jean-Pierre Danthine and Jennifer Hunt (1994)

addressed the question of how international integration of economies would

affect bargaining outcomes. They argued that openness of the economy would

be likely to lower the hump. Greater international competition would reduce

the ability of domestic industry unions to raise wages. At the same time the

national price level would be less affected by the wage decisions of domestic

unions. Their analysis suggests that the practical significance of strategic

interactions between unions and the ECB might be very limited.

Posen and Gould (2006) provide valuable empirical evidence on this

question. They explore recent data on wage restraint and its determinants for

a panel of OECD economies. Their exhaustive attempts to find empirical sup-

port for the theoretical predictions of Iversen and Soskice (1998, 2000) and

Cukierman and Lippi (1999) find nothing. If anything, the evidence points in

the opposite direction. Increased credibility of the monetary institutions has

caused an increase in wage restraint, rather than the predicted fall.

The concept of corporatism has been discussed and analysed more broadly

in books such as those of Pohjola (1992) and Hartog and Teulings

et al.

(1998). There are many ramifications of the concept. A topic that has been

widely taken up in the literature is the implications of centralized bargaining

for the distribution of earnings and relative pay of different groups of workers.

The compression of differentials, leading to problems in recruiting more

highly skilled workers, was one of the factors that led the Swedish employers

organization to withdraw from centralized bargaining.

There has been a massive rise of research into the relationships between

institutions and growth, which include labour market institutions among

the many factors that have affected growth. Daveri and Tabellini (2000), for

example, link the fall in growth rates and rise in unemployment among some

European economies to the combination of rising taxes on labour and strong

trade unions. Strong unions, they argue, enable workers to shift the burden of

increasing labour taxes on to employers, who respond by substituting away

from employment of labour to employment of capital, lowering the return on

new investment and discouraging saving and growth. They analyse only a short

period of time, less than 40 years, so it is unclear from their work whether the

effects on growth rates are permanent or merely transitory. But their arguments

© 2006 The Author(s)

Journal compilation: © 2006 Blackwell Publishing Ltd


illustrate a mechanism through which the interactions of labour market

institutions and taxation might have a prolonged effect on growth rates.

IV. Changes on the Ground

How have labour markets and economies more generally, changed since 1988?

What has happened to institutions and to macroeconomic performance?

There has been a fall in inflation and convergence at low inflation rates, as

Figure 2 shows. Inflation in OECD economies now lies in a range between

zero and 3 per cent, roughly speaking and has done so for the past three or four

years. Unemployment has not converged, of course. It ranges between roughly

4 and 11 per cent (see Figure 3). The convergence in inflation results from the

current implementation of monetary policy. Independent central banks (start-

ing around 1988) use the short-term interest rate as the policy instrument. The

adoption of inflation targeting, or something very close to it, with an inflation

target around 2 per cent is nearly universal.

The collapse of communism in 1989 and the subsequent opening up of

eastern Europe, has affected some economies greatly. Various eastern European

countries (Estonia, Latvia, Lithuania, Poland, Hungary, the Czech Republic,

Slovakia, Slovenia) are now members of the EU. The collapse of the Russian

economy hit Finland very hard. German unification had a potent effect on

Germany. Very high unemployment persists in eastern Germany. Driffill and

Miller (2003) explain this by noting the effects of the extension of west German

wage-bargaining institutions to east Germany. The west German unions were

very keen to achieve convergence of wages between east and west Germany

as quickly as possible and so pushed up eastern wages ahead of productivity

growth. The relatively rapid real wage growth in the east may have helped to

reduce the flow of eastern workers to the west and limited eastern competition

for jobs held by west Germans. Meanwhile the effects of high unemployment

in the east were mitigated by generous social security benefit payments and

substantial budget transfers from west to east. The pressures of high unemploy-

ment and low profitability in eastern Germany have been among the factors

leading to greater flexibility within German wage setting in recent years, which

is discussed in more detail below.

Across the OECD countries as a whole, there has been a general downward

drift in trade union membership, in coverage of collective bargains and in the

amount of co-ordination of bargaining among unions. Trade union density

fell from 34 per cent in 1970 to 21 per cent in 2000 (over a weighted average

of a constant sample of OECD countries) (see Table 4). Coverage has fallen

from 45 to 39 per cent on the same basis. Centralization of bargaining, on the

OECD’s measure has fallen notably in Australia, Denmark, New Zealand,

© 2006 The Author(s)

Journal compilation © 2006 Blackwell Publishing Ltd 743


Figure 2: Inflation Rates CPI inflation rate (%)

30 Portugal

rate 25 Greece

inflation 20 Ireland

15 UK Spain

CPI New Zealand



Yearly Ireland

5 Switzerland

Japan Netherlands

0 Japan

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

–5 Year

United Nations (2005).


Figure 3: Unemployment Rates


(%) Spain

rate Finland


unemployment Ireland

Canada UK


Standardized 5 Switzerland Luxembourg

0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004


OECD (2004a).


Spain and Sweden and slightly in one or two other countries (see Table 5).

The ‘Swedish model’ of highly centralized and co-ordinated wage-setting

disintegrated during the 1980s. The Danish approach labelled ‘flexicurity’ by

the OECD has been widely noted. Co-ordination of bargaining, again on the

© 2006 The Author(s)

Journal compilation: © 2006 Blackwell Publishing Ltd


Table 4: Trade Union Density (%)

1970 1980 1990 2000

Australia 44 48 40 25

Austria 63 57 47 37

Belgium 41 54 54 56

Canada 32 35 33 28

Denmark 60 79 75 74

Finland 51 69 72 76

France 22 18 10 10

Germany 32 35 31 25

Greece 39 32 27

Iceland 75 88 84

Ireland 53 57 51 38

Italy 37 50 39 35

Japan 35 31 25 22

Luxembourg 47 52 50 34

Netherlands 37 35 25 23

New Zealand 56 69 51 23

Norway 57 58 59 54

Portugal 61 32 24

Spain 7 11 15

Sweden 68 80 80 79

Switzerland 29 31 24 18

UK 45 51 39 31

United States 27 22 15 13

OECD 42 47 42 34 a

OECD 42 47 42 36 b

OECD 34 32 27 21 c

OECD 34 33 26 21 d

OECD (2004a, Ch. 3, Table 3.5).

Source: unweighted average; unweighted average (unchanged sample); weighted average; weighted

Notes: a b c d

average (unchanged sample).

OECD’s measure, has fallen in Australia, Austria (slightly), Denmark (slightly),

New Zealand, Portugal (slightly), Spain, Sweden and the UK (slightly), though

there have been increases in Belgium, Italy and the Netherlands (see Table 1).

These changes have accompanied the continued downward drift in the share

of employment in manufacturing and other industries in which unions were

© 2006 The Author(s)

Journal compilation © 2006 Blackwell Publishing Ltd




447.89 KB




+1 anno fa

Corso di laurea: Corso di laurea in scienze internazionali e istituzioni europee
Università: Milano - Unimi
A.A.: 2011-2012

I contenuti di questa pagina costituiscono rielaborazioni personali del Publisher Atreyu di informazioni apprese con la frequenza delle lezioni di Politica comparata e studio autonomo di eventuali libri di riferimento in preparazione dell'esame finale o della tesi. Non devono intendersi come materiale ufficiale dell'università Milano - Unimi o del prof Giuliani Marco.

Acquista con carta o conto PayPal

Scarica il file tutte le volte che vuoi

Paga con un conto PayPal per usufruire della garanzia Soddisfatto o rimborsato

Ti è piaciuto questo appunto? Valutalo!

Altri appunti di Politica comparata

Democrazia - Dahl
Scala di astrazione dei concetti
Sistema istituzionale ed elettorale in Bosnia - Erzegovina
Democrazia maggioritaria e consensuale - Lijphart