The Great Trough in Unemployment:
A Long-Term View of Unemployment, Inflation, Strikes and the Profit/Wage Ratio*
Walter Korpi, Swedish Institute for Social Research, Stockholm University,106 91 Sweden
Phone +468 163 450, Fax +468 154 670, e-mail Walter.Korpi@sofi.su.se
Walter Korpi is professor of social policy at the Swedish Institute for Social Research,
Stockholm University, and works on comparative welfare state development, political
sociology, political economy, class and gender. In English he has published The Working
Class in Welfare Capitalism: Work, Unions and Politics in Sweden (1978) and The
Democratic Class Struggle (1983) as well as articles in American and European journals of
sociology, political science, economics, and social policy. Recent articles include
“Contentious Institutions: An Augmented Rational-Actor Analysis of the Origins and Path
Dependency of Welfare State Institutions in the Western Countries” (Rationality and Society,
2001), "Faces of Inequality: Gender, Class, and Patterns of Inequalities in Different Types of
Welfare States" (Social Politics, 2000) and (with Joakim Palme) "The Paradox of
Redistribution and the Strategy of Equality: Welfare State Institutions, Inequality and Poverty
in the Western Countries" (American Sociological Review, 1998).
The third quarter of the twentieth century with full employment in most Western countries is
a historically unique period, forming The Great Trough in unemployment. This paper
analyses the beginning, continuation and demise of The Great Trough, contrasting a supply-
and-demand framework derived from economic theory with a power sensitive approach
focusing on long-term positive-sum conflicts involving major interest and reflected in
unemployment, inflation, industrial disputes and the functional distribution of national
income. Comparative empirical data from 18 countries are used in analyses of hypotheses
implied by the different theoretical perspectives.
1. THE GREAT TROUGH AND ITS EXPLANATION
In the Western countries, the labor market forms the context in which citizens create the
wealth of nations and participate in distributive processes generating socio-economic
stratification and inequality. For most individuals of working age, gainful employment is the
foundation for participation in distributive strife. This foundation is however highly
precarious and under permanent risk of succumbing to unemployment. In market economies,
variations in levels of unemployment over time and among countries are therefore of key
interest for the analysis of distributive strife and inequality.
Scholarly attempts to explain the causes of unemployment have typically focused on recent
periods. If we extend the time perspective to include the whole twentieth century, we find that
in most Western countries short-terms variations in unemployment levels are superimposed as
ripples on the long waves of unemployment. Here we discover an overlooked but clearly
deviant time period – roughly simultaneous in most of them – when what can be called The
Great Trough in unemployment comes forth. The Great Trough was created when from
traditionally very high levels, unemployment rates drastically fell during the first decade after
the end of the Second World War, and when after a quarter-century of full employment, these
rates dramatically rose after the first half of the 1970s
Both at the beginning and
the end of The Great Trough we do however find interesting differences among countries,
with the United States significantly deviating from most other Western countries.
(Figure 1 about here)
The beginning as well as the demise of The Great Trough have taken both citizens and social
scientists by surprise. Thus, for example, basing his views on the recurring experiences of
mass unemployment during the half-century before the Second World War, the British social
reformer William Beveridge saw 8.5 to 10 percent as a realistic estimate of unemployment
levels likely to dominate in post-war Britain.
Here Beveridge reflected a consensus shared by
most policy makers and economists up to that time: Like original sin, unemployment is a
subject for regret, not for reform. Only a couple of years later, however, as an effect of the
spread of Keynesian ideas, Beveridge proclaimed full employment – defined as an
unemployment level below three percent – as a goal for British post-war policy.
decade of the post-war era ushered in the "Golden Age" of Western capitalism, and after the
mid-1950s full employment in the Beveridgian sense came to prevail in most although not in
all Western countries.
Among policy makers as well as economists, the new consensus view was that unemployment
is a curable disease; economic policy should focus on maintaining full employment, and a
long-run trade-off between unemployment and the rate of inflation was assumed. Social
scientists such as Gunnar Myrdal and Arthur Okun were among the many observers who
believed that the pattern of recurrent high unemployment had become obsolete, since
governments now wanted to have full employment and knew how to achieve it.
makers saw preferences for full employment in the electorate as major forces pressing for
such a policy stance. As formulated by Anthony Crosland, a leading British Labour politician:
"Any Government which tampered seriously with the basic structure of the full-employment
Welfare State would meet with a sharp reverse at the polls".
The arrival of full employment,
breaking the long tradition of recurring mass unemployment, can be seen as a silent
revolution, reshaping distributive processes and citizens' life chances.
The Great Trough ended in connection with the oil shocks in 1973 and 1979, and in the 1980s
two-digit levels of unemployment were back in most but not in all Western countries. Soon,
however, social scientists as well as many others came to view mass unemployment as
something very difficult to avoid. The consensus view on unemployment and inflation was
again turned on its head. Economic policy should focus on maintaining a low and stable rate
of inflation, while unemployment came to be regarded as less important or as a means for
controlling inflation. As noted by an American economist, "in the United States and,
especially in Europe, those in authority often accept high unemployment with an air of
resignation, as if it stemmed from acts of nature rather than from acts of man"
. Thus almost
irrespective of its level, once an unemployment rate has prevailed for some time, conceptions
about what is normal and natural tend to change. In Europe, rates of unemployment which up
to 1973 were regarded as unacceptably high, came to be seen as nearly unattainable goals for
government policies only a decade later.
The key role of employment in citizens' everyday lives and in the generation of national
wealth makes variations in unemployment over time and across countries a central area for
the study of macro-level social, economic, and political change. In recent years and from
different perspectives, scholars have analyzed such change in a number of important works.
The present study explores the background to these changes in a novel perspective, focusing
on the role of long-term variations in unemployment in distributive conflict. Because of the
essential role of the labor market in the production and distribution of our worldly goods,
unemployment can be seen as a pivotal variable, which affects and reflects other basic
changes and can serve as something of a litmus indicator of their outcomes. Recent analyses
of unemployment have overwhelmingly focused on the explanation of the dramatic increase
in unemployment levels after 1973. Here the two oil shocks of the 1970s have often been
portrayed as the main causes of this increase; they are seen as generating demands on wage
adjustments which rigid labor markets and welfare state institutions have been unable to
handle. This paper explores an alternative perspective, where the oil shocks appear not as
major causes but instead as catalysts providing opportunities for changing key parameters of
long-term distributive conflict in Western societies. The paper therefore analyses not only the
return of mass unemployment but also the even more remarkable and historically unique
general fall in unemployment levels forming the beginning of The Great Trough, a
phenomenon which has attracted little scholarly attention.
Empirical data here refer to 18
countries with uninterrupted political democracy after the Second World War and at least one
million inhabitants, that is Australia, Austria, Belgium, Canada, Denmark, Finland, France,
Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Sweden, Switzerland,
the United Kingdom and the United States.
Previous studies by economists on the causes of unemployment have broached distributive
conflict primarily in terms of tensions among employed insiders and unemployed outsiders.
While tensions among employees can be significant, it is however important to also consider
potential conflicts of interests generated around the major faultline on the labor market, that
between employers and employees. The paper discusses diverging approaches to the
explanation of variations in unemployment and evaluates their relevance by confronting them
with comparative empirical data covering the beginning as well as the end of The Great
Trough. In such an undertaking it is fruitful to contrast two classes of approaches to the
explanation of unemployment; one consisting of what can be called equal exchange
approaches, the other of what can be called distributive conflict approaches. These classes are
of course broad and internally heterogeneous. Yet an important difference between them
concerns assumptions made about the distribution of power resources among actors, resources
defined in terms of capabilities of actors to reward and to punish other actors.
In equal exchange approaches to unemployment, differences between employers and
employees in terms of power are assumed to have relatively little significance. The clearly
most important variant in this class of approaches is neo-classical economic theory, where the
core theoretical idea of the perfect market largely defines away differences in power by the
assumption that all actors are so small that they have to accept market prices. The extent to
which power is defined out from the vocabulary of modern economics is indicated by the fact
that the term "power" does not appear in the main dictionary on economic terms and concepts,
The New Palgrave: A Dictionary of Economics
. Differences among the social sciences are
indicated by a quick comparative study showing that in
The International Encyclopedia of the
, the concept "power" covers ten pages.
In economics power difference enter into analyses of monopolies, monopsonies and
bargaining, and few regard the labor market as an atomistic one. Yet there is a tendency to
view relations between employers and employees as voluntary cooperation on relatively equal
terms. With various modifications, such a theoretical approach has informed a large part of
the many studies by economists on the increase of unemployment after 1973. In recent
decades, a broader New Keynesian approach has become significant, underlining the
importance of involuntary unemployment, aggregate demand and economic policy as well as
the operation of open economies and institutions such as unemployment insurance and wage
From more or less different starting points, however, a stress on the role
of labor market rigidities and distortions affecting wage setting looms large in great many
economic explanations of the post-1973 rise in unemployment. Yet, in spite of the very strong
position of neo-classically oriented economic thinking in recent decades, among economists
we also find dissenting voices underlining the need for broader perspectives and these voices
will here serve as links to the alternative approach discussed here.
Distributive conflict approaches assume a positive-sum conflict between employers and
employees and attend to the role of differences in power in the employment relationship.
Among conflict approaches we will here focus on what has become known as the power
resources approach, which assumes an unequal but varying balance of power between
employers and employees, typically with an advantage for employers.
The positive sum
nature of conflicts of interest in the employment relationship indicates that both sides have an
interest in enlarging the pie; yet the way in which the pie is to be divided is likely to be
conflictual. Looking at unemployment from this perspective, conflicts between employers and
employees concerning the distribution of results of production as well as conditions in the
production process become central. Such conflict is likely to have a variety of expressions,
including unemployment but also inflation, industrial disputes and the functional distribution
of national income. This perspective provides for a major role for variations in societal power
distributions and for political actors in the explanation of unemployment.
The present paper argues that while economic factors are important especially for short-term
variations in unemployment, when it comes to accounting for The Great Trough in
unemployment, the supply-and-demand framework derived from neo-classical economics
focusing on labor market distortions as central factors behind unemployment is of only partial
relevance. The purpose is however not to provide an alternative monocausal explanation but
instead to underline the need for considering the role of distributive conflict among unequally
positioned societal interest groups such as employers and employees as well as partisan
politics for variations in unemployment. To bring in such issues into the discussion of
unemployment brings up the difference between what is correct and incorrect, politically
speaking; in some areas of the social sciences this difference is easily significant at the three-
star level. As naturally as we talk about the weather, we can discuss moral hazards among the
unemployed, self-seeking politicians and public bureaucrats as well as rent-seeking unions
and insiders, but to introduce a political class-related element in the analysis of
unemployment is likely to be seen as a
. For good reasons, when doing so Robert
Solow introduced his talk in the following way:
"Now I want to speak about the unspeakable: I am almost tempted to suggest that women
and young people leave the room. The subject is one that, if it is mentioned at all in polite
company, is grouped with witchcraft, drunkenness, and the abuse of children, things that
we know are there but that are best denied.
It is possible that one source of continued
high unemployment in Europe is that the domestic demand for goods and services, and
therefore for labor, has been forced to unnecessarily and unhealthily low levels."
Solow called in question the widely accepted view that labor market rigidities and distortions
can explain the return of mass unemployment in Europe, and pointed to important Keynesian
elements for this return.
By arguing that we have to consider politics and distributive conflicts
among major interest groups, among them employers and employees, as factors significant for
long-term variations in unemployment and for differences among countries, this paper takes a
further step into the unspeakable
To give the reader an overview, the organization of the paper and some of the main questions
and issues are here summarized.
contains a brief review and discussion of
unemployment theories with a focus on a contrast between the neo-classical view and the
power sensitive approach to the explanation of unemployment. While neo-classical economic
theory defines away differences in power between employers and employees, the power
sensitive approach acknowledges potential power inequalities. Furthermore, this approach
points to the key relevance of differences between markets and democratic politics in terms of
the degree of inequality in the distribution of central power resources as well as in the socio-
economic patterning of resource distributions.
discusses basic assumptions and empirical support for interpretations of the rise of
unemployment in terms of increasing educational mismatch as well as within the supply-and-
demand framework inspired by neo-classical economic theory. According to the latter
interpretations the rise in unemployment after 1973 is largely an outcome of increasing labor
market dysfunctioning and rigidity, major distorting factors being institutions such as
unemployment insurance associated with moral hazards and decreasing search efficiency.
Sociological and psychological analyses point to the multi-dimensional positive role of
employment, making wages only one among several potent motivations to work. Empirical
data on life satisfaction among the unemployed and on the role of unemployment insurance
for the over-time development of unemployment are presented. The conclusion is that while
important, the contributions of economic studies in the neo-classical mold to the explanation
of the post-1973 rise in levels of unemployment remain partial.
begins the discussion of the relevance of the power sensitive approach by looking at
the relationship between levels of unemployment and wages. While the demand-and-supply
framework predicts that high wages will be associated with high unemployment, empirical
studies indicate that this association is instead a negative one. This negative association
supports the hypothesis that variations in levels of unemployment tend to affect the relative
power positions of employers and employees and thus to influence the outcome of distributive
conflict. As noted by some economists, unemployment can be seen as a worker disciplinary
device. Here we also discuss to what extent the shift in policy goals from full employment to
low inflation can be seen as based in electoral demands and on evidence indicating that
inflation is a greater menace to public welfare than unemployment.
compares the relevance of neo-classical hypotheses and the power sensitive
approach when it comes to the explanation of the beginning of The Great Trough. One key
issue is to what extent a mirror image of the distortions framework used to explain the rise of
unemployment after 1973 also can explain the decline in unemployment after the Second
World War. The power resources hypothesis suggests that the beginning of The Great Trough
can be seen as reflecting the sea change, when for the first time in the history of most of our
countries left parties emerged as a major political force, directly engaged in governments or
indirectly influencing policies through a "contagion from the left." Empirical analyses focus
on the extent to which this hypothesis receives support by inter-country differences in levels
of unemployment during The Great Trough, the United States here providing one major
analyses the changing patterns of distributive conflict in terms of inflation and
industrial disputes in the context of the new constellations of political power during the Great
Trough. Except for relatively short spells, inflation had been nearly low or absent before the
Second World War; after the war inflation came to increase at accelerating rates. At the same
time industrial disputes dramatically changed their shape, from drawn-out conflicts to brief
spells, however with drastically increasing participation rates. This change in shape is
discussed in terms of the relationship between industrial conflict and inflation. In Europe in
the late 1960s, strikes escalated and major attempts to reform traditional employer-employee
relationships emerged on the political agenda.
examines the concern and responses among economic policy makers to the situation
when during The Great Trough labor and the left had achieved historically strong power
positions. A largely neglected policy document by the OECD from 1970 provides valuable
insights into the ways that the problems of full employment, inflation and wage costs were
conceived prior to the oil shocks of the 1970s. What here is termed the Kalecki-Rehn
hypothesis indicates that the oil shocks generated by the OPEC in the 1970s opened up
windows of opportunity for governments to allow unemployment to rise without major
electoral backlash. In most Western countries the top priority policy goal now changed from
full employment to low and stable inflation. We discuss the distributive consequences of this
shift and its possible background in electoral demands.
looks at inter-country differences in unemployment increases at the end of The
Great Trough and relates them to models of wage bargaining and conflict resolution. To
explain differences in patterns of unemployment before and after 1973 between the United
States and European countries in the EEC and in the EFTA, we analyze interest rates set by
central banks and changes in interest rates in connection with the second oil shock.
Differences in the macro-political positions of central banks in Europe and in the United
States are discussed.
tests the hypothesis generated by the power sensitive approach that long-term
changes in societal power distributions will be reflected in the functional distribution of
national income, indicated by the ratio of business profits to employee wages. National
account data available from 1960 indicate that in most of our countries the profit/wage ratio
shows the expected U-shape; it tends to decline from 1960 to about 1980, then again to
increase up to the late 1990s. In the final
, the results are discussed.
2. EXCHANGE AND CONFLICT APPROACHES TO UNEMPLOYMENT
In the social sciences, the nature of the employment relationship is conceived in widely
different ways, differences which are reflected in theoretical approaches to the explanation of
unemployment. Central issues here concern assumptions made about power relations between
employers and employees, and the roles accorded to distributive conflict, politics and unions
for unemployment. As noted above, in neo-classical economic theory the basic model of the
perfect market defines away differences in power. A few economists appear to apply this
atomistic market model to the labor market, as evidenced in the now classical statement by
Alchian and Demsetz:
"The firm … has no power of fiat, no authority, no disciplinary action any different in
the slightest degree from ordinary market contracting between any two people. …
[The] presumed power to manage and assign workers to various tasks … [is] exactly
the same as one little consumer's power to manage and assign his grocer to various
Most economists are likely to view such a statement as an extreme. Yet there it a tendency to
treat the employment relationship as a voluntary, contractual relationship between near-
equals, thereby giving short weight to power differences in this relationship. Such a view
invites neo-classically oriented economists to explain unemployment in terms of the general
condition for market clearing; if markets do not clear, prices (wages) are too high. Within this
perspective it is very difficult to handle issues of potential distributive conflict between
employers and employees.
Both political science and sociology have been more hospitable to conflict perspectives.
fact that capital hires labor and is thereby accorded important managerial prerogatives is often
seen as indicating a basic power inequality in the employment relationship. It has long been
recognized that variations in unemployment play a crucial role for relations of power between
employers and employees. This role of unemployment once used to be expressed without
euphemisms. Thus in 1943, the leading conservative British newspaper,
"Unemployment is not a mere accidental blemish in a private enterprise economy. On
the contrary, it is part of the essential mechanism of the system, and has a definite
function to fulfil. The first function of unemployment (which has always existed in
open or disguised form) is that it maintains the authority of master over man. The
master has normally been in a position to say: 'If you do not want the job, there are
plenty of others who do.' When the man can say: 'If you do not want to employ me,
there are plenty of others who will,' the situation is radically altered."
As the above statement illustrates, changes in the level of unemployment will inversely affect
the nature of alternative choices available for employers and employees and thereby
differences in the relations of power between them. In the power sensitive perspective,
although both employers and employees can improve their outcomes by co-operating in the
production processes, the distribution of the revenues of production between profits and
wages generates conflicts.
The basis for conflicts of interests in the employment relationship
is the fact that both profits and wages have to be taken from the revenues of product sales,
and there is no natural rule for how the results of production are to be divided between
employers and employees.
The distributive process is therefore characterized by bargaining
and manifest conflicts, the outcomes of which are affected by the resources which employers
and employees have at their disposal for safeguarding their interests. Since variations in
unemployment affect relations of power between employers and employees, these variations
are likely to be reflected in distributive outcomes. Furthermore, since the employment
contract in many ways is an open-ended one, also questions about work performance and
discipline on the shop-floor become important issues of conflict.
The contrasting conceptualizations of power relationships between employers and employees
in exchange and conflict approaches are reflected in views on the nature of the relationships
between democratic politics and markets. In the conflict perspective, a starting point is that
within the firm employer power ultimately derives from ownership of shares, a power
resource with a highly skewed distribution among citizens. As is the case with other forms of
economic resources, this skewed distribution is relatively closely correlated with the socio-
economic system of stratification, "higher" strata owning more than the "lower" ones. Within
democratic politics, however, power – in principle – derives from votes of which each citizen
has one and only one; in the sphere of politics the central power resource is equally
distributed and thus uncorrelated with socio-economic position. We all know that in practice,
this principle is only more or less partially realized in the Western countries. With some
significant differences among countries, in all of them and in different ways political
decisionmakers are made aware of business interests. Yet democracy has made a difference.
Because of differences between the sphere of markets and the sphere of democratic politics in
the socio-economic patterning of the distribution of central power resources, assuming
boundedly rational actors we can expect individuals who are disadvantaged in the market
sphere to attempt to combine for collective action in the political arena, where their power
position is relatively better.
From a conflict perspective, the promotion by unions and
political parties of legislation on issues such as unemployment insurance, social rights,
employee rights and progressive taxation can thus be seen as reforms to redress some of the
consequences of inequalities inherent in the employment relationship. In a neo-classical
perspective, such reforms appear instead as rigidities and distortions in market processes,
hindering wages from falling to market-clearing levels.
Depending on our theoretical
glasses, political measures affecting conditions for and outcomes of market processes can thus
appear as "reforms" or as "rigidities."
For a long time economists saw unemployment as largely voluntary, and wage decreases as
the only way to counteract unemployment. In fact it took the Great Depression of the 1930s
before the presence of involuntary unemployment became widely accepted among
The Keynesian revolution against neo-classical economic thinking took place in
the context of the Great Depression. Keynes recognized that nominal wages tend to be
downward sticky. This is explained in terms of a concern for relative wage levels; workers are
unwilling to lower their wages to get jobs in which others are still working at previous wage
levels. Yet Keynes saw their unemployment as involuntary; if real wages were generally
decreased workers would be willing to accept a job at the previous nominal wage level.
Underbidding of wages in competition for jobs is unusual also because it can disturb social
relationships; employers tend to avoid hiring applicants who are willing to work for lower
Therefore labor markets can be stuck with high levels of unemployment for long
periods of time. As is well-known, Keynes therefore recommended that macro-economic
policies be used to increase demand. With increasing demand, more people will find jobs and
– via multiplier effects – generate virtuous circles decreasing unemployment. In the 1950s, the
so called Phillips curve, indicting that there was a trade-off between unemployment and
inflation, came to serve as a background to counter-cyclical economic policies.
In the 1960s, however, with Milton Friedman as a catalyst, the Keynesian approach to
demand management came under attack.
The attack was founded on the assumption that
with rational and correct expectations, economic actors will learn to expect the consequences
of government economic policies in terms of price increases and inflation, and to adjust their
actions to these anticipated consequences. In the long run, the trade-off between
unemployment and inflation is thus not possible, and policies to decrease unemployment will
only increase inflation. This view therefore led to the recommendation that governments
should avoid policy activism to counteract temporary recessions; in the long run the economy
would anyway return to its equilibrium level but without increased inflation rates. In analyses
of post-1973 events, the pre-Keynesian view that unemployment is voluntary saw a return in
the influential search theory, where workers were assumed to voluntarily quit employment in
order to be able to search more effectively for jobs which they expect will provide higher
wages than what they had had in their previous jobs.
In later research, however, the
assumption of voluntary unemployment has receded to the background and the focus has been
on various aspects of the search process.
3. EDUCATIONAL MISMATCH AND UNEMPLOYMENT INSURANCE
Economists' explanations of the rise of unemployment after 1973 first focused on commodity
price increases and wage changes associated with the oil shocks.
When high unemployment
turned out to be persistent the focus shifted to other causal factors, the central ones including
technological changes generating an educational mismatch among available jobs and the
education of the unemployed, the adverse effects of institutions creating labor market
rigidities, and the interaction between shocks and institutions.
As is well-known, the risk for unemployment has traditionally been negatively correlated with
education and skills. The fact that the unemployed on the average have lower educational
qualifications than those in employment has come to be widely interpreted in terms of an
increasing educational mismatch in the labor force resulting from an assumed widespread
upgrading of educational requirements on the job market generated by technological change.
But while a significant degree of upgrading of job requirements undoubtedly has taken place,
during the past few decades there has also been a very marked increase in the educational
qualifications of working-age populations.
Empirical studies in different countries indicate
that the rapid increase in educational levels in the labor force may have been even faster than
the educational requirements of jobs, leading to tendencies towards "over-qualification"
among those employed.
When formerly unemployed persons get jobs, their degree of "over-
qualification" tends to be higher than among those in continuous employment.
tendencies towards "over-qualification" can partly be the result of employers' positive
statistical discrimination in terms of education among job applicants, the numbers of which
increase with increasing unemployment.
Studies of institutions of relevance for the resurgence of unemployment have primarily
focused on unemployment insurance benefits, job protection legislation, and wage setting
For reasons of space we will here pay most attention to the role of
unemployment insurance, the classical
in economic analyses of unemployment
The explicit purposes for the introduction of unemployment insurance have been that benefits
should improve possibilities for finding a good match between jobs and workers, counteract
poverty among the unemployed, and maintain demand during recessions. Since the 1970s,
however, economists' attention has been centered on the potential negative consequences of
unemployment insurance, which has come to be seen as a key institution distorting labor
Attempts to explain the rise of unemployment after 1973 have turned around one central
"stylized fact": While in the United States unemployment relatively soon fell back close to
earlier post-war levels, in Europe mass unemployment became persistent. In mainstream
economics as well as in much of policy oriented discussions, the dominant explanation for
this contrast in the development and persistence of unemployment has been formulated in
terms of differences in the degree of distortions between labor markets in the United States
and Europe. With only modest unemployment benefits, little employment protection and
weak unions, in the United States wages have been seen as flexible, allowing for wage
adjustment to maintain labor demand but at the same time also generating drastic increases in
inequality of earnings and incomes. European labor markets are instead seen as increasingly
distorted, a development caused by multiple factors, central among which are generous
unemployment insurance in combination with high employment protection, high tax levels,
and strong unions. As a result, the structural component of unemployment – often termed the
equilibrium rate or the natural rate – is assumed to have increased.
insurance is assumed to increase unemployment in at least two ways; by generating "moral
hazards," that is fraud and misuse, and by lowering search intensity of the unemployed and
counteracting the decrease of reservation wages, that is, the wage demands of the
unemployed. We will here discuss these arguments in turn.
Layard, Nickell and Jackman argue that unemployment insurance increases unemployment
because it is "subject to massive problems of moral hazard (unless well administered) …"
this context we must however distinguish between what could be called misuse by policy
makers and misuse by unemployed individuals. Thus, for example, in branches with strong
seasonal variations, recurring periods with unemployment benefits have traditionally served
as indirect subsidies to some types of economic activities, something which from a national
perspective can be seen as undesirable.
Since 1973 in several European countries – contrary
to what was originally intended – policy makers reacted to the return of mass unemployment
by using unemployment, sickness and work accident insurance programs as means for
pushing and enticing out elderly workers from the labor force and for providing minimum
subsistence to the long-term unemployed. In several European countries, thus about half or
more of men 55 years and older disappeared from the labor force.
In countries such as the
Netherlands and Denmark, a plethora of programs providing long-term economic assistance at
minimum levels came to approach what could be called a citizens' wage.
In the context of
very high rates of unemployment, it is however debatable to what extent such policies can be
seen as constituting moral hazards for unemployed individuals.
It goes without saying that economic incentives are important and that unemployment
insurance programs can be misused, for example, to complement earnings from the black job
market or for a period of subsidized leisure. Hypotheses on the likelihood of individual
misuse are however likely to reflect our assumptions about rewards from work. Within
mainstream economics, work is generally conceived as a disutility with wages as the only or
the primary reward from work. In this frame of reference, given an unemployment insurance
program a spell of unemployment is typically treated as a period of leisure which can be
bought at a relatively low cost. Sociological and psychological theories do however predict
that for most individuals work and employment offer potent rewards in addition to the
The characteristics of one's occupation thus place a person in the socio-
economic stratification system reflecting differences in status and prestige within the
community as well as in closer social relations. In this way, the job becomes a main part of
the identity of most individuals and contributes in major ways to structure an individual's
social relations. Furthermore, in most contexts, being unemployed involves a considerable
stigma; as other categories of "undeserving" poor, also the unemployed come to be seen as
morally suspect. A young man introducing himself as living on unemployment assistance
should perhaps not expect to be immediately embraced by his prospective parents-in-law.
Employers tend to use unemployment as a basis for negative statistical discrimination, as a
short-cut indicator of undesirable characteristics of job applicants. Contrary to the original
assumption of search theory, workers who voluntarily quit a job in order to search more
effectively for better-paying ones would act in a highly irrational manner.
A large number of studies in different countries indicate that the experience of unemployment
is associated with mental stress and decline in psychic as well as physical health.
simple indicator of relevance here we can turn to the Euro-Barometer, where a standard
question asks respondents:
"On the whole, are you very satisfied, fairly satisfied, not very
satisfied or not at all satisfied with the life you lead?"
For the period 1977 to 1999 and for
eight countries (France, Belgium, Denmark, Germany, Ireland, Italy, the Netherlands, and
Great Britain) we can look at the percentages of the unemployed and of those in employment
responding "Not very satisfied" and "Not at all satisfied" to this question
those in employment, relatively few, about 20 percent, have chosen the two negative response
alternatives to this question; differences between women and men are here negligible. Among
unemployed women, this percentage is however about twice as high, and among unemployed
men more than twice as high as among those employed.
(Figure 2 about here)
Such indicators of an unusual level of mental stress among the unemployed should warn
against hasty conclusions about the level of moral hazards in unemployment insurance.
Unemployment, especially if prolonged, is a form of leisure that most citizens probably would
like to avoid. In some categories of the labor force, where work identification and work roles
are less clearly crystallized, periods of unemployment may cause less stress.
example, middle-aged or elderly women with irregular or intermittent work histories may
experience it less negatively. This circumstance may contribute to the somewhat lower level
of dissatisfaction among unemployed women noted above. Yet it appears highly unlikely that
the number of "chiselers" has increased so much as to reasonably account for any significant
proportion of the increase in OECD unemployment since the early 1970s. In this context we
must also consider what can be seen as positive consequences of unemployment insurance.
Thus by counteracting poverty, excessive economic strain and social exclusion of the
unemployed, this insurance is of great importance for maintaining the employability of the
It goes without saying, however, that there are moral hazards connected with
unemployment insurance as with other forms of insurance programs. For the maintenance of
public legitimacy of unemployment insurance, such misuse must be strongly discouraged.
A great number of empirical studies have been carried out in attempts to determine the effects
of levels and duration of unemployment benefits on search behavior and exits from
These studies have to a large extent been guided by search theory, focusing
on the process of job search among the unemployed, a process assumed to be negatively
affected by unemployment insurance. The results from these studies vary from no effects to
modest or even sizable effects of insurance on the probability of leaving unemployment.
Thus, for example, data from the United States as well as Sweden indicate that while the
probability for leaving unemployment decreases with its duration, among those remaining in
unemployment the probability of finding a new job shows some increase just before the
exhaustion of unemployment benefits. Since one of the purposes of unemployment insurance
is to provide the unemployed some time for finding a satisfactory job match, it should come
as no surprise if this insurance to some extent increases the duration of individual spells of
unemployment. The question is however if unemployment insurance contributes to
prolonging unemployment spells to the extent that the level of aggregate unemployment is
increased. As pointed out by Atkinson and Micklewright, even if unemployment insurance to
some extent prolongs the duration of individual spells of unemployment, this need not
increase aggregate unemployment.
The refusal of a job offer by one individual may only
mean that the offer goes to another job seeker. An analysis of findings in a large number of
empirical studies on the job search process indicates that unemployed persons searching for
jobs rarely refuse a job offer; a key conclusion is that "variation in unemployment durations
appears to arise primarily from variation in the likelihood of receiving offers".
observation underscores the role of the demand side for duration and levels of unemployment.
The distortions and inflexibility of labor markets generated by unemployment insurance and
other policy measures are assumed to be reflected in increasing levels of natural or structural
unemployment, often also termed the non-accelerating inflation rate of unemployment
(NAIRU), that is the level below which actual unemployment cannot fall without increasing
the rate of inflation. Estimates generally indicate that the NAIRU has increased with
increasing actual unemployment levels. Such estimates have come to serve as major guides
for economic policy making by central banks and governments in the Western countries. An
important question is therefore to what extent this concept has construct validity in the sense
that in the real world, these estimates actually reflect indicators of increasing labor market
rigidities and decreasing employability among the unemployed. Unfortunately, however, the
empirical bases for estimating NAIRU are not measures of real-world rigidities assumed to
mediate between the experience of unemployment and actual search behavior or
employability of the unemployed; instead these estimates are typically based on historical
data on unemployment and inflation or wages.
These studies thus assume – but do not show
– that labor market rigidities have increased or that the search activity and employability of
the unemployed has actually deteriorated.
The presence of long-term unemployment is widely explained in terms of hysteresis; with
prolonged unemployment, search efficiency of the unemployed is assumed to decrease and
their job skills, work morale and health to deteriorate. Thus the long-term unemployed not
only become less capable of filling vacancies on the labor market; they are also assumed to
become less effective as "inflation fighters," adding little to the downward pressure on wages
assumed to be required for market clearing.
Swedish data from the high levels of
unemployment in the 1990s, however, show that while the length of unemployment is
associated with decreased psychic well-being, this does not appear to decrease search activity
or work involvement among the unemployed.
Such findings are in agreement with the
conclusions drawn by the British economists Machin and Manning in a review of studies on
the causes and consequences of long-term unemployment in Europe:
"[T]he rise in the incidence of longterm unemployment has been 'caused' by the collapse
of outflow rates at all durations of unemployment; while the longterm unemployed do
leave unemployment at a slower rate than the shortterm unemployed, this has always
been the case and their outflow rates has not fallen over time; there is no evidence that,
for a given level of unemployment, the incidence of longterm unemployment has been
ratcheting up over time; once one controls for heterogeneity of the unemployed, there is
little evidence of outflow rates that decline over a spell of unemployment."
The strong influence of the neo-classical perspective with its focus on labor market rigidities
would appear to generate serious problems in economic analyses of and debates on the effects
of unemployment insurance. Here, one example can suffice. In what probably is
official policy analysis of the 1990s,
The OECD Jobs Study
came to the conclusion that in a
number of member countries, improvements of unemployment entitlements have generated
major increases in unemployment levels, however with time lags varying from 6 up to 17
years. Thus, for example, this report concludes: "Entitlements rose in Sweden in 1974 and in
Switzerland in 1977 with major rises in unemployment in 1991 in both cases".
In one of the
very rare critical comments by economists on this piece of writing, Manning describes such
causal interpretations as "absurd. In fact, one could write a very similar paragraph relating
performance in the Eurovision Song Contest to unemployment. Sweden won in 1991 (as well
as in 1974) and Switzerland in 1988, so this alternative hypothesis would seem better able to
explain the rise in unemployment in 1991 in both cases."
It has been argued that there is an interaction between economic shocks and labor market
institutions such as unemployment insurance. In good times, the unemployed can find jobs at
their previous wages relatively easily. When negative shocks hit the economy, however,
unemployment insurance will counteract the downward movement of reservation wages now
required for the unemployed to find a job. This assumed interaction between external shocks
and insurance generosity has made Ljungqvist and Sargent describe unemployment insurance
as a time bomb; when countries with generous unemployment insurance are hit by adverse
shocks, unemployment will explode.
This time-bomb hypothesis leads us to expect low
cross-country correlations between indicators of insurance generosity and unemployment
rates in The Great Trough 1960-72, correlations which will become high and positive after
As a preliminary test of the time-bomb hypothesis, we can look at annual cross-country
correlations among our countries during three time periods, the good times 1960-72, the bad
period 1975-87 and a post-shock period 1990-97. The generosity of unemployment insurance
is here measured with two indicators. One is the net replacement rate, here defined as net
benefits after taxes and social security contributions at the wage levels of average production
workers. Duration is measured in terms of weeks with the right to insurance benefits.
results do not support the predictions of the time-bomb hypothesis
. Net replacement
rates show low negative correlations for all three time periods. Correlations between
unemployment levels and duration have negative signs in the first period and the expected
positive signs in the following two periods; yet these correlations are very low and can hardly
form the base for strong interpretations. On the whole there would thus not appear to be any
marked inter-country associations between generosity of unemployment insurance and
unemployment levels during the 37-year period examined here.
(Table 1 about here)
The above discussion indicates that the supply-and-demand framework derived from neo-
classical economics has only been partially successful in accounting for the rise of
unemployment after 1973. We will therefore now turn to a discussion of the relevance of
hypotheses derived from a conflict perspective attending to power differences on the labor
4. UNEMPLOYMENT, WAGES AND DISTRIBUTIVE CONFLICT
In power sensitive perspective on the role of unemployment in distributive conflict, one key
question concerns the relationship between unemployment and wages. The demand-and-
supply framework derived from neo-classical economics predicts a positive association; high
wages are expected to generate high unemployment. Empirical studies by economists do
however indicate that this relationship is a negative one with a major part of causality running
in the opposite direction, that is lower unemployment tends to generate higher wages. Based
on aggregate data from the United Kingdom during a century, in a now classical study Philips
found an inverse relationship between the level of unemployment and the rate of change of
From a partly different angle, the relationships between individual wage
levels and aggregate unemployment levels within regions and industries have been penetrated
in meticulous studies over a wide range of countries and local labor markets.
point to the existence of what has been called the wage curve, showing a marked inverse
relationship between levels of unemployment and levels of wages; individuals who work in
labor markets with lower unemployment tend to have higher wages than similar workers in
markets with higher unemployment
. These findings provide clear support to the
hypotheses that variations in levels of unemployment will affect power relations between
employers and employees and thereby become central for distributive outcomes within the
(Figure 3 about here)
Also different varieties of efficiency wage theories explain the existence of unemployment by
assuming causality to run from wage levels to unemployment.
Borrowing from sociological
theories on social exchange, Akerlof and Yellen view wage increases above market clearing
wages as part of a mutual gift giving between employers and employees in situations where
employers cannot directly monitor work performance.
With full employment and market
clearing wages, dismissal of a worker found shirking will be an inefficient punishment since
the fired worker can quickly find a new job at the same wage level. Employers can solve this
monitoring problem by paying wages above the market clearing level, wages to which
employees respond by improving work performance and decreasing turnover. But if it pays
one employer to increase wages, it will also pay other employers to do it. When many
employers voluntarily choose to pay wages above market clearing levels, unemployment will
result. As argued by Shapiro and Stiglitz, the threat of unemployment will now serve as a
"worker disciplinary device," since with unemployment dismissal will be a punishment.
Theories on efficiency wages capture important parts of the employment relationship; the
interaction between employers and employees is also a social exchange, where norms on
fairness are significant. Here, however, the historically widespread phenomenon of strikes and
lockouts indicates that the employment relationship differs from the picture of conventional
gift exchange; it is associated with hard bargaining and manifest conflicts to an extent
reminding more of a marriage on the verge of break-up than of a Christmas party. In
efficiency wage approaches, employers are seen as being in a Prisoner's Dilemma game; in
the absence of collective employer action they have to pay extra so that unemployment will
become a worker disciplinary device. Here unemployment thus appears as something of a
collective good from employers' point of view. Assuming boundedly rational actors and also a
conflictual part in the employment relationship, if employers are willing to pay for
disciplinary devices, we would expect them to chose the cheapest way to achieve levels of
unemployment high enough to discourage shirking. In this context it is not clear that an
efficiency wage is the most cost-effective choice. From a conflict perspective, one alternative
to using the costly carrot of higher wages can be to support macro-economic policies likely to
generate the stick of high unemployment. In the long run we can expect that among
boundedly rational employers, not only social norms of fairness in mutual gift-giving but also
relative cost-effectiveness of potential threats are likely to affect their choice between carrots
and sticks as worker disciplinary devices.
A perspective on the employment relationship as a positive-sum conflict between boundedly
rational actors unequal in terms of power thus generates the hypothesis that
low levels of
unemployment are likely to be dearer to employees than to employers
. In democratic
countries, we can expect this difference in priorities to become reflected in the collective
actions of interest organizations and political parties. In policy making at the national level,
left parties with more or less close relationship to trade unions and a working-class electorate
are therefore likely to give a higher priority to full employment than are centrist and
conservative parties and employers' organizations. In this vein, Michal Kalecki argued that
"the assumption that a government will maintain full employment in a capitalist economy if it
only knows how to do it is fallacious." According to Kalecki, "under a regime of permanent
full employment, the 'sack' would cease to play its role as a disciplinary device. The social
position of the boss would be undermined, and the self-assurance and class consciousness of
the working class would grow. Strikes for wage increases and improvements in conditions of
work create political tensions."
Kalecki furthermore assumed that business interests have a
strong indirect influence over government policy. From the employers' point of view,
however, the situation is complicated. Boom periods tend to be associated with low
unemployment but also with relatively high profits; yet long periods of full employment may
affect the functional distribution of income and also have the problematic consequences for
employers pointed at by Kalecki.
5. FULL EMPLOYMENT AND CONTAGION FROM THE LEFT
Let us now compare neo-classically oriented theory and the power sensitive approach when it
comes to explaining the driving forces during the formation of The Great Trough, when mass
unemployment was followed by the period of full employment ushered in during the 1950s.
Since full employment came to characterize most but not all Western countries, it is here also
important to analyze inter-country differences in unemployment levels. Can the general
downward turn be interpreted with reference to factors many economists assume to have
generated the upward turn after 1973? This is argued by Layard, Nickell and Jackman, who
view long-term variations in unemployment as stemming from two sources.
One is long-
period change in social institutions, central among which are unemployment insurance and
systems for wage determination; the other big shocks such as oil price rises or major wars
with long-lasting effects. Other economists have interpreted the fall in unemployment levels
after the Second World War as a result of factors favoring economic growth, factors such as
the liberalization of trade barriers, demand generated by the reconstruction after the war, and
faster technological progress.
In this context it must however be remembered that the First World War, also a major war,
did not result in full employment; on the contrary it was followed by high, periodically
extremely high rates unemployment. And if we focus on factors many economists assume to
have increased rates of unemployment in these countries after 1973, the prospects for
drastically falling unemployment levels in the decades following the Second World War
would appear to have been bleak. Thus in the 1930s unemployment insurance – a main culprit
in the neo-classical conception – was found in only half of our countries but became a
universal part of the welfare states in all these countries in the 1950s and 60s, with increasing
benefits levels as well as coverage.
Furthermore, taxes which had been raised as part of war
efforts were generally maintained at levels much above those found in the 1930s, and in all
countries they continued to increase.
Yet, these theoretically adverse changes were
combined with drastically decreasing rates of unemployment. In the period 1960-73, although
not all countries had full employment those with relatively high unemployment did not differ
from the others in terms of GDP per capita growth.
A more promising perspective is sketched by a few economists, who have underlined the
central role of government policies for developments during The Great Trough. Thus
Shonfield underlines the importance of the "conscious pursuit of full employment" by
governments as a crucial factor for the achievement of full employment and the central role of
"political will and skill" in this process.
In the same vein, Boltho has shown that the
influence of government macro-economic policy activism during this period was of major
importance for full employment during the postwar years.
What factors could have motivated governments after the Second World War – unlike their
predecessor after the First World War – to use their political will and skill in the conscious
pursuit of full employment? Here a focus on changes in power relations provides an
alternative hypothesis for the beginning of The Great Trough.
From such a perspective, it is
striking that the shift from mass unemployment to full employment was contemporaneous
with another sea change taking place in the political arena – the full-scale appearance of a
new type of major contender for government power. The political scene was thus drastically
changed, when in the years after the end of the Second World War the formerly rather weak
left political tendency greatly increased its political strength as did its political allies in the
For first time in the history of capitalism, left parties now emerged
either as government parties or major opposition parties in most of our countries.
(Figure 4 about here)
Left parties emerged as major contenders for government power in a period when their
traditional preferences for full employment had received backing by practice as well as by
theory. Thus, during the war years full employment had become a reality as a result of war
efforts and the mobilization of a large number of men into the armed forces. For the majority
of citizens, full employment had come to be seen as a realistic goal; they had reasons to
believe that it could be politically achieved. What had been achieved in practice was given
theoretical backing by the growing support for Keynesian ideas among considerable sectors of
policy makers and economists. And in all Western countries the presence of the Soviet block
served as a warning against allowing the expected post-war disorganization to take the same
expressions as after the First World War.
Against such a background, with J.M. Keynes as one of the intellectual architects, the
international and national monetary institutions created at this time came to differ in crucial
ways from the pre-war ones. Now all countries had abandoned the gold standard, but the
United States, unquestionably the dominant military, economic and political power in the
world, retained a tie between the dollar and gold. In the 1944 Bretton Woods treaty, exchange
rates were pegged to the dollar but adjustable. A key part of the agreement was that national
governments were to control cross-border capital movements. While in the pre-war period
there were barriers for cross-border trade but free capital movements, to a significant extent
capital now became captive within national borders while trade was gradually liberalized. The
emergence of international institutions enabling governments to control capital movements
came to be of major significance for national policymaking during The Great Trough.
How did these changes in national and international contexts comet to affect the views of
policy makers in the Western countries? As indicated above, a power sensitive perspective on
unemployment leads us to expect that left parties are likely to give higher priority to low
unemployment than to low inflation, while centrist and conservative parties are likely to
reverse this order of priorities. A detailed analysis of legislative proposals, technical
documents and policy reports produced by parties in eight Western countries in the 1960s
supports this hypothesis.
While full employment tended to be the dominant objective for
parties on the left, centrist and conservative parties had retained their traditional view giving
first priority to low inflation.
Furthermore, as shown in Figure 3 on the average left parties
only participated in governments during less than half of the 1946-72 period. Full
employment during The Great Trough can thus not be explained in terms of a convergence of
political preferences of major interest groups, nor as carried out by the left parties alone.
In this context we must therefore remember that government policy makers are likely to
consider their strategic situation in the context of the perceived attractiveness of proposals
from other parties among voters. Here what Duverger once called "contagion from the left" is
therefore likely to have been at work.
In spite of continued own first preference for low
inflation, with the policy alternative for full employment favored and seen as realistic by the
electorate, in the presence of relatively strong left parties also centrists and conservative
government parties may come to see it as necessary for electoral survival to strive for full
employment. Our hypothesis is thus that such a contagion from the left will take place
primarily in countries where left parties are major contenders for government power.
To what extent is the hypothesis on the role of left governments and left contagion congruent
with empirical data on differences among countries in levels of unemployment during The
Great Trough? As an indicator for the probability of such contagion we will here use the
percentage of left votes and the percentage of time with left government participation in the
period 1919-59, that is in the period
to 1960-73, constructing a simple indicator based
on the average rank for each country on these two indicators.
When this indicator is related
to the average level of unemployment in 1960-73, we find no linear relationship between left
strength and later levels of unemployment; instead there appears to be a threshold effect in
terms of left political strength below which left contagion did not work
. Of the five
countries with the lowest ranks in terms of potential left influence, four – Italy, Ireland,
Canada and the United States – had clearly higher unemployment levels than the others.
(Figure 5 about here)
Of the above four countries, as is well known the United States never had a socialist party of
any significance. Also in Canada, the socialist tendency had been very weak although to some
extent present. The Irish Labour party was small and relatively insignificant in a party system
originating in different stands on the question of independence from Britain. Italy shares a
history relatively similar to that of Germany. In both countries left parties were repressed
during the periods of authoritarian rule. Immediately after the end of the war, the left share of
votes was actually somewhat higher in Italy than in Germany. However, the Italian left was
seriously split and dominated by a Communist party, the democratic credentials which were
widely questioned. Furthermore, during the Cold War, it was highly unlikely that Italy's major
allies would have allowed the Communist party into government. In Italy left parties had
participated in governments only during the exceptional years of 1944-48. In Germany,
however, the relatively strong Social Democratic party had formed governments before the
Nazi period, and after the war it was a major contender for government position. In
Switzerland and the Netherlands, social democrats participated in coalition governments after
the war. France had experienced a brief period of socialist government already in the late
1930s, and during the first fifteen years after the war some of the left parties participated more
or less regularly in coalition governments. In the other countries left parties had had even
stronger traditions of government participation. The political will to pursue full employment
policies thus reflected important changes in the distribution of political power resources in our
countries and the strategic adaptation of parties to this new situation, however with what
appears as a threshold effect.
During The Great Trough, the United States did thus not belong to the countries with full
employment as defined at that time in Europe.
During this period, in the United States and
Canada relatively high unemployment levels were associated with higher central bank real
interest rates than in Western Europe (cf below).
Japan is here an exception with absent or
weak left tendencies during the authoritarian period prior to the end of the Second World
War. Its very low unemployment rates during the first decades after the war probably partly
reflect statistical artifacts.
After the war the left as well as the unions were internally
divided. However it could be hypothesized that the conservative leaderships in politics and
business were willing to accept full employment to maintain control over the labor force.
The threat from the Soviet Union was perceived as very serious by policy makers in social
democratic as well as in centrist and conservative parties, and was an important consideration
behind the Marshall Plan providing assistance from the United States to support European
recovery. Yet the above analysis indicates that it was only in countries where left parties were
realistic contenders for government power that full employment in the Beveridgian sense of
unemployment levels below three percentage points came to prevail during The Great Trough.
In this context the national rather than the international political situation would thus appear
to have been decisive.
6. INFLATION AND STRIKES DURING THE GREAT TROUGH
The change in power relations generated by full employment during The Great Trough was
likely to have significant consequences for patterns of distributive conflict, important among
which are inflation and industrial disputes. In the neo-classical view, inflation tends to be seen
as an operational malfunctioning of the economy reflecting political intrusion and outside
forces which have distorted the normal balance of market processes. From a power
perspective, however, inflation can be seen as an outcome of distributive conflict reflecting
social divisions within market economies and channeled via the labor market.
above, there is a clear inverse relationship between levels of unemployment and nominal
wages. The step from wage increases to corresponding price increases is often taken as
A look at the historical developments indicates, however, that changes in
unemployment have not always affected inflation rates.
In the period before the First World War, with the exception of singular episodes in some
countries, rates of inflation were generally quite low; in the period 1900-13 with an average of
only 1.5 percent in our countries
. This situation is likely to have reflected
institutional factors, including the prevalence of the gold standard which tied the value of
paper money to metal.
During the First World War and up to 1920, inflation rates were
however very high, and many countries were forced to abandon the gold standard. In
Germany and also Austria, inflation reached skyhigh levels in 1922-23. During the 1920s,
several countries had returned to the gold standard. In response to the Great Depression of the
1930s, however, some countries again abandoned this tie. Excluding episodes of hyper-
inflation, however, in the period 1921-38, on the average there was some disinflation (
percent). As discussed by Hirsch and Oppenheimer, during this period in some countries
central bankers and governments were involved in debates on whether economic strains
should be placed on the monetary economy in the form of inflation or on the real economy in
the form of unemployment.
On the whole, however, low inflation was the dominant goal
and the burdens came to be placed on the real economy, a choice generating periods of
exceptionally high unemployment.
(Figure 6 about here)
As discussed above, after the end of the Second World War, as a result of the clear leftward
shift in political power in most western countries and the new threat from the Soviet Union,
there was a widely shared feeling that it was necessary to avoid protracted problems with
depression and mass unemployment similar to those which had followed the end of the First
World War. Reflecting the strategic political choice to give top priority to full employment,
inflation rates started to increase. They reached an average of 4.1 percent in the period 1949-
72, and escalated up to
percent 1973-84 in connection with increases in commodity prices
and the Vietnam war.
Another highly relevant indicator of distributive conflict in the Western countries is the level
and shape of industrial disputes. Before the First World War, reflecting the growing capacity
of collectivities of wage earners to engage in distributive conflicts, in 1900-13 employee
involvement in industrial disputes increased to an annual average of 1.9 percent, a figure
rising to 3.3 percent in the inter-war period (cf Figure 6).
After the end of the Second World
War, industrial conflict did not "wither away" as many had expected.
Instead we find a
divergence in patterns of conflict. In a few countries – Sweden, Norway and Austria – where
social democratic parties were in strong and stable government position, previous very high
levels of strikes and lockouts fell drastically, reflecting that distributive conflict had been
largely moved from the labor market to the political arena.
In most of our countries,
however, involvement in industrial conflict markedly increased in the 1946-72 and even more
so in 1973-84.
Thereafter conflict involvement drastically declines.
A rarely observed fact is that over these years the shape of industrial conflict was markedly
transformed. Before the First World War as well as during the inter-war years, industrial
disputes were protracted. On the average, involved persons participated 23-25 days, that is
about four weeks. During the post-war period, the average conflict became much shorter, one
week or less per participant.
In spite of the marked increase in involvement, this decrease in
duration did drastically bring down the number of man-days idle as a result of industrial
disputes. From a political perspective, however, the greatly increased involvement in disputes
is a central aspect. This change in the shape of strikes – from very long disputes to short
disputes but much higher involvement among employees – is likely to reflect parallel changes
in inflation rates. Before the Second World War, with the near absence of inflation assured by
gold standards and central bank policies, nominal wages resulting from bargaining between
employers and employees were in practice closely reflected in real wages. Therefore these
settlements often generated long disputes, fought by both parties to the bitter end. In the
period after the Second World War, wage negotiations turned instead into first-round
skirmishes on the determination of nominal wages. Employers were however now able to
withdraw rather quickly from costly industrial disputes to the second round, the raising of
product prices and thus the determination of real wages and inflation rates in connection with
In a historical perspective The Great Trough thus stands out as a period in the Western
countries during which the position of wage-earners and employees in distributive conflicts
was stronger than ever before, something reflecting the combination of their increased
organizational capacity via political parties and unions and the generally high labor demand.
Involvement in industrial disputes escalated. In the 1960s and 1970s, widespread conflicts
flared up, such as the events in France in May 1968, the "hot autumn" in Italy 1969, the
German rise in 1971, and the prolonged miners' strike in Britain 1973-74. Furthermore, in
several countries we find major proposals or legislation improving the position of workers at
the workplace. Countries such as Britain, Denmark and Sweden also saw attempts to encroach
on traditional managerial prerogatives as well as on capital formation. Such a situation was
likely to generate concern as well as counter-moves.
7. CHANGING POLICY GOALS AND THE KALECKI-REHN HYPOTHESIS
It is not surprising that the developments outlined above were viewed with increasing concern
by business interests and political elites in many Western countries. Yet up to the late 1960s,
the situation was grudgingly accepted, to some extent probably because fast economic growth
gave more to all. In government policy making, full employment continued to be given top
priority; here the fear that serious recessions would have major electoral repercussions is
likely to have played a crucial part.
Income policies were tried in attempts to keep wage
increases within limits.
During the tumultuous end of the 1960s, however, conservative and
centrists policy makers came to look for alternatives to these policies. This search was
facilitated by a change in the mood within the economics discipline, led by Friedman's attack
on Keynesian policy activism, and strengthened by the appearance of "stagflation," the
combination of high unemployment with rapid inflation, something seen as contradicting
An important but largely neglected early expression of the search for alternative economic
policy strategies is a policy document produced in 1970 by the Organization for Economic
Co-operation and Development (OECD), playing an important advisory role in relation to
governments in member countries. This document,
Inflation: The Present Problem
, gives a
very interesting illustration of the views of influential policy advisors in the years
the first oil shock and the ensuing period of high inflation. In outlining the problems
generated by inflation, this document points at issues such as insecurity, wage-price spirals,
and increasing interest rates. The report also pointed at the role of high labor demand for
increasing the share of wages and salaries in the national product while decreasing the
corporate share. In its conclusions, the OECD argues for "the urgent need to give higher
priority to price stability," making the key observation that "the problem of inflation arises in
part from the very success of post-war economic policies in other directions – notably in
achieving high levels of employment." The OECD policy report realistically recognizes that
"giving higher priority to price stability means giving lower priority to something else" and
that "in a number of countries this may temporarily have to be [growth and employment]." To
restore price stability the first recommendation of the OECD is that "excess demand should be
eliminated and governments should be prepared, where necessary, to accept a temporary
reduction in the rate of activity until there are signs that better price stability has been
The OECD policy advisors were however keenly aware of the political difficulties in carrying
out the recommendations for downgrading the top priority widely accorded to full
employment. In a cautious discussion of the proposal to recreate "the fear of a real recession"
as a cure for inflation, echoing widely shared beliefs expressed by persons such as Crossland,
Myrdal and Okun, they were very concerned with the potential of negative electoral reactions
to governments attempting to implement their proposal. The report recognizes that "a major
recession would no longer be a
deux ex machina"
but would instead be seen as "a result from
a miscalculation or a deliberate act of government. Indeed, the problem of inflation today
arises precisely because, when faced with a choice, governments have hesitated to give high
priority to price stability." Thus "today a serious recession would be clearly recognized to be
the result of a deliberate policy being followed by the government," while in the 1930s "it
could be seen as a natural disaster." Therefore "there is still too great a tendency to think in
terms of protecting production and employment regardless of costs and prices." In a succinct
summary of the policy dilemma as conceived at that time they write that
"… the fundamental
problem is how to get people to exercise the moderation that they would do if they believed
that a major recession was possible, without actually having to administer the lesson ".
At the time of the preparation of this policy document, the Swedish economist Gösta Rehn –
one of the two intellectual designers of the active labor market policies in Sweden – was head
of the OECD's man-power section.
Within the OECD staff, Rehn argued against the
formulations in this document, maintaining that it would be widely interpreted as an official
recommendation for increasing levels of unemployment. His objections were however
overruled, and internationally the policy document came to be interpreted in the way Rehn
OECD's "fundamental problem" – how governments could avoid electoral blame
for increasing unemployment – found a surprising and radical solution. The 1973 oil shock,
released by the Organization for Petroleum Exporting Countries (OPEC) after the Arab-Israeli
war, hit the Western nations almost like a natural disaster. In most countries, levels of
unemployment jumped up. This shock thus became the outside catalyst for the actual
administration of the lesson of high unemployment, a lesson the blame for which
governments now could largely place on the OPEC. In spite of increasing levels of
unemployment, in the McCracken report the OECD urged caution with expansive policies,
arguing that it is better to accept "a less rapid reduction of unemployment now, in order to
achieve lower levels of unemployment later on".
As noted above the Bretton Woods agreement had enabled governments to control cross-
border capital flows, something giving them considerable leverage in relation to national
capital interests and possibilities steer economies to maintain full employment. After the
dissolution of this system in 1971, from the mid-1970s to the mid-1980s in most of our
countries there was a wave of deregulations of capital movements. Governments thus came to
lose a major tool for full employment economic polices. The result was a globalization of
capital, a globalization which affected national relations of power in favor of capital and
limited the scope of government policy making.
During the 1970s and first years of the 1980s in many Western countries the top priority goals
of economic policy changed from full employment to low and stable inflation. This shift came
along quite different paths, some of which can here be sketched. In Europe, Germany with its
economic strength and independent central bank – Bundesbank – played a key role.
the dissolution of the Bretton Woods system in 1971, Bundesbank chose a tight monetary
policy increasing unemployment. When the first oil shock hit Germany, the coalition
government between social democrats and the small liberal party began a fiscal expansion to
stimulate investment and employment. Bundesbank responded by a very strict deflationary
policy, thereby creating a severe recession and forcing the government to revert to a tighter
As a result of the G7 summit in Bonn in the summer of 1978 urging countries
such as Germany to act as "locomotives" to pull Western countries out of the recession, the
government again moved to an expansionary policy. But when the second oil shock hit the
country in 1979, increasing unemployment was combined with deficits in government
budgets as well as in the current account. Bundesbank now saw this experiment as a failure
and reverted to monetary stringency. The electoral popularity of the social democrats sank
with increasing unemployment, a development increasing disagreements on economic policy
within the coalition government. The liberals now made their crucial "Wende", turned their
backs to the social democrats and entered a coalition government with the christian democrats
in 1982, this new coalition elevating low inflation to its major goal.
In the United Kingdom, the Labour governments (1964-70 and 1974-79) struggled to defend
the pound as a reserve currency and to fight inflation and unemployment by attempting
incomes policies, but were unable to create working alliances with trade unions.
inflation, increasing unemployment and a winter of severe strikes, the Conservatives came to
power in 1979, with Margaret Thatcher as a convinced neo-liberal leader keyed to anti-
inflationary policy. In France, during the 1970s the Socialists, long outside government,
criticized the ruling Gaullists for increasing unemployment by too market-oriented policies.
Under Mitterand, they almost won the presidential election in 1974 and finally did it in 1981.
Parliamentary elections in the same year led to a left government. Planning on the basis of an
expected US-led recovery in 1982, this government began an expansionary policy. In the
United States the new Republican administration and Federal Reserve did however delay its
expansion, something forcing Mitterand to make a U-turn in its economic policy in 1983.
With considerable internal conflicts over policy issues, during the late 1970s and early 1980s
major European countries thus changed economic policy goals, degrading full employment
and giving top priority to low and stable inflation. What were the driving forces behind this
shift? Layard, Nickell and Jackman argue that it was prompted by widespread electoral
demand; after the second oil shock electorates "in most countries declared that enough was
enough: inflation must be reduced. There followed massive demand deflation in all countries,
and by 1985 OECD inflation had been reduced to the same levels as existed in 1969. At the
same time, OECD unemployment had risen by over a half".
However, they give no
evidence for the assumption that the downgrading of full employment was largely driven by
Relevant empirical data on voters' reactions to inflation and unemployment have been
produced within research on "economic voting." These studies indicate that voters are
primarily influenced by their overall assessment of government economic performance rather
than by its specific aspects. They also indicate differences between the United States and
Europe, something we might expect in view of the transatlantic differences in the role
accorded full employment in the Beveridgian sense. In the United States, it would appear that
during the period 1960-1980, inflation rates had more effects on presidential popularity than
In Europe, however, where full employment below three percent was
political touchstone during the Great Trough, in the 1970s and 1980s
unemployment appears to have been at least as important as inflation for voters' party choice
and level of government support.
Surveys on most important political issues indicate that in
many countries electorates appear to have viewed unemployment as about equally important
or most of the time even more important than inflation.
Thus in the United Kingdom when asked: "What would you say is the most important issue
facing Britain today?", from 1977 to 1980 – with unemployment increasing to about six
percent and inflation more than twice as high – although inflation is pointed out by two thirds
of the respondents, unemployment is mentioned by about one half of them. During the first
half of the 1980s, when government policy had shifted from full employment to low inflation,
unemployment is mentioned as the most important issue by as much as 80 percent of the
respondents, inflation only by about 20 percent.
Also German election studies from 1976,
1980, and 1990 indicate that problems related to unemployment were mentioned as most
important issues much more often than inflation and prices. In Denmark, where inflation
increased to about 10 percent already after the first oil shock but unemployment only reached
this level after the second shock, unemployment was clearly seen as the most important
political problem already after the first shock while inflation was rarely mentioned.
1979-85, when Sweden had an unemployment rate below three percent but an inflation rate at
least three times as high, the employment issue was consistently mentioned among the two
most prominent issues for party choice while inflation was seen as much less important.
It would therefore not appear possible to view the downgrading of full employment in favor
of low inflation as a response by European governments to clear electoral demands and
It is often argued that while inflation hurts all citizens, unemployment only hits small
minorities. In this context the reliance on labor force surveys giving averages of weekly
unemployment rates can be misleading, since during longer time periods larger proportions of
individuals are likely to be sometimes unemployed. A panel study in Sweden showed that
among persons 24-55 years in 1991, during the period 1992-2000 close to 40 percent were
sometime hit by unemployment; for three quarters of them unemployment lasted more than
A possible explanation for this shift in policy goals is that high inflation has turned out to be a
greater menace for citizens' welfare than the negative consequences of high unemployment
documented in sociological, psychological and medical research (cf above). In view of the
role now widely accorded to low inflation as
goal for macro-economic policy, one could
expect it to be easy to pinpoint serious negative consequences of inflation. Apart from short
periods of hyper-inflation in some countries, this task does however turn out to be surprisingly
difficult for the normal rates of inflation appearing in the OECD countries during the period
after the Second World War.
In fact, as observed in one macro-economics textbook:
economists have difficulties in identifying substantial costs of inflation"
The costs of
inflation are typically seen as generated, for example, in connection with the need to adjust
prices, in distortions of tax schedules, and as transfers of wealth from people holding cash to
those with real assets. Variable inflation may also generate uncertainty, and high rates of
inflation are disliked by electorates. While these factors are drawbacks, there are also
potential benefits of inflation. One is that as pointed out by Tobin, since it is difficult to cut
nominal wages, adjustments to external shocks are easier to achieve when inflation is allowed
to change real wages.
In this context it is therefore fruitful to consider the rarely discussed distributive consequences
of the shift in policy goals: Is it neutral in terms of the distribution of the national product
between employers and employees, between profits and wages? Here the weakness of a
theory of distribution in mainstream economics generates problems; most studies silently
bypass the question of distributive neutrality, a few note it before bypassing. Thus, fore
example, Layard, Nickell and Jackman state that
"unemployment is the mechanism which
assures that the claims on the national product [by employers and employees] are
, however without discussing the distributive consequences of the use of
variations in unemployment rates to achieve this compatibility.
Noting that unemployment
affects wages in the long run and is related to the level of autonomous upward pressures on
wages and prices, Nickell (1998:805) writes: "If wage pressures rise because unions become
more militant, say, policy makers will, in the long run, be forced to lower real demand in
order to stabilize inflation. … But it is the autonomous increase in wage pressure that is the
ultimate cause of the unemployment increase. And this wage pressure must be reduced if
unemployment is to be reduced without rising inflation."
Here the role of increased
unemployment as a means towards lower inflation is recognized but distributive consequences
and relative costs of alternative policies are not discussed, and policy makers are seen as
impartial umpires bringing unions to order by contracting the economy.
While we must recognize inflation as a problem in itself, the inverse statistical association
found between unemployment and wages clearly indicates that an increase of unemployment
is likely to affect the bargaining positions of employees and employers in opposite directions,
reducing the bargaining power and aspirations of employees while increasing those of
, economic policies which work via variations in levels of
unemployment to ensure that inflation remains stable are thus likely to have distributive
consequences. When the goal of economic policy changes from full employment to stable
inflation, we can thus expect employees to be on the losing side. In this context it is therefore
fruitful to consider the hypothesis that distributive considerations have been one potential
driving force behind demand deflation.
In the mid-1980s, Gösta Rehn proposed such an alternative interpretation for this shift in
policy goals. In his view the relatively mild electoral repercussions after the first oil shock
strengthened the courage of many policy makers. As put by Rehn: "It is possible that the
sudden increase in unemployment after the first oil price jump in 1973 was not planned by
governments. But when it occurred, they were pleased to see that unemployment was not as
dangerous politically as many had thought". In Rehn's terse summary, when governments
appeared determined to stick to the goal of low inflation in spite of greatly increasing levels of
unemployment, the economic policy stance of most Western governments now was:
employment is no longer permitted,"
with the implicit message that "manpower utilization
would be kept so much below full employment that the trade unions and their members were
constantly reminded of 'the lesson' ". In this context Rehn 1987:65) also notes: "Of course no
government admits that high unemployment is a goal or a central means for the achievement
of price stability and a desired rate of economic growth."
For obvious reasons governments
find it politically expedient instead to pronounce low inflation as their primary policy goal.
As noted above, Kalecki argued that in the long run business interest would not
accept full employment.
We can now formulate what can be called the
: The high unemployment rates after 1973 partly reflect attempts by business and
conservative interests to reshape relations of power and patterns of distribution prevailing
during the full employment era into more favorable ones from their point of view. Instead of
being the major problem, unemployment thus comes to be seen a solution to other problems
now considered more serious.
8. PATTERNS IN THE RETURN OF MASS UNEMPLOYMENT
The return of mass unemployment after the oil shocks of 1973 and 1979 shows clear
differences among groups of countries
. The most dramatic increases took place in
In our eight countries which at that time were members of the European Economic
, the average increased from below three percentage points during The
Great Trough to nine percentage points in 1980-95. Among our members of the European
Free Trade Agreement (EFTA)
this increase was much lower; here the dramatic rise came
only in the early 1990s. In the United States there was only a mild increase with a couple of
percentage points from the relatively high levels during The Great Trough.
(Figure 7 about here)
These three distinctive patterns of long-term changes point to problems with the standard
explanation for the "stylized fact" that in the 1980s levels of unemployment in the United
States were lower than those in Europe because of the lower levels of labor market rigidities
and distortions in the United States. Such an explanation in terms of relatively durable
patterns of rigidities and distortions can not easily account for the
unemployment in the United States during The Great Trough. What cries out for an
explanation here is instead the remarkable shift in Europe, bifurcated into two different
patterns; in the United States no basic long-term changes would appear to have taken place.
The above patterns in the return of mass unemployment after the oil shocks are likely to
reflect not only the severity of shocks but also government policies used to cope with
economic strains. Here we can expect that differences in the relative strength of political
parties along the left-right continuum as well as long-term models for the resolution of
conflicts of interests were of importance for the strategies chosen by governments to handle
pressures generated in connection with these shocks. This leads us to into the "varieties of
capitalism" debate. To indicate differences among countries, scholars have used terms such as
"neo-corporatism," concertation, political bargaining, social partnership, and centralized wage
Such concepts capture partially differing country traits, which themselves are
likely to be only partially stable over time. In this context it appears fruitful to distinguish
between, on the one hand, a group of countries where a long-term presence of left parties in
governments in combination with constitutional factors has resulted in what can be described
as macro-political tri-partite bargaining and, on the other hand, categories of countries with
differing levels of centralized wage bargaining.
Macro-political bargaining involving governments, unions and employers' organizations can
be seen as a pattern for conflict settlements generated primarily in countries where left parties
have had a long-standing and strong control over governments and where unions are strong
and centralized. In the period considered here, these conditions can be said to have existed in
Austria, Norway, and Sweden. Such a pattern of bargaining can however also appear in
countries where left parties are weaker but where constitutions give minority parties a pivotal
role in political decisionmaking. Here Finland and Switzerland are the prime cases. The
Finnish constitution has required large majorities in the
for significant legislation
to become valid for more than one year, thus favoring broad coalition governments.
the early 1970s, Finland thus had a relatively stable Social Democratic minority participation
in different coalition governments and long periods of state involvement in incomes policies.
Switzerland is an exceptional case for many reasons, one of them being that Swiss
constitution allows small interest groups to introduce referendums and thus to render the
political decisionmaking process very cumbersome and uncertain.
This has driven the
conservative and liberal parties as well as business interests to bring in the relatively small
Social Democratic party and the weak unions into formal and informal negotiations before
decisions are taken, and to give the Social Democrats a minority representation in the
practically permanent coalition government, the
In discussions of the role of wage-setting procedures for unemployment levels, social
scientists have pointed to the potential relevance of centralized wage bargaining with
encompassing unions which are likely to practice wage restraint, thereby partly internalizing
the consequences of wage demands on employment levels.
Using the Calmfors-Drifill
classification, among our remaining countries a relatively high degree of centralized
bargaining can be said to have existed in Denmark, Germany, the Netherlands and Ireland, a
medium level in Australia, Belgium and New Zealand, but a low level in Canada, France,
Italy, Japan, the United Kingdom and the United States.
The pattern of changes in average levels of unemployment from The Great Trough (1960-72)
to the period after the two oil shocks (1980-9) show that on the average, the smallest
percentage point increases came in the five macro-political bargaining countries (1.6 percent)
. Among these countries, during this period changes in government complexion did
not have major consequences for unemployment levels.
The bourgeois governments in
Sweden (1976-82) and in Norway (1981-86) came to power when long periods of full
employment had become associated with and also largely maintained by social democrats.
Full employment was thus seen by voters not only as desirable but also as a realistic policy
goal. The new bourgeois governments therefore faced strong strategic incentives not to be the
first ones to allow major increases in unemployment, something which at that time would
have been likely to discredit them for a long time to come.
In Norway, the exploitation of
North Sea oil contributed to reverse the oil shock. The extremely low level of unemployment
in Switzerland reflects a very different type of class compromise, where full employment was
limited to the native male labor force and the brunt of unemployment was borne by the so-
called guest workers, a large proportion of which had to leave the country.
(Table 2 about here)
In the remaining countries, the degree of centralization of wage bargaining does not appear to
have been of relevance for the extent of unemployment increases.
Countries classified as
having high or medium levels of centralized bargaining had large increases. An exception
here is New Zealand with a history of extremely low levels of unemployment during the post-
war period, ending up with a relatively low unemployment in the 1980s. Among countries
with low levels of centralized bargaining, increases were relatively modest in Italy, Canada
and the United States, countries with relatively high unemployment rates already during The
Great Trough. Here the major exception with low levels of unemployment is Japan, which has
been described as having state-led capitalism (Pempel 1982). It differs from the Anglo-Saxon
as well as the European forms by combining a long-term exclusion of left parties from
government position with strong state-involvement in the economy via planning and co-
ordination. In Japan, the first oil shock came when popular support for the ruling Liberal
Democratic Party was crumbling for the first time since the period after the Second World
War. The government had a strong incentive, as well as a capacity to avoid drastic increases
in unemployment. In this context the distinctive organization of employment relations in
Japan may also have been of relevance.
The trends noted above were however not to
remain stable. Thus during the early 1990s, apart from France with an already high
unemployment level, the largest increases came in Finland and Sweden, where full
employment gave way to mass unemployment. Also New Zealand saw a major increase
during this decade.
The patterns of the return of high unemployment during the 1970s and 1980s indicate that in
itself, a high degree of centralization in wage bargaining does not appear to have been of
major importance. It was primarily in Austria, Sweden and Norway, where a combination of
centralized wage bargaining and a long-lasting left hold of government still generated a
"contagion from the left" that lead to policies largely retaining full employment. In Finland as
well as in Switzerland with a weaker left, the rise of unemployment was also largely
contained, but there constitutions were essential in generating macro-political bargaining.
To what extent were country differences in the rise of unemployment after The Great Trough
affected by differences in economic policy? Because of various limitations, the analysis here
will be restricted to one economic policy indicator, the short-term interest rates set by central
banks, exemplified by the prime rate of the Federal Reserve Bank in the United States. The
setting of prime rates by central banks is of course affected by many factors. In European
countries, for example, a concern with exchange rates has often been important. Yet this type
of short-term rate has increasingly come to be used by central banks as a main instrument for
controlling inflation and indirectly also growth and employment. It can thus be seen as one
central indicator of the macro-economic policy stance in each country.
the development of annual real short-term central bank rates since 1961 are given
for eight countries of the EEC and four countries of the EFTA as well as for United States.
During 1961-72, in the EEC and EFTA we find average real interest rates fluctuating around
zero (-0.2). In the United States as well as in Canada, real prime rates were however
considerably higher, 2.3 and 1.6 percent, respectively. In this period, the two North American
countries with considerably higher unemployment rates than Western Europe thus appear to
have had more restrictive policy stances than what was the case in countries with full
(Figure 8 about here)
Along with accelerating inflation after 1973-79, real interest rates fell and were clearly
negative in the EFTA countries (-3.4 percent) and also in the EEC (-2.0 percent). Again in the
United States the average real prime rate was considerably higher (1.4 percent), while Canada
now was down to -0.9 percent. After the second oil shock in 1979, however, real interest start
to climb, largely reflecting drastic nominal increases in the EEC countries as well as in the
United States. In the EFTA countries real interests rates show only some increase, about 1.0
percentage points up to the late 1980s.
To what extent are differences in macro-policy stances as indicated by increases in central
bank prime rates reflected in changing levels of unemployment? We will here concentrate on
shifts taking place in connection of the second oil shock in 1979, when most countries raised
nominal interest rates. With only annual information on interest rates, changes in trends can
be only crudely indicated. We here look at the largest increase in interest rates observed
between consecutive two-year periods from 1977 to 1982, increases which for most of our
countries came from 1977-78 to 1979-80. The relevant increase in unemployment can be
expected to come with some time lag and is here taken as the increase of average levels of
unemployment from 1975-82 to 1983-84.
In this context the two largest economies, the
United States and Japan, are clearly outliers. Both countries made major increases in their real
interest rates, but as noted below in the United States with considerable short-term variation.
In the United States average annual interest rates increased from already relatively high
levels; in Japan the increase came from an originally negative level. In neither country did
unemployment levels however show much change; in the United States it did not increase
much from a relatively high level and in Japan it remained at a very low level. We will here
disregard these two cases as well as Norway, where the oil economy makes interpretations
difficult. Among the remaining fifteen countries, we find a fairly good correlation (0.68)
between changes in real short-term interest rates in connection with the second oil shock and
changes in unemployment after this shock
. With the exception of France, real
prime interest rates increased clearly less in the four macro-political tri-partite bargaining
countries (Austria, Finland, Sweden and Switzerland) than in the others.
It would thus
appear that partisan politics and the choice of political strategy in reaction to the external
shocks were of considerable importance for the pattern of return of mass unemployment
during the 1980s.
As noted above, however, in the early 1990s unemployment levels jumped up in Finland and
In both countries the fast escalation of unemployment was associated with marked
increases in real interest rates.
In Sweden the background was a major economic crises
precipitated by policy mistakes and a change in macro-economic policy stance, where the
traditional top priority to full employment was replaced by top priority for low inflation.
Finland the background was partially similar, however to some extent aggravated by the
collapse of the Soviet Union, its major trading partner. In both countries, interest rate
increases were also part of attempts to defend the exchange rate. Yet, when these attempts
turned out as failures, both countries continued very restrictive policies. A policy orientation
directed towards qualification for membership in the new European Union and its monetary
cooperation was probably of relevance for these policy shifts.
During the period 1979-1990, a significant part of difference in levels of unemployment
within Europe between our EFTA and EEC countries can thus be interpreted in terms of the
considerably higher interest rates set by central banks in the EEC. Can differences in central
bank monetary policies also account for some part of the differences in unemployment during
the 1980s between the United States and the EEC? Here we have to consider not only the
explicit goals set by formal instructions to central banks but also the strength of the potential
resistance they meet in carrying out their goals. In this period, the average interest rates was
4.8 percent in the EEC and 5.1 percent in the United States. As noted by Ball, however, in the
United States this high average hides a pattern of frequent changes, where sharp increases
were followed by marked decreases when inflation had slowed down.
In the EEC, on the
contrary, interest rates were kept high during long periods in spite of decreasing inflation and
high unemployment levels.
These differences partly reflect that the Federal Reserve Bank has to consider not only price
stability but also growth and employment. In Europe as noted above, the German
Bundesbank, which traditionally has set the pattern for central banks in many other European
countries, has since 1971 instead persistently focused on maintaining low inflation. But here
also differences in societal power structures are of relevance. In their roles as guardians of a
stable currency, central banks have to face the pressure on wages and prices emanating from
trade unions as well as from potentially interventionist governments. These pressures on
wages and prices are likely to be considerably stronger in most European countries than in the
United States. The confrontations between relatively independent central banks, unions and
governments of different complexions in setting the agenda for economic policy are therefore
likely to unfold differently in the United States and Europe, and also to vary among European
countries. Where central banks and conservative-centrist governments keyed on maintaining
low inflation face strong unions and a potentially interventionist opposition parties, they are
likely to have to take recourse to tougher "worker disciplinary devices" than is the case where
the opposition is more malleable.
Observations made by Carlin and Soskice on what they refer to as the "threat strategy" of the
German Bundesbank suggest such a hypothesis:
"Faced by potential inflationary demands from … powerful actors, the Bundesbank's
strategy is to make threats which actors then have to take into account. To contain
union wage demands and to stiffen the resistance of the employers' associations, the
Bundesbank's threat is to raise nominal interest rates and possibly produce an
appreciation of the exchange rate. … Moreover, if inflationary wage increases occur,
the Bundesbank will need to rebuild its anti-inflationary reputation. It is therefore
unwilling to respond to the economic slack it causes by symmetric cuts in interest
rates. If punishment is to be meted out, policy settings can only be relaxed very
gradually and this fails to provide the necessary confidence about the growth of
demand to drive up investment."
Carlin and Soskice note that a punishment strategy of this type makes the Bundesbank differ
from the Federal Reserve Bank in the United States, "where the absence of powerful unions
makes a punishment phase unnecessary so that a fall in inflation to some target levels is the
signal for the relaxation of monetary conditions."
Here we thus find an explanation of the
differences observed by Ball in the setting of interest rates by the Fed and European central
banks. During the 1970s and 1980s, the Bundesbank played a leading role in setting policy
strategies also for other central banks in within the EEC.
Based on this observation by Carlin and Soskice we can suggest the hypothesis that the
exceptionally high levels of unemployment within the EEC during the 1980s to a significant
extent reflected a conflict strategy by central banks and governments focusing on maintaining
low inflation to counteract pressures from strong unions and potentially interventionist
political parties. This hypothesis thus goes against conventional interpretations that the higher
unemployment rates in Europe compared to those in the United States are due to higher levels
of rigidities and distortions on European labor markets. In contrast to the rigidity hypotheses,
this conflict hypothesis can also explain the lower levels of unemployment in the EFTA
countries; as seen above, in these countries there was some degree of cooperation between
unions, governments and central banks helping to keep interest rates relatively low during this
8. CHANGING PROFIT/WAGE RATIOS
From the perspective of the role of unemployment in distributive conflict, a key question is to
what extent the changes in levels of unemployment generating The Great Trough have been
reflected in the functional distribution of national income between employers and employees.
The wage curve indicates that lower levels of unemployment are associated with higher
wages. If employers can only pass on part of wage increases to consumers in the form of
higher prices, we could expect unemployment levels to affect the profit/wage ratio. This
hypothesis can be illuminated by data from national accounts on the distribution of the gross
domestic product (GDP). Since 1960, OECD national accounts include a breakdown of the
cost components of the GDP providing estimates of two here relevant components. One is
compensation paid by employers to employees; the other is operational surplus, referring to
the value of what is left to employers of gross output once they have paid compensation to
employees and intermediate production costs.
These two components will here be used to
define the relative size of the parts of national product going to employers and employees in
terms of the profit/wage ratio.
As generally is the case in comparative studies, there are problems associated with available
data. Since operational surplus includes incomes of the self-employed, changes in the
structure of the labor force, especially the marked decline of the proportion of farmers in most
of our countries since 1960 will contribute to a decline of the profit/wage ratio.
interest here is however if this ratio will show a U-shape, that is if it falls from 1960 to about
1980 but thereafter has an increasing trend up to the late 1990s. During last couple of decades,
the decline of the farm population is of less importance and will furthermore in effect
counteract the above hypothesis. In spite of efforts at standardization, problems with
definition and measurement are probably important. These problems are likely to generate
differences between countries especially when it comes to the level of the profit/wage ratio.
To the extent that countries have been consistent in the handling of national data, such
differences will be of less importance for within-country trends, which here are of main
To crudely describe the development of the profit/wage ratio over the period 1960-1998, we
have regressed this ratio on year and a quadratic term of year. The results indicate relatively
similar U-shaped patterns for almost all countries; the profit/wage ratio first decreases,
reaches a minimum in the 1980s and thereafter shows an increase
). This U-shape is
especially pronounced in Ireland and New Zealand, while in Canada it is nearly absent. The
quadratic component is small also in the United States and in the United Kingdom, partly
reflecting that in these two countries there was only a modest decline of the profit / wage ratio
in the years up to the 1980s. Japan and Switzerland have somewhat deviant developments,
with a declining trend stagnating in the 1990s, however without a marked upturn. Norway has
a W-shaped profile, which may reflect changes in the oil economy of the country.
(Table 3 about here)
To illustrate these patterns, average values for the profit/wage ratio 1960-1998 are shown for
four categories of countries in
. The "EEC" category includes Belgium, Denmark,
France, Germany, Ireland, Italy, the Netherlands, and the United Kingdom. Because of the
deviant patterns for Norway and Switzerland, the "EFTA" group includes only Austria,
Finland and Sweden, while the United States forms a category of its own. The "EEC"
countries show a marked U-shape with the lowest levels in the years up to 1980.
average the three "EFTA" countries have a decline up to the late 1970s followed by a
relatively flat development but then an increase after 1990. In the United States, both the
decline of the profit/wage ratio up to 1982 and the following increase are relatively modest.
(Figure 10 about here)
The above data thus indicate that in most of our countries there was a marked profit squeeze
during the couple of decades up to and around 1980, while thereafter the profit/wage ratio has
tended to increase. Although the decline in the farm sector and other types of changes in
economies undoubtedly are of relevance here, in countries where information is available on
manufacturing, relatively similar patterns of changes in the profit/wage ratios are found in
most of our countries, data which also reproduce increases after about 1980.
countries we thus find that the profit squeeze was turned into profit growth in tandem with
changes in economic policy stances to low inflation, the drastic increase in rates of
unemployment, and the decline in industrial disputes. In Finland and Sweden the most marked
increase of the profit/wage ratio came in connection with the return of mass unemployment
after 1990. The relatively small changes in this ratio in the United States and Canada may
possibly reflect that they have had relatively small changes in levels of unemployment during
the post-war period. Developments in Japan and Switzerland are difficult to interpret.
9. THE SILENT COUNTER-REVOLUTION
In this paper we have approached the radical changes in the political economies in the
Western countries during the past century with a focus on the role of distributive conflict as
reflected in long-term changes in levels of unemployment, a focus which opens up new
perspectives on these changes. In the analysis we have contrasted an economic supply-and-
demand framework keyed to market dysfunctioning with a power sensitive approach focusing
on long-term positive-sum conflicts among differently endowed actors and interest groups.
Often characterized by elegance in mathematical modeling, analyses in the neo-classical
supply-and-demand mold have concentrated on the return of mass unemployment after 1973,
viewing labor market distortions and rigidities hindering wages to adjust to market-clearing
levels as central causes. Although in many ways valuable, these studies must be seen as
partial. Important limitations are that they have come to largely neglect the element of
distributive conflict in the relationship between the parties on the labor market, give scant
attention to the role of power differences in this relationship, approach the labor market as any
other market, and do not consider the longer time perspective.
– unemployment insurance – resurrected to create one principal part in
neo-classically inspired accounts of the return of mass unemployment was ill-prepared for a
confrontation with basic facts from earlier periods. Thus over the long haul, the development
of unemployment insurance moves contrary to the trend in unemployment levels. Before the
Second World War when this insurance was absent or underdeveloped, our countries were
plagued by high unemployment; when it became universal and increasingly generous during
the 1950s and 1960s, in most of our countries unemployment fell to historically low levels.
The return of mass unemployment after the 1970s shows little correlation with the provisions
of this insurance. The standard interpretation that in a comparison with Europe, the relatively
low unemployment rates since the 1980s in the United States reflects its lower levels of labor
market distortions is called into question by the fact that up to 1973 the United States had
considerably higher unemployment than Europe. Furthermore, within Europe during the
1970s and 1980s, the EFTA countries had much lower rates than those in the EEC, a contrast
which is difficult to interpret in terms of differences in rigidities. Yet, the distortion and
rigidity hypotheses must of course not be rejected
. Undoubtedly we can expect some
degree of misuse of unemployment insurance as of other insurance programs, this insurance
may to some extent prolong individual spells of unemployment, and labor market legislation
may generate some degree of unemployment. Yet the main forces driving the end of The
Great Trough are likely to be located elsewhere.
Scholars working within the supply-and-demand framework have made conflicts of interests
between employed insiders and unemployed outsiders into a major area of research on
unemployment. In contrast yet the potential faultline between employers and employees has
been given short weight. Since profits as well as wages have to be taken from the revenue of
firm sales, we can not discuss wages without at the same time dealing with profits. And we
know that levels of wages tend to be inversely related to levels of unemployment. In market
economies it is of course necessary to have a place for profits, but the unresolved question is
what this place is to be. Therefore we have to face the reality of distributive conflict on the
labor market, and the fact that political means can be used to influence the level of
unemployment in ways which are likely to be of relevance for wages and profits. Those
viewing Western democracies as societies in basic harmony, where free competition between
equals generates welfare for all, may see the underlining of the political element in the rise of
unemployment as an accusation of conspiracy. At a time when most if not all governments in
the Western countries in practice have given up aspirations to reach full employment defined
in the traditional Beveridgian sense of the concept, reminders of this recent transition easily
appears politically inconvenient. But to close our eyes for the reality of distributive conflict in
society would jeopardize the objectivity of social science. In this context we must however
recognize that for employers, the choice of a preferred level of unemployment is complicated
by the fact that boom periods with lower unemployment also tend to be associated with better
profit levels. Here it is fruitful to make a distinction between the short and the long run; this
choice situation is likely to become especially acute after long periods of full employment.
While some economists such as Alchian and Demsetz do not believe that the labor market
differs from ordinary markets in essential ways, Solow notes that "labor as a commodity is
sufficiently different from artichokes and rental apartments to require a different mode of
analysis" and concludes that "the labor market is really different."
One basic characteristic
making labor markets different is that unlike artichokes, commodities on the labor market
have the capacity to learn and to act collectively. Employers as well as employees are likely to
have learned the lesson summarized by
in 1943 that the level of unemployment is
of major significance for relations of power between them and therefore for distributive
outcomes. Furthermore, history tells us that employees have recognized that they have a
relatively more advantageous power position in democratic politics where votes count, than
they have in markets where economic resources decide. Within a neo-classical framework, it
is difficult to rationally account for attempts by employees to act collectively via political
parties and unions and to use democratic politics to affect conditions for and outcomes of
distributive processes on the labor market. Viewing institutions such as unemployment
insurance as generating distortions easily implies that they are seen as largely irrational;
without them we would all be better off. Attempts at rational accounts of such institutions are
considerably facilitated by a recognition of significant contrasts between the socio-economic
patterning of the main power resources on markets and in democratic politics.
When we extend the analyses to include not only the end of The Great Trough but also its
beginning and continuation, it appears clear that here the role of politics is likely to have been
of major importance. Thus while a manifold of factors undoubtedly affected and interacted at
the beginning of this trough, a reasonable interpretation is that the arrival of full employment
after the Second World War was made possible by the sea change in political power when for
the first time in history, left parties with a low level of unemployment as their primary goal
emerged as major government parties or opposition parties in most of our countries. Yet,
however, the change into full employment policies did not reflect a convergence of political
goals; conservative and centrist parties maintained their traditional first preference for low
inflation rather than full employment. In the context of Bretton Woods institutions and the
presence of strong government alternatives seen as willing and able to support full
employment, one important part of the motivation for governments to maintain full
employment was instead the perceived – and to a considerable extent realistic – threat from
the electorate to punish policies increasing unemployment. In retrospect it is perhaps
surprising that the Damoclean sword of electoral punishment held sway over politics for such
a long time and allowed for full employment in most Western countries; North America
however being located towards the periphery of the effective Damoclean reach. In Western
Europe, much of the scenario once sketched by Kalecki now became a reality.
in industrial disputes increased to historically very high levels, and in some countries
managerial prerogatives were seriously questioned and political stability appeared to be
In discussions of the return of mass unemployment after 1973, the two oil-shocks have widely
been seen as main causes; the present analysis indicates that they can instead be seen as
catalysts. As we have seen, already before the first oil shock the OECD argued that full
employment was the decisive cause behind the rise of inflation and employer costs and
urgently recommended that top priority should instead be given to price stability. The fact that
this policy report was published after some internal conflicts among its top staff members
indicates that it was not produced by accident. The desire by several influential interest groups
to dismantle full employment from its top priority position had thus matured already by the
late 1960s. While OECD policy advisers pondered over how to create the fear of a recession
without an electoral backlash, OPEC's two oil shocks jolted western economies and thereby
opened windows of opportunity for a major policy change. As indicated by Rehn, it should
come as no surprise if several conservative-centrist governments were eager to avail
themselves of this window of opportunity opened up by the OPEC to change economic
policy goals to better suit their traditional preferences.
During the following decades our
countries saw intensive political conflicts centered around economic policy of relevance for
unemployment; in most of them the outcome was increasing unemployment. If the beginning
of The Great Trough in the 1950s can be seen as a silent revolution, its demise since the 1970s
is an almost equally silent counter-revolution. While these two revolutions have had only mild
repercussions in North America, they have decisively reshaped the European political
economy and institutions.
In this perspective, the light-weight nature of arguments advanced to explain the
pronouncement of low and stable inflation as the main policy goal in most Western countries
poses a riddle. Thus it is hardly possible to argue that this shift was a response to widely
shared electoral demands, or that inflation has more negative consequences for public welfare
than does unemployment. Assuming boundedly rational actors, it would instead appear that
the downgrading of full employment reflects traditional distributive conflicts and
longstanding policy preferences by conservative and centrist parties, preferences which could
find expression after the OPEC-generated shocks. We have also seen that this shift in policy
goals parallels the upward turn in the profit/wage ration since the 1980s, a decline of
industrial conflict, and as is wellknown, in many countries also an increasing inequality in the
distribution of disposable household income. The observation by Rehn that "of course no
government admits that high unemployment is a goal or a central means for achieving price
stability and a desired rate of economic growth" is likely to provide a significant clue for
governments publicly announcing low inflation as the main goal variable.
While hypotheses on labor market dysfunctioning fail to account for differences between the
United States and Europe both during The Great Trough and after its demise as well as for
differences between EEC and EFTA, a focus on the role of the distribution of political power
and differences in economic policy can do it to a considerable extent. We have seen that up to
1973, central bank real interest rates as well as levels of unemployment were considerably
higher in the United States and Canada than in Europe. As partly reflected in changes of real
interest rates in the years around 1980, inter-country differences in economic policy stances
by central banks and governments appear to have been of importance for differences in the
rise of unemployment rates after the second oil shock. Here the divergence between the EFTA
and EEC countries indicate that centralized bargaining was of relevance for delaying mass
unemployment only in the context of strongly entrenched left governments or constitutions
generating broadbased coalitions.
In accounting for differences between the United States and the EEC countries in terms of the
persistence of high unemployment rates, it is fruitful to consider differences in the strength of
the main actors that central banks and governments have to confront in attempting to reach the
goal of low inflation. In Europe, central banks face strong unions and potentially
interventionist governments and opposition parties. Based on observations on actions by the
German Bundesbank, we have suggested the hypothesis that to keep these strong players in
check, European central banks may have to rely on a punishment strategy requiring prolonged
periods of high real interest rates and economic contraction; such a strategy is unnecessary for
the Federal Reserve Bank facing the weak unions in the United States. The differences
between the EEC and the EFTA countries in levels of unemployment in the 1980s reflect
differences in economic policies but partly also the relatively strong position of left parties
and unions long making the threat of electoral retaliation for high unemployment a reality in
most of the EFTA countries. In the 1990s, adherence to the Maastricht convergence rules are
likely to have contributed to the rise and persistence of unemployment in the European Union.
The European Monetary Union and the European Central Bank are institutions which will be
of major importance for the future development of unemployment. This central bank is
modeled after the German Bundesbank and has an equal devotion to low and stable
The shift of top priority goals from full employment to low inflation has involved a major
reshaping of international and national institutions determining societal power relations.
Under Bretton Woods institutions and a Keynesian policy model with low unemployment,
unions and employees are partners in formal or informal macro-level tri-lateral bargaining
with employers and governments. In this bargaining the power of employees rests on their
capacity for exerting wage pressure but also on their capacity to moderate their wage
pressures in exchange for governments' use of monetary and fiscal policy to maintain full
employment, if necessary by demand reflation. In this context the capacity of national
governments are to a significant extent dependent on their possibilities to control cross-border
capital flows. With deregulated capital movements and a monetarist policy model, the central
bank alone is in charge of controlling inflation; since low inflation is the only goal variable,
whenever necessary this goal can be achieved by demand contraction functioning as a worker
disciplinary device. The globalization of capital means that the potential contribution of
unions to wage restraint has now lost much of its value as currency in political exchange. In
this setting unions can maintain employment primarily by keeping wage increases so low that
the central bank does not find it necessary to act. The action alternatives by unions and
employees are now greatly restricted; if they exert their potential power, the central bank can
punish them by higher unemployment. These institutional changes setting the framework for
the exertion of power by major interest groups are likely have important distributive
consequences, enabling some actors while restricting others. To understand these
developments it is necessary to complement analyses in the supply-and-demand framework
with a focus on positive-sum distributive conflicts among major interests groups differently
positioned in terms of power.
Average Annual Correlations Between Unemployment Rates and the Net Replacement Rate
and Benefit Duration in Unemployment Insurance During Three Time Periods in 18
Period* Net replacement rate Duration
1960-72 -0.33 -0.24
1975-87 -0.17 0.11
1990-97 -0.36 0.28
* The number of annual correlations in the three time periods are 9, 9, and 6, respectively
Changes in and Levels of Unemployment 1960-98 in Countries with Tri-Partite Macro-
Political Bargaining and with Different Levels of Centralized Wage Bargaining in 18
Fit of Regressions of the Profit/Wage Ratio on Time, 1960-1998, in 18 Countries
Constant Year (t) Year
Australia 60.7 -1.80 (-7.2) 4.09 (6.7) 0.57
Austria 71.0 -2.55 (-14.6) 4.89 (11.5) 0.90
Belgium 79.7 -3.15 (-12.6) 6.83 (11.2) 0.82
Canada 45.8 -0.46 (-2.2) 0.61 (1.2) 0.31
Denmark 61.4 -2.70 (-18.1) 5.24 (14.5) 0.93
Finland 77.2 -3.39 (-11.6) 6.49 (9.1) 0.85
France 79.0 -3.05 (-11.5) 5.68 (8.9) 0.85
Germany 68.3 -2.68 (-15.0) 5.35 (12.3) 0.89
Ireland 87.6 -4.01 (-13.2) 10.33 (14.0) 0.84
Italy 99.1 -2.86 (-8.9) 6.87 (8.7) 0.67
Japan 107.6 -3.75 (-10.6) 5.11 (5.9) 0.92
Netherlands 69.5 -2.69 (-9.7) 6.29 (9.3) 0.71
New Zealand 103.0 -4.12 (-12.8) 9.13 (11.7) 0.82
Norway 64.5 -1.79 (-3.0) 4.06 (2.8) 0.16
Sweden 43.5 -1.92 (-9.2) 4.22 (8.4) 0.70
Switzerland 62.5 -1.64 (-13.4) 2.16 (7.3) 0.95
United Kingdom 37.8 -1.01 (-5.8) 2.87 (6.8) 0.59
USA 46.2 -1.25 (-9.6) 2.54 (8.1) 0.75
”EEC” 72.8 -2.77 (-17.8) 6.18 (16.3) 0.90
”EFTA” 63.9 -2.62 (-16.0) 5.20 (13.1) 0.90
: OECD: National Accounts
The Great Trough in unemployment
Average for Sweden and the United Kingdom;
Average for Australia, Austria, Belgium, Denmark,
Finland, France, Germany, Japan, the Netherlands, New Zealand,
Norway, Sweden, Switzerland and the United Kingdom.
Figure 2. Overall levels of dissatisfaction with life among unemployed
men and women and among employees (20-54 years of age) in
eight European countries 1977-1999.
Employed men and women
Source: Eurobarometer (Inglehart, Reif and Melich, 1992).
The Wage Curve: With decreasing local unemployment rates,
wage rates tend to increase.
Local unemployment rate
Left party percentage of time in governments, cabinets and
votes, and union density during five periods 1900-1995.
Average for 18 countries.
1900-13 1919-38 1946-72 1973-84 1985-95
eft government time
eft cabinet share
. Percentage unemployment 1960-73 in 18 countries
ranked by average left political strength 1919-1959
. Inflation, relative involvement in industrial conflict (percent)
and duration of disputes (days) during five time periods
1900-1997. Averages for 18 countries.
Reli CPI Duration
. Percent unemployment in Countries of EEC and EFTA
and in USA, 1955-2000.
Real short-term central bank interest rates in EEC, EFTA
and the United States 1961-1997
Percentage point changes in real short-term interest rates and
unemployment in 15 countries in connection with the second
The profit / wage ratio in four categories of countries
Source: OECD. National Accounts
*Work on this paper has been facilitated by generous help from many persons. Special thanks go
to Robert M. Solow for extensive comments an a draft of the manuscript. Earlier versions of the
paper have been presented at meetings of American Political Science Association, Research
Committee 19 on Poverty, Welfare and Social Policy of the International Sociological
Association, Stockholm-Turku Welfare State Workshop, Sociology Department of Gothenburg
University, Trade Union Institute for Economic Research in Stockholm, Labor economics seminar
at the Swedish Institute for Social Research, and European Social Science History Conference. I
want to thank participants in these meeting for valuable comments. For constructive help I also
thank Jan Otto Andersson, Rune Åberg, Klas Åmark, Miriam Cohen, Stefan Englund, Ingrid
Esser, Duncan Gallie, Michael Katz, Tomas Korpi, Ingalill Montanari, Annitta Näsström, Joakim
Palme, Jean Parr, Ola Sjöberg and Pat Thane as well as the editors of this journal. As usual
responsibility for remaining errors rests with the author. This research has been supported by the
Bank of Sweden Tercentennial Foundation.
Data sources are for 1921-38 Sven Hegelund, Lars Jonung, Erling Petersson, and Eskil
Wadensjö, “The Philips' Curve for Sweden? The Determinants of Wages and Prices in Sweden
1922-1971”, Lunds Universitet, Nationalekonomiska Institutionen, Lund, 1975, and C.H.
National Income, Expenditure and Output of the United Kingdom 1855-1965
(Cambridge: Cambridge University Press, 1972); for 1950-59 Angus Maddison,
in Capitalist Development. A Long-Run Comparative View
(Oxford: Oxford University Press,
1991), and for 1960-2000
Labour Force Statistics
(Paris: OECD, various years).
Social Insurance and Allied Services
(New York: The Macmillan Company,
Full Employment in a Free Society
(New York: W. W. Norton,1944), 93ff.
The Political Economics of Prosperity
(Washington, DC: The Brookings
The Future of Socialism
(London: Cape, 1956), 28.
Alan S. Blinder, “The Challenge of High Unemployment,”
American Economic Review
78, no. 2
These works include, for example, Geoffrey Garrett,
Partisan Politics in the Global
,(Cambridge, Cambrige University Press, 1998); J. Rogers Hollingsworth and Robert
Contemporary Capitalism: The Embeddedness of Institutions
University Press, 1997); Torben Iversen, Jonas Pontusson, and David W. Soskice,
Employers and Central Banks: Macroeconomic Coordination and Institutional Change in Social
(Cambridge: Cambridge University Press, 2000); Herbert Kitschelt,
Transformation of European Social Democracy
(Cambridge: Cambridge University Press, 1994);
Herbert Kitschelt, Peter Lange, Gary Marks, and John D. Stephens, ed.,
Continuity and Change in
. Cambridge: Cambridge University Press, 1999); Fritz W. Scharpf and
Vivien A. Schmidt, ed.,
Welfare and Work in the Open Economy: From Vulnerability to
(Oxford: Oxford University Press, 2000); Fritz W. Scharpf and Vivien A.
Welfare and Work in the Open Economy: Diverse Responses to Common Challenges
(Oxford: Oxford University Press; 2000).
Thus, while there is a sizable scholarship exploring the causes of The Great Depression during
the first half of the 1930s, the causes of its inverse – the quarter century of The Great Trough –
have remained largely unexplored.
The selection of countries is thus based on the idea om "most comparable cases", see Arend
Lijphart, “The Comparable-Cases Strategy in Comparative Research,”
8, no 2, 1975),158-77.
The New Palgrave: A Dictionary of Economics,
John Eatwell, Murray Milgate and Peter
Newman ed., (London: Macmillan and New York: The Stockton Press, 1987); David L. Sills, ed.,
The International Encyclopedia of the Social Sciences
(New York:Collier-Macmillan, 1968). .
For a review, see David Soskice, “Reinterpreting Corporatism and Explaining Unemployment:
Co-ordinated and Non-co-ordinated Market Economies,” in Renato Brunetta and Carlo
Labour Relations and Economic Performance
(London: Macmillan, 1990),
chap. 7. Keynesian interpretations stressing the role of investment have also been advanced, see
for example Paul Davidson, “Post Keynesian Employment Analysis and the Macroeconomics of
The Economic Journal
108 May, 1998), 817-31.
For applications of this approach see, for example, Gösta Esping-Andersen,
Markets: The Social Democratic Road to Power
(New Jersey: Princeton University Press, 1985); .
The Working Class in Welfare Capitalism: Work, Unions and Politics in Sweden
(London: Routledge & Kegan Paul, 1978),
The Democratic Class Struggle
and Kegan Paul, 1983), “Power Resources Approach vs. Action and Conflict: On Causal and
Intentional Explanation in the Study of Power,”
3 (no 2):31-45, and
“Contentious Institutions: An Augmented Rational-Actor Analysis of the Origins and Path
Dependency of Welfare State Institutions in the Western Countries,”
Rationality and Society
no 2, (2001):235-83; John Myles,
Old Age in the Welfare State: The Political Economy of Public
(Lawrence, Kan.: University Press of Kansas, 1989); Julia S. O'Connor and Gregg M.
Power Resources Theory and the Welfare State. A Critical Approach
University of Toronto Press, 1998); John D. Stephens,
The Transition from Capitalism to
(London: Macmillan, 1979).
Robert M. Solow, “Unemployment in the United States and in Europe: A Contrast and the
CESifo Working Paper
(No. 231, 2000), 9, italics added).
Armen Alchian and Harold Demsetz, “Production , Information Costs, and Economic
American Economic Review,
62, (no. 5, 777-95), 777.
Quoted in Beveridge,
Full Employment in a Free Society
In contrast to "economic rationality," boundedly rationality assumes that actors are satisfying
rather than maximizing, have limited information and information-processing capability and need
not be exclusively focussed on material self-interest. In this connection the complications with
broad-based collective action pointed out long ago in Mancur Olson,
The Logic of Collective
Action : Public Goods and the Theory of Groups
(Cambridge, Mass.: Harvard University Press,
1965) should however be recognized, as well as the fact that employees are internally
heterogeneous. See John H. Goldthorpe,
On Sociology: Numbers, Narratives, and the Integration
of Research and Theory
(Oxford: Oxford University Press, 2000), chapter 10.
The term "distortion" carries a clear implication of twisting or wresting away from a natural or
true shape. Here the point of comparison is however not nature but the theoretical model of the
The reform-rigidity sides of the coin are to some extent overlapping in that reforms can have
unexpected and unwanted consequences and that some types of political measures may
compensate market failures.
Joan Robinson, "The Second Crises of Economic Theory,"
American Economic Review
, 62, No
For a discussion see Truman F. Bewley, “Why Not Cut Pay?”
European Economic Review
Milton Friedman, “The Role of Monetary Policy,”
American Economic Review,
Edmund S. Phelps,
Inflation Policy and Unemployment Theory
(New York: Norton, 1972).
In the judgement of Robert M. Solow, from perspective not limited to neo-classical economics,
the original version of search theory provides a picture that puts demands on credulity. "People
who give the vague impression of being unemployed are actually engaged in voluntary leisure.
They are taking it now, planning to substitute extra work later, because they think, rightly or
wrongly, that current real wages are unusually low compared with the present value of what the
labor market will offer in the future. … It is astonishing that believers have made essentially no
effort to verify this central hypothesis." Robert M. Solow,. “On Theories of Unemployment,”
American Economic Review,
70, no. 1 (1980, 1-11): 7.
Michael Bruno and Jeffrey D. Sachs,
The Economics of Worldwide Stagflation
Blackwell, 1985). For a discussion see Walter Korpi, “Political and Economic Explanations for
Unemployment: A Cross-National and Long-Term Analysis.”
British Journal of Political Science
21, no. 3 (1981): 315-48.
Also several other factors have figured in the discussion, for example, competition from newly
industrialized countries and regional mismatch. Although such factors can be of some relevance in
some circumstances, they do not appear to be crucial for the patterns of longterm changes
Education at a Glance
(Paris: Organisation for Economic Co-operation and
See Rune Åberg, “Equilibrium Unemployment, Search Behaviour and Unemployment
Cambridge Journal of Economics
25, no. 2 (2001):131-47; Lex Borghans and
Andries de Grip, ed.
The Overeducated Worker?: The Economics of Skill Utilization
Edward Elgar, 2000); Richard B. Freeman,
The Over-Educated American
(New York: Academic
Press, 1976); Sten Johansson, Per Lundborg, and Johnny Zetterberg. 1999.
karaktär och vägarna till full sysselsättning
(Stockholm: Fackföreningsrörelsens Institut för
Ekonomisk Forskning, 1999); Carl le Grand, Ryszard Szulkin and Michael Tåhlin, “Har jobben
blivit bättre? En analys av arbetsinnehållet under tre decennier.” In J. Fritzell, M. Gähler and O.
Välfärd och arbete i arbetslöshetens årtionde. SOU 2001:53
, Stockholm: Fritzes,
Tomas Korpi and Sten-Åke Stenberg, “Massarbetslöshetens Sverige - arbetslöshetens karaktär
och effekter på individers levnadsförhållanden,” in J. Fritzell, M. Gähler and O. Lundberg.
Välfärd och arbete i arbetslöshetens årtionde. SOU 2001:53
, Stockholm: Fritzes, 2001), Chap. 2.
(New York: Macmillan, 1975).
For a review see, for example, Charles R. Bean,. “European Unemployment: A Survey.”
Journal of Economic Literature
Already during the period between the two world wars, pre-Keynesian economists such as A.C.
Pigou, John Hicks, Eli Heckscher and Gustav Cassel emphasized the negative incentive effects of
unemployment insurance. See Bertil Holmlund, “Unemployment Insurance in Theory and
Scandinavian Journal of Economics,
100, no. 2 (1998):113-41.
See Richard Layard, Stephen Nickell, and Richard Jackman,
Performance and the Labour Market
(Oxford: Oxford University Press, 1991); Stephen Nickell
and Richard Layard, “Labor Market Institutions and Economic Performance,” in Orley
Ashenfelter and David Card, ed.
Handbook of Labor Economics, Volume 3
,. Amsterdam: North-
Holland, 1999); Stephen Nickell, “Unemployment and Labor Market Rigidities: Europe versus
Journal of Economic Perspectives,
11, no. 3 (1997):55-74; Horst Siebert,
“Labour Market Rigidities: At the Root of Unemployment in Europe,”
Journal of Economic
11, no.3 (1997):37-54.
Layard, Nickell and Jackman,
For example, in Sweden in 1990 the unemployment insurance fund for fishermen paid out
benefits to 73 percent of its members at least sometime during the year. This reflects not only bad
weather conditions but also largely the effects of fishing quotas imposed by international
agreements. In the funds for musicians and for theater employees, the percentage with benefits at
least sometime during the year were 37 and 29, respectively, something reflecting the very
unstable career prospects in such branches of activity. However, as long as the level of
unemployment remained low, in the great majority of all funds unemployment insurance was
largely used for brief spells of unemployment and there was little "churning" between
unemployment benefits and active labor market programs. See Walter Korpi,
arbetslöshetsförsäkring i Sverige
(Stockholm: Allmänna förlaget, 1995).
See Bernhard Ebbinghaus, "A Way Out of 'Exit from Work'? Reversing the Entrenched
Pathways of Early Retirement." in Scharpf and Schmidt,
Welfare and Work in the Open Economy:
Diverse Responses to Common Challenges,
See Jørgen Goul Andersen, “Marginalization, Citizenship and the Economy: The Capacities of
the Universalist Welfare State in Denmark,” in Erik Oddvar Eriksen and Jörn Loftager, ed.,
Rationality of the Welfare State
(Oslo: Scandinavian University Press, 1996); Jelle Visser and
A Dutch Miracle: Job Growth, Welfare Reform and Corporatism in the
(Amsterdam: Amsterdam University Press, 1997).
See Marie Jahoda,
Employment and Unemployment: A Social Psychological Analysis
York: Cambridge University Press, 1982); Duncan Gallie, Catherine Marsh, and Carolyn Vogler,
Social change and the experience of unemployment
(Oxford: Oxford University Press, 1994);
Duncan Gallie and Serge Paugam, eds.
Welfare Regimes and the Experience of Unemployment in
(Oxford: Oxford University Press, 2000).
See Susanne Alm,
The Resurgence of Mass Unemployment. Studies on Social Consequences of
Joblessness in Sweden in the 1990s
(Stockholm: Swedish Institute for Social Research, Stockholm
University, 2001); Anders Björklund, “Unemployment and Mental Health - Some Evidence from
Journal of Human Resources
, XX, no. 4 (1985); Anders Björklund and Tor Eriksson,
“Unemployment and Mental Health: A Survey of Nordic Research.”
Scandinavian Journal of
7, no 3(1998):219-35; Gallie, Marsh and Vogler,
Social Change and the
Experience of Unemployment
; Gallie and Paugam,
Welfare Regimes and the Experience of
Unemployment in Europe;
Employment and Unemployment
; Tomas Korpi,
Unemployment: Studies in the Individual Consequences of Unemployment and Labor Market
(Stockolm: Swedish Institute for Social Resarch 1994); Tomas Korpi,
Process. Studies of Search, Selection, and Social Mobility in the Labor Market.
Department of Sociology, Stockholm University); Mikael Nordenmark, “Employment
Commitment and Psychological Well-being among Unemployed Men and Women,”
42, no 2 (1999): 125-46; Mikael Nordenmark and Mattias Strand, “Towards a
Sociological Understanding of Mental Well-being Among the Unemployed: the Role of Economic
and Psycho-Social Factors.”
33, no. 4 (1999):577-97.
Averages for countries are population weighted. Respondents here include only persons 20-54
years of age classified either as unemployed or as employees in employment. The self-employed,
etc are thus excluded. Since manual workers show lower levels of satisfaction than do white-collar
employees and former manual workers probably are over-represented among the unemployed, the
average level of dissatisfaction among employees is here weighted so that manual workers have
twice the weight of white-collar employees. For the years 1980 and 1981, probably for some
technical reasons, the proportion dissatisfied among unemployed men is considerably lower than
for other years (source Ronald Inglehart, Karlheinz Reif, and Anna Melich,
Communities Studies, 1970-1992
(Michigan: Inter-university Consortium for Political and Social
Cumulative File ICPSR 9361 / ZA 2533,
and later versions).
Rune Åberg, “Ekonomiska och sociala perspektiv på sökeffektivitet och
Arbetmarknad och Arbetsliv,
1, no.2 (1995).
The positive role of social insurance institutions has been discussed by economists such as Jonas
Agell, “On the Benefits from Rigid Labour Markets: Norms, Market Failures and Social
109, no.453(1999):143-64, and Anthony B. Atkinson,
Economic Consequences of Rolling Back the Welfare State
(Cambridge, Mass.: MIT Press, 1999).
See Åberg, “Equilibrium Unemployment, Search Behaviour and Unemployment Persistency,”
The Resurgence of Mass Unemployment
Anthony B. Atkinson and John Micklewright, “Unemployment Compensation and Labor
Market Transitions: A Critical Review.”
Journal of Economic Literature
29, no. 4 (1991):1679-
727; Theresa J. Devine and Nicholas M. Kiefer.
Empirical Labor Economics. The Search
(New York: Oxford University Press, 1991); Holmlund, Unemployment Insurance in
Theory and Practice; Layard, Nickell and Jackson,
; Peder J. Pedersen, and Niels
Westergard-Nielsen, “Unemployment: A Review of the Evidence from Panel Data.”
No. 20, Spring 1993): 65-114.
Atkinson and Micklewright, Unemployment Compensation and Labor Market Transitions.
Devine and Kiefer
(Empirical Labor Economics,
302) review studies primarily from the United
States and the United Kingdom and find that among jobs applied for by the unemployed, "the
typical acceptance probability [of a job offer] is high – in the range of 0.8-1.0." This need however
of course not mean that the unemployed wold accept
Åberg, “Equilibrium Unemployment, Search Behaviour and Unemployment Persistency,”
Johansson, Lundborg and Zetterberg,
…; 1999; Solow, On
Theories of Unemployment.
Estimates of NAIRU and NAWRU (the non-accelerating wage rate of unemployment) are
typically derived from regression equations based on time series with previous unemployment and
changes in consumer prices or wages as the most typical independent variables. Since these
estimates have not been shown to be associated with the behavior of the unemployed, they need
not have construct validity.
Layard, Nickell and Jackman,
, Talk about the unemployed as "inflation fighters"
to refer to the functions of the unemployed in decreasing wage pressures.
Åberg, “Equilibrium Unemployment, Search Behaviour and Unemployment Persistency,”
Anders Harkman and Fredrik Jansson, “Vad bestämmer sökaktiviteten?”
1, no. 2 (1995).
Howard Machin and Alan Manning, “The Causes and Consequences of Longterm
Unemployment in Europe” in Orley Ashenfelter and David Card, ed.,
Handbook of Labor
Economics, Volume 3
(Amsterdam: North-Holland, 1999).
The OECD Jobs Study. Unemployment in the OECD Area 1950-95
Alan Manning, “Comment on B. Holmlund, 'Unemployment Insurance in Theory and Practice'
Scandinavian Journal of Economics,
100, no.1 (1998): 143-45. In a warning that should be
taken seriously in all social science disciplines, Manning writes: "We have very strong
beliefs from theory that there should be such a link [between unemployment benefits and
unemployment], beliefs that are so strong that we can read paragraphs like the one above and not
think that anything is amiss.
We need to be honest about the fact that theory plays a disturbingly
large part in informing discussion about the impact of unemployment insurance on the labour
L. Ljungqvist and T. Sargent. “The European Unemployment Dilemma,”
Bank of Chicago
, Working Paper 17, 1995).
Ola Sjöberg, “Unemployment and Unemployment Benefits in the OECD 1960-1990.”
Employment and Society,
14, no. 1(2000):51-76.
The net replacement level is computed as an average for a single person and a four-person
family with one earner and two minor children during the first week of unemployment (after
waiting days) and a year consisting of 26 weeks of benefits and 26 weeks of work. In countries
without a defined duration, it is set to 260 weeks, that is five years. Data on unemployment
insurance are from the
Social Citizenship Indicator Program (SCIP)
now under construction at the
Swedish Institute for Social Research, Stockholm University, and have considerably higher quality
than those assembled by the OECD. Benefits levels and durations in unemployment insurance
have been determined as of the end of every fifth year in the period 1960-1995. Since it is
reasonable to assume some time lags in the effects of changes in unemployment insurance on the
behavior of the unemployed, and since changes in programs often are relatively marginal, for each
observation year the characteristics of unemployment insurance is correlated with unemployment
rates for the observation year plus the two following years. Thus, for example, the replacement
rate at the end of 1960 is correlated with unemployment levels 1960, 1961 and 1962. The number
of observation years forming the bases for the computation of average period correlations is thus
9, 9 and 6.
These results are thus in agreement with the conclusion by Steinar
Strøm, “Comment on B.
Unemployment Insurance in Theory and Practice
Scandinavian Journal of
no. 1 (1998:151-152) that "neither microeconomic nor macroeconomic results
give strong and/or convincing support to the predictions of strong positive relationship between
unemployment and unemployment benefits as indicated in microeconomic and macroeconomic
theory. Moreover, there is no support for the strong beliefs, held by many economists and others,
of a strong positive relationship between unemployment benefits and duration of unemployment."
A.W. Phillips, “The Relation between Unemployment and the Rate of Change of Money Wage
Rates in the United Kingdom, 1861-1957.”
25, November (1957):283-99.
David G. Blanchflower and Andrew J. Oswald,
The Wage Curve
, Cambridge, Mass.: The MIT
Press; David G. Blanchflower and Andrew J. Oswald, “An Introduction to the Wage Curve,”
Journal of Economic Perspectives,
9, no. 3 (1995:153-67).
George A. Akerlof “Gift Exchange and Efficiency-Wage Theory: Four Views.”
74, no. 2 (1984):79-83; George A. Akerlof and Janet L. Yellen, ed.,
Wage Models of the Labor Market
(Cambridge: Cambridge University Press, 1986); Janet L.
Yellen, “Efficiency Wage Models of Unemployment,”
American Economic Review
74, no 2
Power and Exchange in Social Life
(New York: Wiley, 1964); George C. Homans,
Social Behavior: Its Elementary Forms
(New York: Harcourt, Brace & World, 1961).
Carl Shapiro and Joseph E. Stiglitz, “Equilibrium Unemployment as a Worker Discipline
American Economic Review,
74, no. 3(1984):433-44.
The Last Phase in the Transformation of Capitalism
(New York and London:
Monthly Review Press, 1943): 175.
Layard, Nickell and Jackman,
Newell and Symon suggest that during the twentieth century the long-term changes in
unemployment in the United Kingdom reflect changing tastes for unemployment in the labor
force, with the generations experiencing the period between the world wars being highly risk-
averse but those grown up during full employment less so (Andrew Newell and James Symon,
“The Passing of the Golden Age,” in Renato Brunetta and Carlo Dell'Aringa, ed.,
Relations and Economic Performance. Proceedings of a conference held by the International
Economic Association in Venice, Italy
(London: The Macmillan Press Ltd, 1990): 353-77.
The problem with this explanation is that we have no independent measure of changes in tastes
and that it is difficult to know why and when some tastes but not others are transformed into
policies. Furthermore, as noted above, also in recent decades those who have tasted
unemployment find it bitter.
See Eero Carroll,
Emergence and Structuring of Social Insurance Institutions. Comparative
Studies on Social Policy and Unemployment Institutions.
Stockholm: Swedish Institute for Social
Research, Stockholm University, 1999); Ingalill Montanari, "Modernisation, Globalisation and the
Welfare State: A Comparative Analysis of Old and New Convergence of Social Insurance since
British Journal of Sociology
, 52, no.3 (2001): 469-494.
From 1960 to 1973, in the smaller OECD countries average tax rates increased from 27 to 36
percent, in the larger ones from 29 to 33 percent (OECD:
, Table 6.6).
During The Great Trough 1960-1973, the correlation between levels of unemployment and rates
of GDP per capita growth among our 18 countries is 0.04. While this type of correlation of course
must be interpreted with utmost care, it would appear to indicate that it is difficult to equate rapid
growth with full employment. It would also appear difficult to explain variations in unemployment
in terms of wage setting institutions.
Modern Capitalism: The Changing Balance of Public and Private Power
(New York: Oxford University Press, 1965), 63.
Andrea Boltho, 1989. “Did Policy Activism Work?”
European Economic Review
Political and Economic Explanations for Unemployment.
Measured as the average percentage of votes for the traditional social democratic parties and
parties to their left in our 18 countries, left parties had received only 14 percent of votes 1900-13,
and 28 percent in 1919-38 but garnered on the average 37 percent of the votes in 1946-72. Before
the First World War, left government participation had been exceedingly rare. In terms of the
average left share of cabinets seats, their participation doubled from 13 percent between the two
world wars to 28 percent 1946-72, while the percentage of time with left representation in
governments doubled from 22 to 46 percent. Average union membership as a percentage of
employees climbed from 22 percent in the inter-war period to 41 percent in the latter period.
Before the Second World War, in most of our countries universal suffrage had typically only
existed during 2-3 decades. In four of them – Austria, Germany, Italy and Japan – authoritarian
regimes had been in power in the inter-war years.
The countries covered were Belgium, France, Germany, Italy, the Netherlands, Norway, the
United Kingdom and the United States.
E. S. Kirchen, J. Benard, H. Besters, F. Blackaby, O. Eckstein, J. Faaland, F. Hartog, L.
Morissens, and E. Tosco.
Economic Policy in Our Time, Vol I
. (Amsterdam: North-Holland,
Duverger (1958) Duverger, Maurice. 1958. Political Parties: Their Organization and Activity in
the Modern State. New York: Wiley.
One example is provided by the Conservative government in Britain 1951-1964, where the
Treasury has been described as behaving like "a simple Pavlovian dog responding to two stimuli,"
one being "a run on the reserves" and the other "500 000 unemployed" with politicians responding
most often to the latter stimuli and officials to the former one. Samuel Brittan,
The Treasury under
the Tories, 1951-1964
(Harmondsworth: Penguin Books, 1964), 288.
Because of the severe split among social democrats and communists in Finland, France and
Italy, in these countries only one third of the time in government is taken into account.
In an interesting volume on the acceptance of Keynesian economic ideas in the Western
countries before and after the Second World War, Peter Hall discusses the role of professional
economists, institutional structures, and political parties in this context (Peter A. Hall,
Political Power of Economic Ideas: Keynesianism across Nations
, (Princeton NJ: Princeton
University Press, 1989). There would however not appear to be a clear correlation between the
ease of acceptance of Keynesian thinking among economists and levels of unemployment during
The Great Trough. Thus, for example, while economists' acceptance was late in Finland this
country by and large had very low unemployment, while easy and early acceptance in the United
States came to co-exist with relatively high unemployment.
Koji Taira, “Japan's Low Unemployment: Economic Miracle or Statistical Artifact?”
106 (1983):3-10; Angelika Ernst, “Arbeitslosigkeit und Unterbeschäftigung in
Japan. Eine Übersicht zum Stand der Empirischen Forschung.”
Mitteilungen aus der Arbeitsmarkt
11, (1978):34-49; Constance Sorrentino, “Japan's Low Unemployment: An
Monthly Labor Review,
See Fred Hirsch, “The Ideological Underlay of Inflation,” and John H. Goldthorpe, “The
Current Inflation: Towards a Sociological Account,” in Fred Hirsch and John H. Goldthorpe, ed.,
The Political Economy of Inflation
(Cambridge, Mass: Harvard University Press, 1978); Fred
Hirsch and Peter Oppenheimer, “The Trial of Managed Money: Currency, Credit and Prices 1920-
1970,” in Carlo M. Cipolla, ed.,
The Fontana Economic History of Europe
Collins/Fontana, 1976), and Andrew Glyn, “Social-Democracy and Full-Employment,”
no. 211 (1995):33-55.
An example here is that the concept of the Phillips curve has come to mean not a relationship
between unemployment and the change of money wages as it was originally derived by Phillips
but instead a relationship between unemployment and rates of inflation.
For reviews see Michael D. Bordo and Anna J. Schwartz, “Monetary Policy Regimes and
Economic Performance: The Historical Record,” in J.B. Woodford and M. Taylor, ed.,
(Amsterdam: Elsevier, 1999):149-233; Barry Eichengreen and Marc
The Gold Standard in Theory and History
(London: Routledge, 1997). The gold
standard was the most common one but some countries had a tie to silver, a few also a bi-metallic
Hirsch and Oppenheimer,
The Trial of Managed Money
Up to 1959 employees involved in strikes and lockouts are taken as a percent of the non-
agricultural population; after 1960 as a percentage of all employees. Because of changes in
definitions in published data on industrial disputes from the International Labour Organization, the
United States is excluded after 1981 and France since 1983.
Arthur M. Ross and Paul T. Hartman,
Changing Patterns of Industrial Conflict
John Wiley and Sons, 1960).
Walter Korpi and Michael Shalev, “Strikes, Power and Politics in the Western Nations, 1900-
Political Power and Social Theory,
1(1980):301-34. In Sweden the number of unofficial
strikes increased after 1969. The major break in the pattern of relative industrial peace, however,
was the 1980 nation-wide lockout in the private sector declared by the Swedish Employers'
Confederation during the period of Center-Conservative governments 1976-81.
See Colin Crouch and Alessandro Pizzorno,
The Resurgence of Class Conflict in Western
Europe Since 1968
1978) for a contemporary discussion. Behind these
unweighted averages we do however find major inter-country differences. Thus, besides Austria,
Norway and Sweden conflict rates were also low in Germany, the Netherlands and Switzerland
while they were high in Australia, Finland, France, Italy, New Zealand and the United Kingdom.
In the United States and Canada, disputes of long duration did however remained the dominant
pattern. The same was also partially true in Ireland.
The literature on "economic voting" indicates that to a considerable extent these fears were
realistic (cf below).
In retrospect it would appear that high inflation in this period was to a large extent related to the
two oil shocks and the Vietnam war, and thus need not have invalidated the Keynesian policy
Inflation: The Present Problem
(Paris: OECD, 1970), 6-10.
Ibid., 8-9, 35, italics added.
Gösta Rehn and Rudolf Meidner were both economists at the Swedish Confederation of Trade
Unions (LO) in the 1940s and 50s, and developed the intellectual basis for the active labor market
policy introduced during this period.
Gösta Rehn 1990, oral information to the author.
, Towards Full Employment and Price Stability (OECD: Paaris, 1977):26.
For discussions see Ton Notermans, "The Abdication from National Policy Autonomy: Why the
Macroeconomic Policy Regime Has Become So Unfavorable to Labor."
Politics & Society
, 21, no
2 (1993): 133-167; Jonathon W. Moses, "Abdication from National Policy Autonomy: What's Left
Politics & Society
, 22, no. 2 (1994): 125-148; and Geoffrey Garrett, "The Causes of
Comparative Political Studies
, 33, no.6-7 (2000):941-991.
See Fritz W. Scharpf,
Crisis and Choice in European Social Democracy
(Ithaca NY: Cornell
University Press, 1991); Philip Manow and Eric Seils. 2000. “Adjusting Badly: The German
Welfare State, Structural Change, and the Open Economy,” in Fritz W. Scharpf and Vivien A.
Welfare and Work in the Open Economy: Diverse Responses to Common Challenges
(Oxford: Oxford University Press, 2000): 264-307; Wendy Carlin and David Soskice. “Shocks to
the System: The German Political Economy under Stress,”
National Institute Economic Review,
no. 159 (1997):57-76.
This policy included a doubling of the nominal interest rate.
Trade Unions. The Logic of Collective Action
(London: Fontana, 1982); Martin
Rhodes, 2000. “Restructuring the British Welfare State: Between Domestic Constraint and Global
Imperatives,” in Scharpf and Schmidt,
Welfare and Work in the Open Economy: Diverse
Responses to Common Challenges,
Governing the Economy. The Politics of State Intervention in Britain and France
(Cambridge: Polity Press, 1986); Jonah D. Levy, "Vice into Virtue? Progressive Politics and
Welfare Reform in Continental Europe," Politics & Society, 27, No. 2 (1999): 239-273.
Layard, Nickell and Jackman,
Helmut Northpoth, "Economic, Politics, and the Cycle of Presidential Popularity," in
Michael S. Lewis-Beck ed.,
Economics and Elections. The Major Western Democracies
Arbor: University of Michigan Press, 1988): Chap. 9.
Economics and Elections.,
G. Bingham Powell and Guy D. Whitten, A Cross-
National Analyses of Economic Voting.”
American Journal of Political Science
37, no. 2
For leads or access to the data discussed below, I want to thank John Curtice, Jørgen Goul
Andersen, Anthony Heath, Sören Holmberg, and Bernhard Wessels.
See Mori Polls & Surveys (http://www.mori.com/polls/trends, October 10, 2001). In the latter
period unemployment rates were about twice as high as rates of inflation.
2001, Table 4). During this period "other problems with the economy" were
however mentioned more often than inflation, which only was pointed out by 0-6 percent of
Stockholm:Liber, (2000):114. Inflation has here been coded
among other economic issues.
Korpi and Stenberg,
In this context is should be remembered that the high rates of inflation observed in the 1970s to
a large extent reflected drastic increases in commodity prices and the Vietnam war.
(New York: McGraw-Hill, 1996):429, italics
James Tobin, “Inflation and Unemployment,”
American Economic Review,
62, no.1 (1972):1-
Layard, Nickell and Jackman,
, 13, italics added.
Stephen Nickell, Unemployment: Questions and Some Answers,"
, 108, no.
Gösta Rehn, “State, Economic Policy and Industrial Relations in the 1980s: Problems and
Economic and Industrial democracy,
8, no. 1 (1987:16-93), 63, italics added.
The Last Phase in the Transformation of Capitalism.
For a discussion see Göran Therborn,
Why Some Peoples are More Unemployed Than Others
Belgium, Denmark, France, Germany, Ireland, Italy, the Netherlands and the United Kingdom.
Austria, Finland, Norway and Switzerland and Sweden.
For helpful reviews M.Crepaz and A Lijphart, “Corporatism and Consensus Democracy in
Eighteen Countries: Conceptual and Empirical Linkages.”
British Journal of Political Science
(2)( 1991): 135-46; Michael Shalev, “Class Conflict, Corporatism and Comparison: The Japanese
Enigma,” in S. N. Eisenstadt and E. Ben-Ari, ed.,
Japanese Models of Conflict Resolution
(London: Kegan Paul International, 1990): 60-93; Alan Siaroff, “Corporatism in 24 Industrial
Democracies: Meaning and Measurement.”
European Journal of Political Research,
36, no. 2
The Finnish Political System
(Cambridge, Mass: Harvard University Press,
1971). A minority of one-sixth of the
could thus postpone some types of legislation.
Up to 1987, reflecting of the relationship to the Soviet Union, the Conservative party in Finland
was considered unfit for government participation.
Giuliano Bonoli and André Mach. 2000. “Switzerland: Adjustment Politics within Institutional
Constraint,” in Scharpf and Schmidts,
Welfare and Work in the Open Economy,
The Political Construction of Interests: National Health Insurance Politics in
Switzerland, France and Sweden, 1930-1970
(New York: Cambridge University Press, 1992); H.
Entscheidungsstrukturen und Entscheidungsprozesse in der Schweizer Politik
Campus: 1980). Franz Lehner, “Consociational Democracy in Switzerland: A Political-Economic
Explanation and Some Empirical Evidence.”
European Journal of Political Research,
1(1984):25-42; Manfred G. Schmidt,
Der Schweizerische Weg zur Vollbeschäftigung: Eine Bilanz
der Beschäftigung, der Arbeitslosigkeit und der Arbeitsmarktpolitik
(Frankfurt: Campus, 1985).
Lars Calmfors and John Driffill, “Bargaining Structure, Corporatism and Macroeconomic
, 6, (1988):14-61; Layard, Nickell and Jackman, Unemployment;
In this context it must however be remembered that in Ireland, a type of macro-political
bargaining with labor in a relatively weak role was instituted in the 1980s. See Niamh Hardiman,
“Economic Governance and Political Innovation in Ireland.” Paper presented at Annual Meeting
of the American Political Science Association, Washington DC, August 31-September 3, 2000.
In New Zealand as well as in Australia wage arbitration by courts has had a long tradition but was
largely discontinued in the 1990s; in Australia however after a period of macro-level bargaining
during the Accord of the 1980s.
In Austria, social democrats dominated the government from 1970 to 1987 (for a discussion see
Anton Hemerijck, Brigitte Unger, and Jelle Visser, “How Small Countries Negotiate Change:
Twenty-Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium,” in Scharpf
Welfare and Work in the Open Economy:Diverse Responses to Common Challenges,
Thus, for example, in Sweden the center-conservative coalition government 1976-81
practically nationalized the important steel industry and the major shipyards to avoid a collapse of
Similar tendencies, although much less marked, were also found in Austria and Germany. From
1973 to 1977 the immigrant proportion of the Swiss labor force declined by 5.3 percentage points
Der Schweizerische Weg zur Vollbeschäftigung
, 30. The equivalent decrease in Austria
and Germany was 1.5 and 2 percentage points (Scharpf,
Crises and Choice
, 72). The Swiss case is
difficult to interpret in terms of patterns of wage bargaining and union strength. While Switzerland
has a high degree of employer coordination, unions are generally weak and wage bargaining
largely takes place on the firm level. The country never practiced Keynesian policy models. Here
the constitution with easy provisions for referenda and the mosaic of linguistic-religious cleavages
forming the background for this constitution appear as basic for the co-operative nature of Swiss
The results do not change much if we use an additive index of co-ordination among employers
and unions, respectively. According to a classification in these terms by Nickell,
Qustions and Some Answers
, 811, this combined index shows low values for Australia, Canada,
Ireland, New Zealand, UK and USA, medium values for Belgium, France, Netherlands, Italy,
Japan and Switzerland and high values for Austria, Denmark, Finland, Germany, Norway and
Sweden. This observation need however not deny that union capacity for centralized bargaining is
important (for consequences of its absence in the United Kingdom, see
Economy. The Politics of State Intervention in Britain and France
Cambridge: Polity Press
and Rhodes, Restructuring the British Welfare State. Furthermore, as argued by Soskice
(Reinterpreting Corporatism and Explaining Unemployemt), employer co-ordination is likely to be
of importance for long-terms business climates.
Ronald Dore, “The Distinctiveness of Japan,” in Colin Crouch and Wolfgang Streeck, ed.,
Political Economy of Modern Capitalism
, (London: SAGE, 1997): 19-32.
For a discussion see Herman Schwarz, “Internationalization and Two Liberal Welfare States:
Australia and New Zealand,” in Scharpf and Schmidt,
Welfare and Work in the Open Economy:
Diverse Responses to Common Challenges
For an early analysis of interactions between institutions and left government power, see Peter
Lange and Geoffrey Garret, "The Politics of Growth: Strategic Interaction and Economic
Performance in the Advanced Industrial Democracies, 1974-1980,"
Journal of Politics
, 47, no. 3
(1991):792-827. See also Geoffrey Garrett and Peter Lange
"Political Responses to
Interdependence: What's "Left" for the Left?"
It has however here not been possible to include other important policy measures, such as fiscal
Data on nominal interest rates come from the OECD. The GDP inflation rate of the current
year is subtracted from the nominal rate to get the real short-term official interest rate. The EEC
category here includes Belgium, Denmark, France, Germany, the Netherlands, Ireland, Italy and
the United Kingdom. The EFTA countries here are Austria, Finland, Sweden, and Switzerland.
Norway is not included because of lack of data. Furthermore since Norway is a small country with
a very large oil industry, the Norwegian economy and economic policy has to a great extent been
affected by the oil sector, something that makes it less comparable with the other EFTA countries.
Ireland and Italy with high unemployment did however follow the European pattern to the
extent that they had negative real interest rates (-5.7 and -1.0, respectively).
In Australia, Italy, the Netherlands and the United Kingdom, however, the largest increases
took place from 1979-80 to 1981-82. For these countries, increases in unemployment are
determined as changes in average unemployment from 1975-83 to 1984-85.
In France the intense electoral competition between the left and the right in the 1970s was
likely to inhibit strong disinflationary policies (cf above).
Jukka Pekkarainen and Juhana Vartiainen,
Finlands ekonomiska politik. Den långa linjen 1918-
(Stockholm: FIEF, 2000).
The increase of EFTA interest rates in Figure 7 during this period reflects primarily
developments in Finland and Sweden. In Austria and Switzerland only limited increases were
For discussions see Lars Calmfors, “Lessons from the Macroeconomic Experience of Sweden.”
European Journal of Political Economy,
9, no.1 (1993):25-72; Lennart Erixon, “A Swedish
Economic Policy - The Rehn-Meidner Model's Theory, Application and Validity, in Henry
Milner and Eskil Wadensjö, ed,.
Gösta Rehn, the Swedish Model and Labour Market Policies.
International and National Perspectives
, (Aldershot: Ashgate, 2001), and Andrew Martin, "The
Politics of Macroeconomic Policy and Wage Negotiations in Sweden." in Torben Iversen, Jonas
Pontusson, and David W. Soskice , ed.,
Unions, Employers and Central Banks: Macroeconomic
Coordination and Institutional Change in Social Market Economies.
University Press, 2000): Chap. 8.
Laurence Ball, “Aggregate Demand and Long-Run Unemployment,”
Brookings Papers on
no. 2 (1999): 189-248.
Carlin and Soskice, Shocks to the System, 70.
Compensation to employees paid by resident producers includes wages and salaries but also
social security contributions, private pension schemes, etc. Operational surplus is defined as the
difference between the value of gross output for producers less producers' costs including
compensation to employees, costs of intermediate goods bought for use in the production process,
the value of wear and tear of production capital and indirect taxes. Operational surplus includes
the incomes of the self-employed. Beginning in 1996, the OECD has changed the definition of the
operational surplus category to include also "mixed income," information for which are given only
from 1970 onwards, and the earlier series was discontinued after 1996. This addition substantially
increases the profit/wage ratio, but for the years during which both series are available the
correlation between the two versions of this ratio is very close to unity. We have therefore used
regressions to estimate the original ratio for the years 1997 and 1998.
The profit/wage ratio is a ratio between the profit/GDP ratio and the wage/GDP ratio. A look at
these two separate ratios indicates that in most countries the former shows tendencies to a U-shape
while the latter tends to reverse this shape.
From 1960 to 1974, the decline of the share of the agricultural sector of civilian employment
has declined by at least 15 percentage points in Finland, Japan, Italy and Ireland, 10-15 percentage
points in Austria, France, and Norway, 5-9 percentage points in Canada, Denmark, Germany,
Sweden and Switzerland, but by less than 5 percentage points in Australia, the Netherlands, New
Zealand, the United Kingdom and the United States. From 1974 to 1997, declines with a least 5
percentage points are found only in Japan, France, Italy, Austria, Denmark, Finland, Ireland and
A similar pattern is also found for Australia and New Zealand.
In several of our countries, data on operational surplus in manufacturing are found only for
shorter time periods and are entirely missing in Austria, Ireland and Switzerland. In Germany,
however, this ratio shows a decline after 1990.
Robert M. Solow,
The Labor Market as a Social Institution,
(Oxford: Basil Blackwell, 1990):
The Last Phase
in the Transformation of Capitalism
Rehn, State, Economic Policy and Industrial Relations in the 1980s.
Rehn, ibid., 65.
This is likely to lead to choices in favor of low inflation. See Robert J. Franzese and Peter A.
Hall, "Institutional Dimensions in Coordinating Wage Bargaining and Monetary Policy," in
Torben Iversen, Jonas Pontusson, and David W. Soskice, ed.,
Unions, Employers and Central
Banks: Macroeconomic Coordination and Institutional Change in Social Market Economies
(Cambridge: Cambridge University Press, 2000): Chap. 6.