Resumé

In the 1970s the level of the distributive conflict increased in most developed capitalist countries. But at the same time it was found that the level of the distributive conflict remained on an unchanged level or even declined in some countries. In this article the role of interest organizations is analysed in explaining both cross-national differences in the level of the distributive conflict and the resurgence of the distributive conflict in the 19705. Using Mancur Olson's new theory of interest organizations, the author shows on a theoretical level that a pluralist mode of interest intermediation is conducive to a high level of the distributive conflict and that a corporatist mode of interest intermediation is conducive to a low level of the distributive conflict. Empirical analyses in 18 developed capitalist countries are then carried on to test this theory. Empirical analyses strongly support the theory.

Introduction

Distributive struggle is an essential part of economic as well as of political life. Economic life deals with allocation. It is about production, distribution and consumption. Political life deals with authorization. It is about values, interests, power and distribution. Distributive problems are an essential part of economic as well as of political life. Distribution is something that connects economic and political life. Values to be distributed are produced in economic life. The actual process of distribution is always influenced by political authorization.

Conflict is immanent in distribution. There are conflicts between workers
and employers, between subgroups of workers, between subgroups of
employers, between different regions, and between different nations.

These conflicts may be manifest and severe, of latent and harmonious.
The critical issues sometimes rise in the economy, sometimes in the polity.

After the Second World War, especially in the 19605, there were many marks of decrease in the distributive conflict in the developed capitalist countries. Economic growth was rapid and steady. There seemed to be new affluence to be distributed to everyone. Governments adapted Keynesian doctrines of economic regulation. According to Keynesian stabilization doctrines and post-Keynesian growth doctrines, governments were powerful

both in activating economic growth and in redistributing produced economic values. In the social science of the 1960s the following arguments were common: The era of class society and class politics is over (Lipset 1963). With active economic policies nations have got rid of economic crises and serious business cycles (Keynes 1936). The redistributive capacity of the new welfare policies is so big that it is meaningless to talk about exploitation in the capitalist society. Economic decision making has been democratized by means of active government policies that are decided in the interplay between political parties and different interest groups (Goldthorpe). As a result of the economic intervention of the government the laws of economic development have changed. Therefore, traditional capitalism has ceased to exist. Now it is reasonable to talk about mixed economy (Samuelson 1976), post-capitalism and post-industrial society (Bell 1974). Because of the redistributive capacity and the welfare programmes of the government it is also reasonable to talk about a welfare state (Wilson & Wilson 1982; Flora & Heidenheimer 1981).

In this process, distributive struggle that originally was situated mainly in the economy was partly translated into the polity (Hibbs 1978) and then deradicalized and demilitarized by the redistributive and growth policies of the welfare state (Holloway & Picciotto).

In the 1970s something new happened. Economic growth rates declined and became unstable. Large increases in unemployment worsened the standard of living of these unemployed groups (OECD 1982a). Because of high rates of inflation the distribution of income became very unstable (Hirsch & Goldthorpe 1978; Medley 1982). Rising energy, raw material and wage costs threatened the profits of business corporations (McCracken 1977; Dahrendorf 1982; Cox 1982). On the other hand, price inflation diminished the real value of wage and salary earnings. In this situation distributive problems became manifest (Walsh 1983). There was a resurgence of class conflict (Crough & Pizzorno 1978).

In this paper, distributive conflict and its relationship to the power structures of interest organizations and to different modes of interest intermediation will be analysed in developed capitalist countries with democratic political institutions. Countries to be studied are Australia, Austria, Belgium, Canada, Denmark, Finland, France, the Federal Republic of Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Sweden, Switzerland, the United Kingdom and the United States.

Research questions, on the general level, are as follows:

1. What is the role of interest organizations and modes of interest intermediation
in explaining cross-national differences in the distributive conflict?

2. What is the role of interest organizations and modes of interest intermediation
in explaining the resurgence of distributive conflict in the 19705?

Labour Dispute Activity

In this paper distributive conflict will be analysed in terms of labour dispute activity. In the ILO Yearbook of Labour Statistics we have three figures for labour disputes. We have figures for the number of disputes in each country in each year. Secondly, we have figures for the number of employees involved. When compared to the labour force these figures can be used as an illustration of the extensiveness of labour disputes. It illustrates how large a share of the labour force has been involved in labour disputes. Then, thirdly, we have figures for the days lost in labour disputes. When compared to the labour force these figures can be used as an illustration of the over-all volume or activity of the labour disputes. Both strikes and lock-outs are included in these statistics, and no differentiation is made between strikes and lock-outs.

I have chosen the third option. Table 1 gives figures for labour dispute activity in 18 OECD countries. The figures measure the average number of days lost annually in labour disputes per 1000 workers in the 1960s and in the 19705. Annual time series for the same figures are presented in the Appendix. Figures for civilian employment (OECD) have been used as an indicator of the labour force.

The level of labour dispute activity seems to be rather constant in each country. In Austria, the Federal Republic of Germany, Japan, the Netherlands, Norway, Sweden and Switzerland, labour dispute activity was at a low level both in the 1960s and in the 19705. In Australia, Canada, Finland, Ireland, Italy, the United Kingdom and the United States, on the other hand, labour dispute activity wais at a much higher level during the same period.

In Table 1 countries have been classified into three groups according to the extent of their labour dispute activity in the 19705. We may speak of countries with a low level of distributive conflict, with a medium level of distributive conflict and with a high level of distributive conflict. Putting it another way, it may be said that countries with a low level of distributive conflict are countries with a strong economic consensus and countries with a high level of distributive conflict are countries with a weak economic consensus.

Column three in Table 1 shows the percentage difference in the level of labour dispute activity between the 1960s and the 19705. We can see that in most countries labour dispute activity rose in the 19705. Indeed it rose in all the countries with a weak economic consensus. And it fell only in four countries, Austria, France, Norway and Switzerland. The big decline in the case of France is mostly due to the "fool year" 1968 that dramatically increased the average for the 19605.

In column four we have another measure for the change in labour dispute
activity. It is the linear trend of labour dispute activity between 1960 and 1979.

At this point we should ask two questions. Firstly, how can we explain
continuity and change in labour dispute activity inside each country? Secondly,

how can we explain the marked continuity in the cross-national differences in labour dispute activity — a continuity that makes it reasonable to talk about countries with a strong economic consensus, with a medium economic consensus, and with a weak economic consensus? Spearman's rank order correlation between average levels of labour dispute activity in the 1960s and the 1970s is 0.81. In the following correlation analyses, however, Pearson's product moment correlations will be used. In these correlation analyses a logarithm transformation of the labour dispute figures in Table 1 has been used.

Interest Organizations and the Distributive Conflict

What is the role of interest organizations in explaining distributive conflict? Do they promote distributive struggle, or do they prevent and control distributive struggle? And if they can both promote and prevent distributive struggle, under what conditions do they promote conflict and under what conditions do they prevent conflict?

Olson's Theory

Mancur Olson has presented an interesting theory of the relationship between interest organizations and the distributive conflict (Olson 1982). According to Olson, interest groups have an ambivalent position in society. On the one hand, interest organizations aggregate and promote interests of different groups in society. They try to maximize the welfare of the group members. On the other hand, in a society with a multitude of interest organizations, the agendas of decision making will be crowded with issues of redistribution rather than issues of economic growth. A society with a multitude of interest organizations is also more rigid compared to a less organized society. Individuals always make decisions faster than organizations they are composed of.

One important argument in Olson's theory is that group size is of crucial significance for determining the style of group actions. According to Olson, small groups are in a position where they usually need not think about the social effects of their redlstributive pressure policies. Large interest groups, on the other hand, are in a position where they have to think about both the redistributive effects and the social costs of their pressure policies. This situation is demonstrated by the following example. Suppose that the benefit of a redistributive policy for a particular group is worth b, that the group size is n, and that the redistributive policy in question contributes a social loss of c (for instance a decrease in real income). Suppose that this social loss of c is proportionally distributed in society. Suppose, finally, that the size of the whole society is N. The total effect of the redistributive policy for a group is then positive if

We can see that small groups need carry only a small proportion of the costs of their action. Therefore the costs are often smaller than the benefits. Just the opposite is true in the case of large groups. They have to carry a large proportion of the costs of their action. Therefore, Olson supposes, large and encompassing organizations are more willing than small ones to search for solutions in which the social costs of the action will be minimized. The word encompassing has a special meaning in Olson's theory. It does not refer to

the size of the group only. The goals and interests of an encompassing organization are closely connected to the larger community, are indeed a part of it. Thus, for instance, Japanese enterprise unions are encompassing in relation to a particular firm or industry. Such unions may be small in relation to society as a whole, but the welfare of the members of the unions is closely connected to the success of the firm. The situation is different in a union that controls the supply of some specialized skill which a firm or an industry needs, but controls only a small percentage of the employees involved. Such a union would have only a marginal influence on the profitability of any firm and accordingly would have little incentive to avoid inefficient practices or to help the employer or industry in any other way.

Another interesting element in Olson's theory is his dynamic view of the development of interest organizations in society. According to Olson, stable societies with unchanged boundaries tend to accumulate more collusion and interest organizations for collective action over time. Olson seems to think that there are always more interests than interest organizations. It always takes time before unorganized special interests are organized. And alreadyorganized interest organizations do not usually cease to exist if social conditions are stable and harmonious. Therefore the amount of interest organizations tends to increase, the bargaining tables and agendas will be more and more crowded, and the society's capacity to adapt new technologies and to reallocate resources in response to changing conditions will slow down.

Pluralist and Corporatist Modes of Interest Intermediation

During recent years, political scientists have talked much about the new corporatism (for example Wilensky 1976; Schmitter 1977; Schmitter & Lehmbruch 1979; Lehmbruch & Schmitter 1982; Harrison 1980; Armingeon 1983). According to the analysts of corporatism, the power structures of industrial relations have in many western democracies changed from pluralist into corporatist. According to Philippe Schmitter, these two types of interest intermediation can be differentiated as follows:

Pluralism can be defined as a system of interest intermediation in which the constituent units are organized into an unspecified number of multiple, voluntary, competitive, non-hierarchically ordered, and self-determined (as to type or scope of interest) categories that are not specifically licensed, recognized, subsidized, or otherwise controlled in leadership selection or interest articulation by the State and that do not exercise a monopoly of representational activity within their respective categories.

Corporatism can be defined as a system of interest intermediation in which the constituent
units are organized into a limited number of singular, compulsory, non-competitive,
hierarchically ordered, and functionally differentiated categories, recognized or licensed
(if not created) by the State and granted a deliberate representational monopoly wkhin
their respective categories in exchange for observing certain controls of their selection of
leaders and articulation of demands and supports.
(Schmitter 1977, 9)

Corpoiratization may also be analysed as a process that has taken place after
the Second World War in many developed capitalist countries. Especially the
following features seem to be typical of this process:

1. Well-organized employer and labour organizations collaborate with each
other and with the government. All economically important social groups
are represented in this co-operation by their interest organizations.

2. As time passes, the relations between organizations and the government become more and more centralized. It has been found that a centralist tendency in one sector of interest representation induces centralist tendencies in other sectors of interest representation (Wilensky 1976, 25). This centralist development seems to be more or less irreversible.

3. In this collaboration each group has a veto. The system of consensual decision-making guarantees that the interests of every collaborating group will be respected. It thus seems that optimization in this kind of a system is optimization basedon a Pareto criterion. There is therefore also something very conservative in decision-making of this kind.

4. In its internal decision-making, every group is relatively independent.

5. A principle of proportionality is used in the distribution of goods. Every
group therefore has an incentive to support economic growth.

6. A strong economic and social consensus is a precondition for democratic
corporatism. But consensual decision-making and proportionality in the
distribution of goods may also reproduce an atmosphere of social consensus.

7. In the co-operative decision-making between government and interest groups, the traditional line between private and public affairs becomes blurred. The distinction between private bargaining and public, governmental decision-making, also becomes blurred. Interest organizations have official positions in public administration. Governments also delegate tasks to interest organizations, and in the process the role of interest organizations changes. Traditionally interest organizations have articulated and aggregated interests; now they also execute the policy that has been agreed at the top level of decision-making.

8. In the corporatist arena there is a tendency to make large policy packages. In the implementation of proportional distribution every group tries to utilize the benefits of public policy while leaving the burdens to others. Public finance has been an important means in this redistributive game.

9. Economic growth, wages and salaries, prices, employment, balance of payments, government consumption, and transfer payments are important questions in the corporatist arena. Different policy sectors, such as social policy, regional policy and cultural policy will be absorbed into economic policy.

Consensus building and consensual decision-making seem to be important
features of the new corporatism. Therefore, I think, we may analyse new

corporatism as a way of controlling the distributive conflict and as a way of preventing problems induced by an open distributive conflict. Corporatism is a way of making interest organizations encompassing in the Olsonian meaning.

Hypotheses for Empirical Analyses

Using Olson's theory and the theory of new corporatism we can now make some hypotheses as regards the role of interest organizations, especially the role of trade unions and employers' organizations in explaining the distributive conflict.

Hypothesis 1: We suppose there to be a stronger economic consensus in
countries with large and encompassing labour unions.

Hypothesis 2: We suppose that the relations between unions contribute to the level of the distributive conflict. If the unions are competing for members in the same population we suggest that there would then be a higher level of the distributive conflict than in a situation where there is a clear division of labour between the unions (Armingeon 1982, 234; Franzmeyer 1982, 30-31; Seidel 1983).

Hypothesis 3: How does the rate of unionization influence the level of the distributive conflict? According to the traditional class theories there is a connection between the rate of unionization and labour militancy. When a large proportion of the labour force is unionized, then it is also powerful and militant (Marx 1958a; Marx 1958b). So it is supposed.

Using Olson's theory we could suppose that the number of distributional
coalitions increases with the rate of unionization.

With the above-mentioned arguments we can present the following
hypothesis 3a: There is a positive linear relationship between the rate of
unionization and the level of the distributive conflict.

Using theories of corporatism we can present an alternative hypothesis 3b: We may suppose that if there are favourable conditions for consensual decision-making, with a high rate of unionization supports this consensual style of life. On the other hand, if there are favourable conditions for an intensive distributive conflict, a high rate of unionization may still increase this conflict.

Hypothesis 4: In countries with corporatist incomes policies there has been a tendency to adopt centralized decision-making. We suppose that centralized federations are more encompassing than decentralized. Therefore we also suppose that the centralization of the labour unions contributes to a stronger economic consensus (see also Armingeon 1982, 234).

Hypothesis 5: The relationships between employers and the labour unions
are modified by the institutional structure of both parties. In the case of large
and encompassing organizations, with a high rate of unionization, and with

centralized power structure, we would expect to have consensual relations
between employers and the labour unions.

Hypothesis 6: Olson supposes that stable societies with unchanged borders tend to accumulate more interest organizations over time. We may now suppose that in cross-sectional comparisons the longer the time of stable democracy with unchanged borders in the country, the bigger the competition between labour unions and the higher the level of the distributive conflict.

Empirical Analyses

Data set. It is not easy to find empirical indicators to test all the abovementioned
hypotheses. In Table 2 I have collected data to test most of the
above-mentioned hypotheses.

We do not have data to test hypothesis 1 directly. This hypothesis will be
tested indirectly while testing the other hypotheses.

In order to test hypothesis 2 we need data on the competition between labour unions. In the third column (variable 7) in Table 2 we have a dichotomized index of this competition in the 19705. The figures are my own calculations based on many different sources (Kauppinen 1980a-m; OECE) 1979a; OECD 1979b; Stephens 1979; Blum 1981; Armingeon 1983). Countries have been classified into two groups, that of strong inter-union competition (code 2) and that of weak inter-union competition (code 1). If the division between unions is based on political factors, or if there is not a rather clear functional differentiation between unions, countries have been classified into the group of strong inter-union competition. On the other hand, if unions are functionally differentiated either on industrial or craft basis, countries have been classified into the group of weak inter-union competition.

In order to test hypothesis 3 we have data on the union membership as a percentage of the labour force in 1950 (variable 5) and 1970 (variable 6). The figures are taken from John Stephen's book "The Transition from Capitalism to Socialism" (Stephens 1979).

In order to test hypothesis 4 we need figures for the centralization of labour federations. In column four (variable 8) in Table 2 we have a measure of centralization of the labour federations. This is an additive scale constructed by Harold Wilensky (1976) based on the influence of central federations on collective bargaining, their power in the control of strike funds, their staff compared to the number of members and their share of collected dues. This data describes situation in the 19605. In some countries there have been clear changes in the power structures of labour federations after the 19605. I have corrected Wilensky's scale of the centralization in two cases. The influence of the central federations on collective bargaining has clearly increased in the Federal Republic of Germany and in Ireland.

We do not have figures for the power structures of employers' organizations.

But we have figures for the monopolization of production in each country. In column five (variable 9) in Table 2 we have an index of the monopolization of production constructed by Frederick Pry or (1973) and John Stephens (1979). The basic statistics used in the measure is a four-firm concentration ratio, that is, the percentage of employment, shipment, or production that is accounted for by the four largest enterprises in a narrowly defined industry. Using pairs of similarly defined industries, for the United States and each other country, weighted averages have been computed for each country.

We assume that the centralization of production correlates with the centralization
of employers' organizations.

The level of collective wage bargaining will be used to analyse the interplay between the partners of the labour market. In Table 2 (variable 10) we have constructed a trichotomized index of the level of wage bargaining in the 19705. Firstly, we have countries where collective wage bargaining is centralized on the national level. The central organizations play the first violin in the bargaining. In Table 2 these countries have been denoted with code 3. According to my classifications, wage bargaining was centralized in Austria, the Netherlands, Norway, Sweden and Finland.

Secondly, we have countries where collective wage bargaining is decentralized
down to the union or firm level. In these countries individual unions
carry on bargaining with little reference to each other. In Table 2 these

countries have been denoted with code 1. According to my classification,
wage bargaining was decentralized in Japan, France, Australia, Canada,
Italy and the United States.

In between we have countries with both centralist and decentralist elements in their collective wage bargaining. In Table 2 these countries have been denoted with code 2. According to my classification Germany, Switzerland, Belgium, Denmark, New Zealand, Ireland and the United Kingdom belonged to this middle group.

In column seven (variable 11) we have a measure of years of stable democracy with unchanged borders to be used in testing hypothesis 6. This measure is modified from Christopher Hewitt's index of full democracy (Hewitt 1977). Hewitt defines full democracy by the simultaneous fulfillment of the following requirements: universal male suffrage, secret ballot, and responsible government. But it is not the age of democracy as such that is of interest in testing Olson's theory. We are interested in uninterrupted periods of social stability with unchanged borders. Erich Weede (1983) has modified Hewitt's index for this purpose. War time occupations and after war changes of borders have been taken into consideration in this modification. In column seven (variable 11), then, we have the number of years of uninterrupted stable democracy with unchanged borders in each country before 1965.

In column eight (variable 12) and nine (variable 13) we have figures for
the openness and size of the national economies. We assume that they have an
effect on the centralization of economic organizations.

Corporatist and pluralist countries. Table 2 gives us a possibility to classify countries according to the style of interest intermediation. In a typical corporatist case we have a high rate of unionization, centralized labour federations, a low competition between different labour unions, and wage bargaining on national level. Austria, Norway, Sweden, Belgium and Finland fit very well into this model. The Netherlands is also a corporatized country with a somewhat smaller rate of unionization. Then we have some countries with typically corporatist styles of interest intermediation but on a more decentralized level. In the Federal Republic of Germany and in Denmark we have slightly less centralized power structures in the labour market compared to the above-mentioned countries. In the cases of Japan and Switzerland we may discern corporatist interest intermediation, too, but with extremely decentralized power structures.

We can now see that all countries with a strong economic consensus (measured with a low level of labour dispute activity) are more or less corporatized. Finland seems to be a deviant case. It is the only country with a weak economic consensus and with strongly corporatized interest intermediation.

In a typically pluralist case of interest intermediation we have decentralized

labour unions, competition between different labour unions, and wage bargaining on union or firm level. With the exception of Finland and Ireland, all countries with a weak economic consensus belong to this category. Ireland has a more or less pluralist background. But since 1970 wage bargaining have been centralized on the national level.

France and New Zealand also fit very well into the pluralist type.

Determinants of distributive conflict. Fig. 1 represents a causal interpretation of the relations between distributive conflict and the structures of economic organizations. Interpretation is based on the above-mentioned hypotheses and the data set in Table 2. In Figure 1 we have bivariate correlations between a logarithm transformation of the labour dispute activity in the 1970s (variable 2) and variables in Table 2. Techniques of causal modelling are not used for two reasons. Firstly, we have a large number of variables compared to the small number of cases. Secondly, we do not presume transitive causal relations, only. On the contrary, our causal interpretation includes both transitive causal relations, indicated with single-ended arrows, and reciprocal causal relations (feed back), indicated with double-ended arrows.

Table 2 and Figure 1 can be used to test the above-mentioned hypotheses.
Inter-union competition seems to increase the level of the distributive struggle,
as was supposed in hypothesis 2.

Centralization of the labour federations seems to lower the level of the distributive struggle, as was supposed in hypothesis 4. In a similar way centralization of collective wage bargaining seems to lower the level of the distributive struggle, as was supposed in hypothesis 5.

There is almost a zero correlation between the rate of unionization in 1970 and the level of the distributive struggle in the 19705. Is it now plausible to interprete this result as falsifying hypothesis 3a? I think not. The relationship between the rate of unionization and level of the distributive struggle should be analysed in a multivariate context. In Figure 1 we can see a strong positive correlation between centralization of the labour federations and the rate of unionization. In a similar way there is a positive correlation between centralization of wage bargaining and the rate of unionization. Centralization, in its turn, lowers the level of the distributive conflict. Therefore, we may suppose that the consensus-prone effect of centralization covers the conflictprone effect of a high rate of unionization. This hypothesis will be tested with the following regression equation:

(1)

where:

lnLD70_79 = Natural logarithm of variable 2
TUM = Variable 6
LWB = Variable 10

s.e. = Standard error of the partial regression coefficient
beta = Standardized partial regression coefficient

In this equation the partial regression coefficient of the rate of unionization is positive, although not statistically significant, while the partial regression coefficient of the centralization of wage bargaining is negative, and statistically almost significant. This equation gives some support to our hypothesis that the consensus-prone effect of centralization covers the conflict-prone effect of a high rate of unionization.

Equation 2 is an enlarged version of equation 1. Into equation 2 we have
included also our variable of the inter-union competition (variable 7).

(2)

where:

lUC = Variable 7

The overall level of explanation is far better in equation 2 compared to equation 1. This result as well as the beta coefficients of equation 2 show that competition between different labour unions is a more important predictor of the distributive conflict than the rate of unionization or centralization of industirial relations. According to this equation, too, centralization of the wage bargaining lowers the level of the distributive conflict and a high rate of unionization increases the level of the distributive conflict.

But what about the possible conditional relationship between the rate of unionization and the level of the distributive conflict? In hypothesis 3b we supposed that a high rate of unionization may be conducive to economic consensus in a consensual context and conducive to economic conflict in a conflict-prone context. We analyse this hypothesis by making separate analyses in countries with a strong economic consensus and in countries with a weak economic consensus. Spearman's rank-order correlation between the rate of unionization and the level of the distributive struggle is 0.25 in countries with a strong economic consensus, and —0.12 in countries with a weak economic consensus. The correlations are small, but they are not in the direction of our hypothesis, either. We would have supposed a negative correlation in countries with a strong economic consensus and a positive correlation in countries with a weak economic consensus. Our data set, then, does not support hypothesis 3b.

The above-mentioned analyses of hypotheses 2, 3 and 4 give us also somearguments to analyse "the logic" of corporatization. Let us begin with some facts. We have a strong positive correlation between centralization of labour federations and the rate of unionization. We also have a strong negative correlation between centralization of labour federations and the level of interunion competition. We can say that one purpose of the corporatist centralization is to prevent harmful effects of inter-union competition and to control the effects of labour militancy in highly organized labour markets.

This process of corporatization is, nevertheless, in a way ambivalent. The process of centralization makes labour federations more encompassing. It also makes the bargaining agendas more encompassing. Therefore all potential distributional coalitions want to participate. In this way centralized corporatism with encompassing bargaining agendas increases the rate of unionization and the number of actual distributional coalitions. And, ceteris paribus, an increase in the rate of unionization increases the level of the distributive conflict.

Let us now turn to the sixth hypothesis. In Table 2 we have variable 10 to test this hypothesis. According to the causal interpretation in Figure 1, the longevity of uninterrupted democratic tradition with unchanged borders has an effect both on the level of inter-union competition and on the centralization of industrial relations. Our independent variable measures time.

Are we now in a position to make historical explanations of the development of interest organizations? Is it plausible to say that as time passes inter-union competition tends to increase and labour federations tend to decentralize? I think we are not allowed to make historical explanations of this kind. One possible explanation is as follows.

In countries occupied during the Second World War many old interest organizations had been destroyed during the war. After the war new interest organizations were established. These organizations were established in a changed ideological and political atmosphere. New styles of macro-economic policies were becoming legitimate. Therefore the organizational structures of interest organizations were modified according to the needs of macro-level economic policies.

But there are also anomalous cases that do not fit into this theory. Both
in France and Italy labour federations are decentralized and inter-union
competition is remarkable.

It is also possible that there are some "deeper" cultural traditions that contribute to the correlation between stable democracy and centralization and to the correlation between stable democracy and inter-union competition. 1 here refer to the cultural differences between Anglo-Saxon and continental European countries. Kenneth Dyson (1980) has called Anglo-Saxon countries "stateless societies" and continental European countries "state societies". He has in mind the different role of state and government intervention in these types of societies.

Could we now say that the corporatist mode of interest intermediation is something that grows out of this continental tradition of "state societies", and that the pluralist mode of interest intermediation is something that grows out of the Anglo-Saxon tradition of "stateless societies"? At least we can say that all corporatist countries with centralized decision-making structures are continental European countries and that all Anglo-Saxon countries have a pluralist mode of interest intermediation.

In Table 2 we can see that there is a strong inter-union competition in every Anglo-Saxon country. And there are only two other countries, namely France and Italy, having a strong inter-union competition. In the case of centralization we can find that there are decentralized labour federations in all the Anglo-Saxon countries and decentralized wage bargaining in all the Anglo countries except Ireland, where wage bargaining has been centralized since 1970.

Resurgence of the distributive conflict. In the light of labour dispute figures in Table 1, we can say that in the 1970s there has been a resurgence of the distributive conflict in most developed capitalist countries. The average level of labour dispute activity rose in all countries with a weak economic consensus, and

in all countries with a medium economic consensus, except France. It rose also in Germany, in the Netherlands and in Sweden. The average level of labour disputes declined only in Austria, Norway, Switzerland and France. In Japan the level of labour disputes remained constant.

There was an increasing linear trend in labour dispute activity in all countries
except Austria, Denmark, France, Japan, Norway and Switzerland.

We may now ask the following question: Do the power structures of the labour market have an influence on the resurgence of the distributive conflict? We could suppose that there would have been a smaller or no resurgence in countries with a low rate of unionization, in countries with centralized industrial relations, and in countries with a weak competition between labour unions.

We will test these hypotheses with a regression equation having the linear trend of labour dispute activity (variable 4) as the dependent variable, and the rate of unionization in 1970 (variable 6), competition between labour unions (variable 7) and level of the collective wage bargaining (variable 10) as independent variables. We have the following regression equation:

(3)

where:

TLD = Variable 4

r = Bivariate correlation between dependent variable and the
predictor

The results of this regression are very similar compared to those of equation 2. Inter-union competition seems to explain rather well both cross-national differences in the level and the resurgence of the distributive conflict. In a multivariate context a high rate of unionization seems to be conducive to a high level of the distributive conflict and the resurgence of this conflict in the 19705. The only remarkable difference between equations 2 and 3 is that the level of collective wage bargaining seems to have no effect on the resurgence of the distributive conflict.

The relationship between the rate of unionization and the linear trend of labour dispute activity may also be analysed as a conditional relationship. We may suppose that a high rate of unionization increases distributive struggle especially in pluralist countries. We will test this hypothesis with bivariate correlations between the rate of unionization in 1970 and linear trend of

labour dispute activity in two subgroups, that of corporatist countries, and that of pluralist countries. In corporatist countries the correlation is 0.013, and in pluralist countries 0.713. The rate of unionization has an open militant effect in pluralist countries, only. In corporatist countries the militant effect of a high rate of unionization was controlled and covered by the corporatist structures of interest intermediation.

Conclusion

We have analysed the role of interest organizations in explaining crossnational differences in the level of the distributive conflict as well as differences in the resurgence of the distributive conflict in the 19705. Using Mancur Olson's new theory of interest organizations, we show on a theoretical level that a pluralist mode of interest intermediation is conducive to a high level of distributive conflict and that a corporatist mode of interest intermediation is conducive to a low level of distributive conflict. In fact, a corporatist mode of interest intermediation can be analysed as a way of controlling and decreasing the level of distributive conflict.

Empirical analyses in 18 developed capitalist countries have then been carried on to test the above-mentioned theory. These empirical analyses strongly support the theory. In the corporatist mode of interest intermediation competition between different labour unions is lower than in a pluralist mode of interest intermediation. In the corporatist mode of interest intermediation there has been a tendency to adopt centralized bargaining and decision-making. In centralized decision-making between all relevant organizations, the policy of pure redistribution loses its meaning. The policy of pure redistribution will not be accepted when all relevant interest groups are bargaining at the same table. Austria, Belgium, Finland, the Netherlands and Sweden are typical examples of centralized corporatist systems of interest intermediation. In the Federal Republic of Germany and in Denmark corporatist structures are a little more decentralized, and in Japan and Switzerland they are very decentralized. On the other hand, Australia, Canada, France, Ireland, Italy, New Zealand, the United Kingdom and the United States are countries with a pluralist mode of interest intermediation.

It is very possible that there are some "deep" cross-national cultural differences that explain the development of corporatism and pluralism in industrial relations. Kenneth Dyson has called Anglo-Saxon countries "stateless societies" and continental European countries "state societies". I suppose that the corporatist mode of interest intermediation is something that grows out of this continental tradition of "state societies" and that the pluralist mode of interest intermediation is something that grows out of the Anglo-Saxon tradition of "stateless societies".

Industrial relations are, of course, only one set of factors influencing the level of the distributive conflict. We can suppose that structures of the political arena as well as policy outputs have also an effect on the distributive conflict. We could suppose that a polarization of the electorate, fragmentation of the party system and intensity of the political competition also contribute to the level of the distributive conflict. The size of governing coalitions may also have an effect on the distributive conflict. Minimum winning coalitions probably increase the level of the distributive conflict while grand coalitions orientate to growth rather than to redistribution. We can also imagine that some policy outputs decrease the level of the distributive conflict while others increase it. And we can assume that there is a connection between economic outcomes and the level of the distributive conflict. To some extent there is a trade-off between growth and redistribution.

Therefore, industrial relations are just one, but an important set of factors
influencing the level of the distributive conflict.

REFERENCES

Armingeon, K. 1982. "Determining the Level of Wages: The Role of Parties and Trade
Unions". In Castles, Francis G. (ed.)- The Impact of Parlies. Politics and Policies in
Democratic Capitalist States. Beverly Hills: Sage.

Armingeon, K. 1983. Neo-Korporaiistische Einkommenspolitik. Eine vergleichende Unlersuchung
von Einkommenspolitiken in westeuropäischen Ländern in der 70er Jahren.
Frankfurt am Main: Haag -f Herchen Verlag.

Bell, D. 1974. The Coming of Post-Industrial Society. A Venture in Social Forecasting. London.

Blum, A.A. (ed.) 1981. International Handbook of Industrial Relations. Contemporary
Developments and Research. London: Aldwych Press.

Cox, A. (ed.) 1982. Politics, Policy and the European Recession. London and Basingstoke:
The Macmillan Press Ltd.

Crouch, C. 1983. "Corporative Industrial Relations and the Welfare State". In Jones, Barry
(ed.). Perspectives on Political Economy. London: Frances Pinter.

Crouch, C. & Pizzorno, A. (eds.) 1979. The Resurgence of Class Conflict in Western Europe
Since 1968, Vol. I, National Studies. London and Basingstoke: The Macmillan Press Ltd.

Dahrendorf, R. (ed.) 1982. Europe's Economy in Crisis. London: Weidenfeld and Nicolson.

Flora, P. & Heidenheimer, A.J. 1981. The Development of Welfare States in Europe and
America. New Brunswick and London: Transaction Books.

Franzmeyer, F. 1982. Approaches to Industrial Policy Within the EC and its Impact on
European Integration. Aldershot: Cover.

Goldthorpe, J.H. The Development of industrial society. Sociological Review Monograph
N: o 8.

Harrison, R.J. 1980. Pluralism and Corporatism. The Political Evolution of Modern
Democracies. London — Eioston — Sydney: George Allen & Unwin.

Headey, B.W. 1970. "Trade Unions and National Wage Policies". Journal of Politics 32
(1970): 2, 407-439.

Hewitt, C. 1977. "The Effect of Political Democracy and Social Democracy on Equality in
Industrial Societies: A Cross-National Comparison". American Sociological Review 42
(1977), 450-464.

Hibbs, D.A. 1978. "On the Political Economy of Long-Run Trends in Strike Activity."
British Journal of Political Science, 8 (1978): 2, 153-175.

Hirsch, F. & Goldthorpe, J.H. (eds.) 1978. The Political Economy of Inflation. Bungay,
Suffolk: Martin Robertson.

Holloway, J. & Picciotto, S. (eds.) 1978. State and Capital. A Marxist Debate. London.

Kauppinen, T. 1980a. Työelämän suhteet Belgiassa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980b. Työlämän suhteet Englannissa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980c. Työlämän suhteet Hollanissa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980d. Työlämän suhteet Italiassa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980e. Työelämän suhteet Itävallassa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980f. Työelämän suhteet Japanissa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980g. Työlämän suhteet Kanadassa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980h. Työelämän suhteet Norjassa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980L Työelämän suhteet Ranskassa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980j. Työelämän suhteet Ruotsissa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980k. Työelämän suhteet Saksan Liittotasavallassa. Helsinki: Työelämän
suhteiden neuvottelukunta.

Kauppinen, T. 19801. Työelämän suhteet Tanskassa. Helsinki: Työelämän suhteiden neuvottelukunta.

Kauppinen, T. 1980m. Työelämän suhteet Yhdysval/oissa. Helsinki: Työelämän suhteiden
neuvottelukunta.

Keynes, J.M. 1936. The General Theory of Employment, Interest, and Money. New York:
Harcourt, Brace & Co.

Lehmbruch, G. & Schmitter, P.C. (eds) 1982. Pattern of Corporatist Policy Making. Beverly
Hills and London: Sage.

Lipset, S.M. 1963. Political Man. The Social Basis of Politics. New York: Doubleday &
Company.

Marx, K. 1958a. "General Rules of the International Working Men's Association". In Marx,
Karl and Engels, Friedrich. Selected Works in Two Volumes, Volume I. Moscow: Foreign
Languages Publishing House.

Marx, K. 1958b. "Wages, Prices and Profit". In Marx, Karl and Engels, Friedrich. Selected
Works in Two Volumes, Volume I. Moscow: Foreign Languages Publishing House.

McCracken, P. et al. 1977. Towards Full Employment and Price Stability. Paris: OECD.

Medley, R. (ed.) 1982. The Politics of Inflation. A Comparative Analysis. New York etc.:
Pergamon Press.

Olson, M. 1982. The Rise and Decline of Nations. Economic Growth, Stagflation, and Social
Rigidities. New Haven and London: Yale University Press.

Organisation for Economic Co-operation and Development 1982a. The Challence of Unemployment.
A Report to Labour Ministers. Paris: OECD.

Organisation for Economic Co-operation and Development 1979a. Collective Bargaining and
Government Policies in Ten OECD countries. Austria, Canada, France, Germany, Italy,
Japan, New Zealand, Sweden, United Kingdom, United States. Paris: OECD.

Organisation for Economic Co-operation and Development. Labour Force Statistics. Several
volumes.

Organisation for Economic Co-operation and Development 1981. National Accounts of OECD
countries. Volume 11. Paris: OE'CD.

Organisation for Economic Co-operation and Development 1982b. National Accounts. Volume
I. Main Aggregates 1951-1980. Paris: OECD.

Organisation for Economic Co-operation and Development 1979b. Wage Policies and Collective
Bargaining Developments in Finland, Ireland and Norway. Paris: OECD.

Pryor, F. 1973. Property and Industrial Organization in Communist and Capitalist Countries.
Bloomington: Indiana University Press.

Samuelson, P.A. 1976. Economics. Tenth edition. Tokyo etc.: McGraw-Hill Kogakusha, Ltd.

Schmitter, P.C. 1977. "Modes, of Interest Intermediation and Models of Societal Change in
Western Europe". Comparative Political Studies, 10 (1977): 1, 7-38.

Schmitter, P.C. & Lehmbruch, G. (eds.) 1979. Trends Toward Corporatist Intermediation.
Beverly Hills and London: Sage.

Seidel, B. 1983. Wage Policy and European Integration. Aldershot: Gover.

Stephens, J.D. 1979. The Transition from Capitalism to Socialism. Trowbridge & Eshen:
The Macmillan Press Ltd.

Walsh, K. 1983. Strikes in Europe and the United States. Measurement and Incidence.
London: Frances Pinter.

Weede, E. 1983. Democracy, Creeping Socialism, and Ideological Socialism in Rent-Seeking
Societies. A paper prepared for the ECPR workshop on "Politico-Economic Interdependence
in the Highly Industrialized Societies".

Wilensky, H.L. 1976. The "New Corporatism", Centralization, and the Welfare State.
London and Beverly Hills: Sage.

Wilson, T. & Wilson, D.J. 1982. The Political Economy of the We/fare State. London —
Boston — Sydney: George Allen & Unwin.