Resumé

Scandinavian countries are characterized by high electoral participation and high enrollment in interest organizations. Consequently, most citizens are represented in politics by both parties and organizations, and the problem is which kind of representation citizens prefer. Danish survey data show that members appreciate their interest organizations. Furthermore, the public wants government to stay out of industrial relations and to consult interest organizations. However, there are limits to the influence which the public permits organizations. Thus organizations should only have a say on policies in fields related to their narrowly defined interest and they should not be allowed to become too strong. The political problem is to balance the principle of parliamentary supremacy and the principle of taking due regard to all interests concerned.

Introduction

This paper deals with the acceptance by the public of interest group influence
in Danish politics. The point of departure is a set of characteristics of the
Scandinavian countries as strongly organized democracies, namely:

— The existence of a double system of representation. Citizens are represented
in politics by both parties and interest organizations.

— The existence of strong parallels between the structure of the party system
and the structure of the organizational system.

In most countries political parties coexist with interest organizations. This, however, does not by necessity imply that most citizens are represented via two channels. Whether this is the case or not depends on both electoral and organizational activity (Fig. 1):

---

* A draft version of this paper was presented for the symposium 'Problems of Governance in The Nordic Countries' organized by the Nordic Political Science Association in Reykjavik, Iceland, 21-28 August 1983. I wish to thank the participants and an anynomous reviewer for their valuable comments.

The Scandinavian countries are characterized by high electoral tournout and high organizational enrolment. Turnout is 80-90 percent without much variation according to social position, and enrolment, in for example trade unions, is almost equally high. In Denmark 75-80 per cent of the labour force (white and blue collar alike) are trade union members.^l

It can be concluded from these figures that at least two-thirds of the labour force are both trade unionists and party voters. This marks the Scandinavian countries off from most other countries, mostly due to the high level of organizational enrolment. In other countries, the party channel is often the only channel for a majority of citizens.²

Dual representation is a widespread phenomenon and a defining characteristic
of Scandinavian political systems. Duality is further accentuated by strong
structural parallels between the party system and the organizational system.

In the literature, organizations are often identified as representing functional interests, whereas parties are viewed as channels for territorial interest. But in Scandinavia parties as well as interest organizations seem to be organized around the same types of (functional) interests. Class voting has for a long time been a predominant characteristic of electoral behaviour. This is still the case, even if there has been some debate about the decline of class voting (Worre 1980, Andersen 1984). Further, territorial variations in party strength are of minor importance for electoral outcomes in systems such as in Denmark and Sweden with nationwide proportional representation.

At the same time, the main interest groups are occupational interest organizations. Trade unions are aligned with labour parties, agrarian organizations with agrarian parties, and employers organizations with some bourgeois parties.

Normally, interest groups focus on their own sectional interests, whereas political parties aggregate the interests of different groups and, in addition, have to take more collective interests into account. But the difference is not clear in all instances.

First, a number of political parties appeal to small segments of the electorate and do not in any way resemble catch-all parties. This is the case in a multiparty system like the Danish, with 9-11 parties in parliament during the last decade. The Communist party concentrates on militant workers, Left Socialists on students and intellectuals, the Socialist People's Party on the same groups plus public employees, etc. Furthermore, it might even be said that the trade

union federation (LO) combining skilled and unskilled sorkers, public and
private employees eta, has to aggregate different sectional interests to a greater
extent than such small parties.

Second, at least the trade union movement has a rather broad interpretation of its aims. Scandinavian trade unions have always been linked with the Social Democratic parties. The ideal of pure 'business trade unionism' does not correspond to Scandinavian realities. And today unions take stands on a wide range of political issues. Danish unions have policies with respect to general economic issues, primary education, peace in Europe etc., not only as a by-product of the Social Democratic connection. To some extent they compete with the Social Democratic party in representing working-class interests.³

The difference between parties and interest groups is still present, but primarily as a matter of degree. A conservative trade unionist often finds, whether he likes it or not, that not only his party but also his occupational interest group speaks on his behalf in matters related to defence policy, primary schools, etc. He is represented via two channels.

Top and Bottom Problems

Many studies of interest organizations have focused on the top level of politics: How are organizations represented in public bodies, what are the policies of governments vis-å-vis interest organizations? This paper takes a different road: The system of dual representation constitutes a distinctive feature of the political system, and the emergence of strong interest groups represents a deviance from traditional notions of parliamentary government. Therefore, it is of crucial importance to find out how the system of dual representation is perceived — accepted or rejected — by those represented through it. The paper deals with attitudes towards four different topics:

1. Own organization (satisfaction with the organization, disagreement with
it, etc.).

2. The organization of relations (a) between different organizations and (b)
between organizations and the state.

3. The proper domain for interest group activities.
4. The political power of the organizations.

There is no basis for a priori assumptions about citizens' attitudes. The ideal of government by parliament would lead to the supposition that citizens would accept that no organization should encroach upon the supremacy of elected politicians. On the other hand, it seems to be a widespread opinion that due regard should be given to all concerned, and that politicians to some extent should seek the consent of interest organizations.

Further, a consideration of selfish interests would lead most citizens to a positive attitude towards their own organization, its involvement in policymaking and its general power, but not necessarily towards interest groups in general. Employers, for example, fearing the influence of tråde unions on Social-Democratic govemments, might be inclined to oppose interest-group influence in politics.

On the other hand, if citizens give high priority to national interests (as opposed to sectional interests) rejection of interest-group influence would be a natural consequence. But they would not necessarily reject any involvement of organizations in the policymaking process. Some involvement may be seen as a means to the attainment of national consensus and national interests.

Finally, political parties have an almost automatic advantage over interest organizations: Discontented voters are free to leave the party — which guarantees some minimum agreement — whereas dissident members of organizations often have to stay.

The Data

The data to be used in the following originate from three different sources.

1. The Danish election surveys in 1979, 1981 and 1984.

2. A special survey of political participation and attitudes towards the political
system from 1979.4

3. A small follow-up survey in June 1984, conducted in connection with the
election to the European Parliament.

For the sake of brevity the paper will focus upon results for the whole population. But due to strong inter-group variations, reference will be made to findings for different occupations, party electorates and organizations. Especially the party split is important, as the overall pattern mainly corresponds to opinions in the centre of the political spectrum.

A short note on terminology is needed for reading the results. In the Scandinavian countries, trade unions and the employers' federations are bound by collective agreements for the whole term of the agreement (Elvander 1974a). This 'peace obligation' means that strikes and lockouts are ruled out most of the time, and they are heavily fined by a special labour court. In the following, strikes at the wrong time are referred to as 'illegal strikes'. It should be stressed that they are not violating criminal law, but only the collective agreement.

Satisfied Members

Although Danish organizations have a very high level of enrolment this does not, by itself, imply that interest organizations are supported in any active sense by large sections of the population. High enrolment may be due to other factors such as tradition, special benefits supplied to members or obligatory membership (de facto or de jure).

It is a dominant feature, however, that members of Danish interest organizations appreciate their organizations. In the 1979 study of political participation, members of interest organizations were asked about what importance they attached to their organization, whether they agreed or disagreed with the local and national leadership etc. The results are given in Table 1 (members of more than one organization were asked about the organization they thought was most important to them).

Inside the organizations, there are hardly any signs of what is often referred to as political distrust. Both the organizations and their leaders are popular.s This pattern of general satisfaction varies only a little from subgroup to subgroup (results are omitted). In general, voters at the political extremes — the left-wing and the Progress party — are more critical than voters in the centre of the political spectrum, and trade-union members are slightly less

enthusiastic than members of business organizations. But satisfaction
predominates in all groups.

In sum, the organizational system has a solid basis. In the case of trade unions, this can be thrown into relief by an Anglo-Danish comparison. The British Election study of 1979 asked respondents: 'In general, how good a job would you say the trade unions are doing for the country — very good, fairly good, not very good or not good at all?' Respondents were also asked whether the unions were doing a good job for their members. These questions were also included in the Danish election study of January 1984.

Table 2 reveals a striking similarity. In both countries, 71-73 per cent feel that unions are doing a good job for their members, with almost no variation between trade union members and non-members. This by itself indicates high membership satisfaction with a crucial part of union activities. The picture is different with respect to 'the job unions are doing for the country'. Here especially non-members criticize the unions and even members are less enthusiastic, but with a substantial difference between the two countries. Whereas only half of British union members think that trade unions are doing a good job for the country, a clear majority of Danish trade unionists share that view.

Relations between Organizations and Government

Equally solid is the basis for the public activities for the organizations. Two fields will be considered: The labour market, where collective agreements traditionally are made between organizations, and legislation in spheres of interest to the organizations. Two problems are raised: Should organizations interfere with the activities of government? And should government interfere

with the activities of the organizations? Fig. 2 presents four logical possibilities.

Only the two first rows are relevant in Danish politics. Danish interest organizations are represented on commissions, consulted by politicians and administrators and are thus clearly involved in politics (Buksti 1984a, Fivelsdal 1979, Kristensen & Johansen 1982). But Danish politicians oscillate betwen interference (row 1) and non-interference (row 2). On the one hand, it is common practice for Danish governments to proclaim their attachment to the principle of free collective bargaining in the field of industrial relations. On the other hand, it is also quite common that they put limits on wage increases, and — if negotiations break down — that they settle wages by law.⁶

All this refers to the elite level and voters may differ in their views. It could perhaps be argued that citizens would endorse the traditional idea of separation of politics and economics (including industrial relations) which keeps organizations out of politics and also opposes government intervention in the non-governmental sector (row 4). But in accordance with the general member satisfaction mentioned above, citizens quite often take stands favouring organizations the most (row 2). A substantial proportion of the electorate supports both autonomy for organizations in the labour market and consultation of organizations by government.

Table 3 shows widespread support for the principle that wage agreements should be made without state intervention (Questions 1-3), but there are limits. First (question 8), opinions are divided as to what should happen when an agreement cannot be reached and industrial conflicts are looming. Second (questions 10 and 12), only a few accept that illegal strikes should be used as a means to press the matter. The results also show some ambiguities. Almost half the population is in favour of close cooperation between government and labour market organizations (question 4), despite the fact that a clear majority rejects state interference in free collective bargaining.

Text of abbreviated questions:

8. And then a question about how the state should react to a threat of industrial conflict:
A Says: Government should interfere and decide matters before the conflict breaks out.
B Says: If the workers have rejected a proposal for a new agreement they should naturally

be allowed to strike to get their claims carried through.

9. Next a question about police action against physical blockades (picketing).
A Says: The police should be much tougher on physical blockades.

B Says: The police should not interfere in industrial disputes. At least the police should not be
tougher than it is today.

10. Let us assume that parliament has decided matters in an industrial conflict in a way with
which the wage-earners have reasons to disagree.

A Says: The Wage-earners should go on strike even if the strike is illegal.

B Says: Neither the wage-earners nor the employers should disobey a decision in parliament
dealing with an industrial dispute, not even when they have good reasons to be discontent.

These details apart, it seems to be widely accepted that labour market organizations ought to have a domain of their own — at least if they handle affairs efficiently. This is no banality. An incomes policy, for example, may be hard to reconcile with the idea that the main area of income determination should be kept out of politics.

However, voters from various political quarters differ about the size of
domain (Table 4), and two party groupings deviate clearly from the 'average
pattern':

— The left wing hardly puts any limits on the actions of organizations and
their members.

— The right wing, especially Agrarian Liberals and Conservatives, agree in principle that industrial relations should be handled by the (labour market) organizations, but they put almost no restrictions on state interventions, especially not when conflicts are looming.

Whereas the left-wing point of view represents a complete denial of the authoril y
of the state, this authority is stressed by the right-wing.

The Social Democratic view is more mixed. A number of ingredients can
be distinguished:

— The organizations should handle affairs themselves, even if that implies industrial conflicts. The latter is at variance with the right-wing point of view. — However, if the state intervenes, its authority should be respected. Illegal actions are not accepted. This is at variance with the leftwing point of view.

Social Democrats accept a division of labour, but not confrontations with the
state. The remaining parties — all minor ones — are placed somewhere between
the Social Democrats and the right wing.

The conclusion for Social Democratic voters also holds for trade unionists, for workers and for employees (Table 5). All these groups are in favour of labour market autonomy. And all reject illegal behaviour. Left Wing calls for militant action are not likely to appeal to a majority opinion among workers and employers.

Part of the left-right difference is easily explained. After all, it is a foregone
conclusion that the left is more in favour of trade unions, strikes etc. than the

right. So far, the most interesting point is the opposition of Social Democrats
(and workers and trade unionists) against illegal strikes.

However, not all questions dealt specifically with trade unions and strikes, but referred to collective bargaining in general and to the organizations on the labour market (See questions 1-4). And the left-right difference was found in these cases as well. So there are reasons to assume that the left is more positive towards organizations, at least in the field of industrial relations. And probably for good reasons. C. Offe argues that the success of managers, factor owners, doctors etc. first of all depends on their own abilities, whereas ordinary wage

earners have to rely more on collective action, and he adds that state intervention
often aims at moderation of wage claims (Offe 1981).

The widespread acceptance of organizations is not confined to collective agreements. Many citizens favour a more general reliance on the consent of organizations in the field of legislation. It is sometimes claimed that Parliament is reduced to a rubber stamp when a minister has reached an agreement with interest organizations. But a clear majority of the public wants parliament to accept agreements with the Association of Municipalities without making amendents, and a majority of the same size is in favour of letting Parliament abide to agreements between trade unions and the Employers' Federation (Table 6).

Once again, however, there are wide variations between the view in workingclass and bourgeois layers. The former is very much in favour of leaving as much as possible to the organizations, whereas feelings are much more mixed on the bourgeois side (Table 7).

Both left-wing woters and Social Democrats have positive majorities on every question concerning involvement of and delegation to the organizations. In the same way workers are very inclined to accept organizational involement and so are, according to other criteria, trade unionists (results omitted). And these questions do not refer specifically to trade unions.

Among the major bourgeois group — Agrarians and Conservatives —
feelings are only positive towards agreements which have already been reached,
but in these parties the following is not accepted:

— That government is obliged to compromise with organizations.
— That politicians in all instances should respect the will of the organizations.
— That wide fields of policy-making should be delegated to the organizations.

The middle classes (salaried employees, independents and farmers) and members of business organizations are close to this description (results omitted). The most interesting point is the position of the white-collar employees. Here they differ from the workers.

In conclusion, left wing and Social Democratic voters take stands that are very favourable towards the organizations (row 2 in Fig. 2). And using other criteria, the conclusion for Social Democratic voters also holds for workers, trade unionists and — with some reservations — for the general public. Government should, according to the views in these groups, refrain from interference with free collective bargaining and it should seek to reach agreements with the organizations.

The position of bourgois voters and of the middle classes is more ambiguous. Certainly, free collective bargaining is accepted, but hardly if it results in industrial conflicts. And whereas agreements with and between organizations are accepted, it is not considered obligatory to obtain the consent of organizations. However, compared with the leftist view there is no doubt that bourgeois voters are more inclined to support the authority of the state vis-å-vis the organizations.

The Proper Scope of Interest Group Influence

So far, the results indicate widespread public acceptance of interest-group influence. But this is only one side of the coin. There are also important restrictions on the amount of influence which the public is willing to allow interest organizations. The first of these relates to the fields within which organizations should have a say.

As mentioned above, Scandinavian trade unions do not limit their activities to industrial relations in the narrow sense of the term. Danish unions have educational policies, are in favour of deficit spending and are troubled about peace in Europe.7

Therefore, in a small follow-up survey in June 1984, respondents were given a list of policy fields and were asked, with respect to each, whether the labour market organizations, trade unions and employers federations, should have a say on policies or leave this to the political parties. In Table 8 those without a definite opinion have been left out. Therefore, those not in favour of organizational influence are in favour of decisions by political parties only.

What emerges is a very clear-cut picture: Labour market organizations should restrict themselves to labour market issues and leave other problems to political parties. This definition of relevant fields is so narrow that some or even most trade unions would have difficulties in mastering even their most narrowly-defined tasks in the labour market. Public expenditures have a crucial impact on the demand for labour (Crouch 1982, 190-220) and organizations can hardly avoid dealing with the subject, and least of all in a country like Denmark with a third of the labour force employed in the public sector. Compared with the practice of Danish interest organizations, the public seems on this point to prefer a more restricted role.

It should further be noted that this narrow definition is found in almost all groups. The left is more positive towards organizations than the right, but not even among left-wing voters is a majority in favour of permitting the labour market organizations a say on anything beyond the labour market.⁸

In sum, although the public favours consultation with organizations, it is
very far from giving them a role on a par with political parties. Votes should
count and resources should not decide.

The Power of Organizations

There is a further restriction on the role given by the public to interest organizations: They must not be too strong. The 1979 study of political participation asked respondents whether they thought that politicians paid too much or too little regard to the views of the organizations. Results are given in Table 9.

Not all interest organizations are seen as having too much power. However,
the two major labour market organizations (LO and DA), especially the LO
are by a large margin considered too powerful.

This result is found in most subgroups of the population (Table 10). The group differences reflect normal left-right and status differences. The political influence of LO is especially unpopular among bourgeois voters and among self-employed and farmers, whereas the reverse holds true for the Employers' Federation (DA).

But there is a pronounced bias against the political influence of LO. Whereas employees would like to see the power of the Association of Employees increased, and farmers are convinced that the politicians ought to listen a good deal more to the Council of Agriculture, workers are in fact very modest on behalf of their organization. They think that LO has too much influence. And even trade unionists oppose the influence of LO and do so to almost the same extent as they oppose the influence of the Employers' Federation. Among the different political groupings, only the left wing wants the influence of LO increased. Among Social Democratic voters the situation is the same as among trade-union members: there is almost equal opposition against the political power of LO and the power of the Employers' Federation.

Even more marked results were found in the election study 1981. Voters were asked whether different organizations had too much or too little power in society. This question does not only refer to the responsiveness of politicians, and the balance of opinion was somewhat more negative towards the organizations.

In the 1981 survey the list of organizations contained both national organizations and their local branches. Results are given in Table 11. What emerges is a substantial difference between evaluation of the power of LO and the power of local trade unions. Both are seen as having too much power, but the latter much less so than the former — and also less so than the Employers' Federation.

The negative perception of the influence of especially LO might be due to a
general dislike of trade unions, a dislike which should be higher than the

dislike of other organizations. However, there are no indications that tråde unions are unpopular. On the contrary, trade-union members appreciate their organization, and most of them are actually affilliated to a LO union. Further, a majority found that tråde unions were doing a good job for both their members and the country. Finally, at the election in 1979 a majority gave LO positive scores on a sympathy scale and there was almost no difference between the scores given to LO and the scores given to the Employers' Federation (Nielsen 1982). Equal ratings were also found in the study of political participation in 1979.

It is more likely that opposition to the power of LO is caused by a perception — right or wrong — of LO as surpassing other organizations in influence as regards politics and society. The study of political participation asked which organizations had the biggest, next-biggest and third-biggest influence on political decisions. 72 percent of the respondents reported LO as their first choice. And this interpretation makes sense of the respondents more positive evaluation of the power of local tradeunions compared with evaluations of the power of LO. LO is clearly more powerful in society than local trade unions.₉ As a consequence, opposition towards such power should be directed against LO rather than against local branches, whereas general dislike of the unions should lead to equally negative evaluations of the local and the top level.

Conclusion

The constellation of attitudes outlined poses problems as regards almost any line of policy: Governments may grant organizations far reaching independence and they may follow the advice of the organizations, in which case they risk reproaches that they are giving the organizations too much power. Or they may intervene in collective bargaining, stop listening to the organizations etc. Then they risk criticism for not being responsive enough.

Much of the difficulty seems to stem from the dilemma between upholding
principles of parliamentary supremacy and at the same time giving due regard
to all interests involved.

To a certain extent, the dilemma is a Social Democratic problem in particular. Left-wing voters are clearly one-sided in their stress on the autonomy of organizations. Such views may present problems too but hardly point to conflicting opinions. And the benevolence of right-wing voters towards organizations is clearly restricted. But Social Democratic and neighbouring bourgeois voters seem to want it both ways, wishing involvement of organizations but opposing their influence.

The dilemma is a Social Democratic problem also in another sense. It was
first of all the power and influence of LO which was criticized, not — or not to
the same extent — the power and influence of, for example, the Employers'

Federation or the Agricultural Council. This does not ease the situation for the Social Democratic party, at least not when it is in government. But the problem should not be seen in isolation. Other results have shown much higher evaluations of the Social Democratic Party than of LO (Nielsen 1982). This does not remove the difficulties. Elut not all odds are against the political parties.

---

NOTES

1. The figures given by Beyme (1980, 75) do not register the accellerated unionization during the late 1970'5. ,

2. UK has a slightly lower level of turn-out (70-75 per cent) and, more pronounced, a lower level of enrolment (50 per cent). Most other countries have even lower enrolment in tradeunions than the UK (Germany and the Netherlands about 40 per cent, Italy and France 20 per cent). The concrete figures refer to trade unions. It is harder to get figures with respect to other types of organizations. However, it is a safe guess that other sections of the population are also highly organized in Scandinavia.

3. This may create problems because the unions have a politically heterogenous membership. At the election in 1984, for example, only 46 per cent of all members of Danish LO-unions voted for the Social Democratic Party, 21 per cent supported a left-wing party and 33 per cent a bourgeois party.

4. The author is grateful to the group conducting the latter study for access to their results. Special thanks should be given to Erik Damgaard, Jørgen Goul Anders" and Lars Nørby Johansen.

5. This impression is underlined by the finding that only 9 per cent of the members state that they planned to leave the organizations, against 89 per cent answering definitely 'no'. To some extent this is probably due to selective goods in the sense of Mancur Olson (Olson 1965) and not to any identification with the general goals of the organization. On the basis of the study of political participation, Buksti shows that 55 per cent of the passive members cite individual benefits and services as the chief reason for their membership, agaist 45 per cent giving other reasons. Among the more active members the picture is different. Only 33 per cent refer to individual benefits and services against 67 per cent giving other reasons (Buksti 1984b, 72).

6. At this point there is a marked difference between Sweden and Denmark. In general, Swedish governments have avoided being held responsible for the conduct of collective bargaining and managed to keep out of labour market negotiations (Elvander 1975b).

7. Thus, the General Workers Union (SID) has a 'peace programme' and LO has both given peace movements financial support and arranged 5 minutes stoppages for the cause of peace.

8. Only the very small and extreme left-wing parties (Left Socialists and Communists), not the Socialist People's Party, differ in this respect.

9. There is no strong shop-steward movement, nor extremely powerful locals. But locals have some say, because the details of wage settlements in some branches are negotiated in the individual firm and the national wage agreements in these cases only stipulate minimum wages.

REFERENCES

Andersen, J. 1984. 'Decline of Class Voting or Change in Class Voting?: Social Classes and
Party Choice in Denmark in the 1970'5, European Journal of Political Research 12, 243-280.

Beyme, K. 1980. Challenge to Power. Trade Unions and Industrial Relations in Capitalist
Countries. London & Beverly Hills: Sage.

Buksti, J. 1984a Organisationernes Folk. Aarhus: Politica.

Buksti, J. 1984b. 'Interesseorganisationer i politik', in Erik Damgaard, ed. Dansk demokrati under
forandring. Copenhagen: Schultz.

Crouch, C. 1982. Trade Unions: The Logic of Collective Action. London: Fontana Books.

Elvander, N. 1974a. 'Collective Bargaining and Incomes Policy in the Nordic Countries:
A Comparative Analysis', British Journal of Industrial Relations, 417-437.

Elvander, N. 1974b. 'The Role of the State in Settlement of Labour Disputes in the Nordic
Countries: A Comparative Analysis', European Journal of Political Research, 363-383.

Fivelsdal, E. et.al. 1979. Interesseorganisationer og Centraladministration. Copenhagen: Nyt fra
Samfundsvidenskaberne.

Johansen, L. & Kristensen, O. 1982. 'Corporatist Traits in Denmark, 1946-1976', in Lembruch
& Schmitter, ed., Patterns of Corporatist Policy-Making. London & Beverly Hills: Sage.

Nielsen, H.J. 1982. 'Electoral Politics and The Corporate System: The Question of Support',
Scandinavian Political Studies 5, 43-65.

Olson, M. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups.
Cambridge, Mass.: Harvard University Press.

Offe, C. 1981. 'The Attribution of Public Status to Interest Groups: Observations on the West
German Case', in Susan Berger, ed., Organizing Interest in Western Europe. Cambridge:
Cambridge University Press.

Worre, T. 1980. 'Class Parties and Class Voting in the Scandinavian Countries', Scandinavian
Political Studies 3, 299-320.