Organizational de-integration of political parties and interest groups in Denmark

Abstract

Political parties and major economic interest groups often used to be closely linked, but over recent decades they seem to have become more and more detached. Until now, this process has primarily been described in almost deterministic structural models that tell us little about how this detachment takes place and imply that it affects all players at around the same time. The article analyses this development with a simple exchange model that explains changes in organizational integration between political parties and interest groups from its effect on their respective goal-seeking as collective actors. The model is tested on four sets of political parties and interest groups in Denmark in a longitudinal analysis covering developments since 1920. The variation found largely confirms the expectations of the model that the timing of the detachment across party families depends on the strength of resources both sides can derive from the other.

Keywords

Denmark interest groups public subsidies social cleavages

Introduction

During the second half of the twentieth century an organizational de-integration seemed to take place in advanced industrialized democracies between political parties and traditionally affiliated interest groups such as Social Democrats and trade unions and right-wing parties and business associations (Thomas, 2001). This article aims to explain such developments. It agrees with the argument that exogenous structural developments in recent decades have led to a situation in which political parties no longer represent clear and distinct groups in civil society in the ways they previously did – a development that also affects relations with interest groups (Katz and Mair, 1995). However, the article qualifies such arguments through an explicitly actor-oriented exchange model of organizational integration between political parties and interest groups, slightly adapted from Allern et al (2007). The model implies that political parties and interest groups loosen their relations when the perceived costs of upholding close institutional relationships exceed the benefits. When the model is applied across sets of political parties and interest organizations, relations do not loosen at the same time, as general structural models imply, but at different times depending on the specific costs and benefits associated with the particular relationships.

The article presents comparable data on development in relations between four political parties and interest groups in Denmark from 1920 to 2008: Social Democrats and trade unions, Liberals and farmers' associations, Conservatives and business associations, and, finally, Social Liberals and the smallholders' associations. With its long-term perspective, the article documents not only the decline in relations during the most recent decades but also during what was often assumed to be the ‘Golden Age’ of close ties before 1960. By including relations between centre-right parties, on the one hand, and farmers' and business associations, on the other, the article seeks to avoid the narrow focus on Social Democrats and trade unions often found in the literature. The main argument presented is that the costs and benefits for each political party and interest group of upholding organizational integration have been affected but not determined by general societal developments. As a consequence, organizational disintegration did not happen at the same time in all cases in Denmark. Furthermore, the article illustrates that political parties may, as autonomous actors, react differently to similar circumstances. Five sections follow. The next section presents the theoretical framework as well as the exchange model. Then comes a section on study design: the selection of the Danish case and empirical indicators. The third section studies the strength of party–interest group links, the dependent variable, and the fourth section analyses the proposed explanations according to the exchange model. The results are compiled and discussed in the conclusion.

Change in organizational integration between political parties and interest groups

In many Western European countries, strong links between major political parties and interest groups developed in the late nineteenth and early twentieth centuries as part of a mobilization of civil society and a struggle for the extension of democratic rights. The most cited example in the literature is undoubtedly the workers' movement with Social Democrats and labour unions, but Lijphart’s (1968) description of Dutch society points out how interest organizations were then important parts of all ‘pillars' in consociational democracies. While the literature on this topic is limited, it does empirically point out a general decline of party–interest group links, and, theoretically, a structural and a more actor-oriented explanation may be singled out (cf. Allern 2010; Allern and Bale, 2011; Thomas, 2001).

Structural explanations for change

Lipset and Rokkan (1967) noted how the party systems in a number of Western European countries in the 1960s were ‘frozen’ so that they reflected the societal cleavages of the 1920s. However, since the 1960s different authors have pointed out that general processes of post-industrial societal developments have led to value shifts in the electorate away from class-based voting (Fuchs and Klingemann, 1995: 11 ff.; Inglehart, 1987). According to Katz and Mair’s (1995) cartel party thesis, public subsidies have made the parties independent of economic resources from interest groups and correspond with an observation of de-linking between political parties and interest groups as representatives of civil society. Working with the Nordic party systems, Sundberg (2003: 89–94) expects developments like the ones described above to have weakened relations between parties and interest groups. Although he points to the different historical backgrounds of political parties, a general expectation is outlined: there were close relationships between political parties and interest groups during the age of the mass party and a general detachment took place during the age of the catch-all party and beyond. Accordingly, one should expect such a development to take place at around the same time across different parties and interest groups. True, structural developments like changes in employment patterns may vary across countries and sectors, but, by and large, emphasis on general structural developments leads to the assumption that party–group links have changed – declined – more or less simultaneously in established democracies.

Changes in party–interest group relations. An actor-oriented institutional explanatory model

In the general structural model, the ‘mechanics' of transformations take place at macro-level without any clear connection with the specific functioning and aims of political parties and interest groups. Therefore, we are dealing with a relatively crude model that may capture major developments but is not well suited to explaining differences between political parties and their relationship with interest groups. Party perceptions of usefulness of close relationships might be influenced but not determined by general structural developments.

In their exchange model, Allern et al (2007) argue that political parties and interest groups use institutional ties to overcome problems of non-simultaneous exchange in order to enjoy mutual benefits of a close relationship when they support each other in reaching their respective goals. Interest groups may provide different kinds of resources for a political party, such as money, manpower, specialized expertise and other types of political support in the form of sanctioning, defending and thereby legitimizing the policies of the party (or simply abstaining from or reducing overt criticism) (cf. Quinn, 2002: 210). Furthermore, the rank and file of the interest organization can be important as members and, not least, as voters of the political party. Likewise, the leaders of the interest groups can be an important channel for candidate recruitment. Political parties may be assumed to pursue multiple goals simultaneously; for example, policy, office and votes but votes may be considered an instrument for achieving the other (cf. Strøm, 1990). At the same time, the political party can further the goals of the interest groups by helping the latter obtain policy influence, not least when the party is in office (cf. Lehmbruch, 1984). Furthermore, interest groups also seek members. Close attachment to a certain political party may affect interest group membership either positively or negatively: political parties may encourage potential members to join the interest group but potential interest group members may also be repelled by the party attachment (cf. Svåsand and Rasch, 1997).

To sum up, political parties and interest groups may gain from mutual exchange in the form of resources and voters (as well as other adherents), on the one hand, and support, political influence, and members, on the other. Development in organizational integration occurs when:

... ‘goods' are delivered less reliably or become relatively less valuable to either the party or the unions (perhaps because of rising costs), then each partner’s cost–benefit calculation of the relationship will change, and one or both should lose interest in an institutionalised relationship. (Allern et al, 2007: 612)

If the structural model is right, de-linking in organizational integration between political parties and traditionally affiliated interest groups should occur across the board from around 1960s onwards. According to the exchange model, structural changes are not likely to have such a strong – and hence standardizing – impact on party–group relationships. Change in organizational integration occurs when the respective goal-seeking of the different parties and organizations is affected. Structural changes have decreased the general usefulness of the relationship for both sides, but how, and to what extent, may well vary across pairs of parties and interest groups. Most importantly, changes in the various relationships, i.e. the timing of the decline, are correlated with the development of the specific resources that each side can derive from the other by being linked as organizations. These are the general assumptions that I analyse in what follows. More specific implications are outlined afterwards, although changes in the costs involved will not be dealt with in any detail.

Case selection and measures

Examination of the argument requires us to look at an established industrialized democracy with a multiparty system and a history of organizational integration between political parties and interest groups. As a unitary state – with a high degree of linguistic, religious and ethnic homogeneity but still a multiparty system – Denmark provides a good and relevant case for cross-party comparisons. While still allowing a cross-party study of organizational integration, a within-country design ‘holds constant’ variables of the political system. By and large, it can be assumed that political parties and interest groups were exposed to the same general structural changes at the same time in Denmark (cf. Gerring, 2007: 97).

Early organization of Danish political parties was accompanied by the establishment of major urban and rural trade organizations as well as unions (Thomsen, 1992). To a large extent these dividing lines in the economic–organizational system came to be reflected in the party system, expressing similar interests at different levels of society (Elklit, 1988: 28). Other insightful political–historical observers, such as Rasmussen (1971: 172), Bregnsbo (1975: 170–1) and Thomsen (1992: 35), point to the very same four important ‘pairs' of political parties and major socio-economic organizations in Denmark: the Social Democrats and the trade unions, the Liberals and the farmers' associations, the Conservative People’s Party and the trade associations, and finally the Social Liberals and the smallholders' associations. These are the four relationships studied over time and in depth here.

Owing to the historical background, this study commences around 1920 when universal suffrage, proportional representation and the party- and class-based interest group systems were fully established, together with the four old parties that divided the electorate. From a structural perspective, one should expect the 1920s until the 1960s to be an era of strongly organized integration between political parties and interest groups, and the period from 1960s until today to be a period of weaker – if any – links. In order to cover each decade, data have been collected and are reported every eighth year from 1923 until 2004.

Organizational integration is measured by multiple indicators. Lijphart (1968: 61–6) listed how many of the leaders of each Dutch political party simultaneously held leadership positions in corresponding interest groups. In Sundberg (2001), a group of Nordic scholars¹ defined closeness between political parties and interest groups in terms of four variables: institutional ties (like joint committees); personal overlaps and transfer between leading governing bodies; and economic (and other material or manpower) support. In a Danish context, different authors have pointed to similar variables when selecting the most important characteristics of relations between the four old political parties and the major economic interest groups (Meyer, 1965: 150–8; Rasmussen, 1971: 163–75; Thomsen, 1992: 29). Thus, one could argue that the more these conditions are present between a political party and an interest group, the more organizational integration there is. They are all features that – at least over time – may be affected by party or group leadership decisions. However, as in Allern et al (2007), financial contributions are here treated as an independent variable, not as a part of the relationship itself.

Hence, the degree of organizational integration of the four old parties and four major economic interest groups in Denmark, is measured by, first, institutional ties, second, personal overlaps, and, third, personal transfer of leadership. In order to reduce complexity all indicators are kept at national level. Institutional ties are measured in two ways: formal statutes linking party and interest group – such as guaranteed access to mutual decision-making bodies – and non-statutory but regular joint leadership forums. Overlapping leadership is primarily measured at the level of the governing bodies of the four political parties and the four corresponding interest groups. The governing body of an interest organization is identified as its national committee; the leadership of the interest groups is operationalized as the members of their executive committees or governing boards. The following are included:

The Danish Confederation of Trade Unions (LO), as well as the members of the national committees of the three unions in Denmark, which over time has had most members.²

The Danish Employers' Association (DA).

The Agricultural Council (LR) and the board of the Danish Farmers' Unions (DDL).

The Danish Family Farmers' Association (DFL).³

Here, leadership of the political parties encompasses the national committees and parliamentary groups. The latter are included because they often wield more political influence than the leadership of the party organization, especially when it comes to centrist and centre-right parties. Only cases of overlapping leadership between party and interest groups holding full voting rights are included. Data have been collected using the primary sources: the official minutes of the organizations, official rules and party accounts mostly collected from party or organizational archives. Sources such as annual books, reference works, etc., have been used if the former were insufficient or unavailable.

Concerning personal transfer of leadership, the sources are official parliamentary biographies from which data are collected on the number of parliamentarians who, at the time of their election or at any time in their past, had either held elected office at any level or been employed in the corresponding type of interest organization.⁴ Such a background would indicate the relevance of an interest organization as a recruitment channel for the political party. Similar data have not been found on the party organizational leadership.

The independent variables of the study are found in the exchange model. For interest groups this is measured by, first, number of members provided to the interest group by the party and, second, policy influence. The former cannot be operationalized directly but has to be deduced from secondary literature. The latter is operationalized as voter support and years in government office by the political party. For political parties, the goal-promoting value of interest groups is measured as, first, vote potential and, second, as resources. Vote potential is measured as the share of the social group and the strength of its class voting.⁵ Resources are measured as financial contributions as they appear in party accounts (public since 1995). A number of older party accounts were found in archives.

Party–interest group integration in Denmark 1923 – 2004

Institutional ties

A formalized relation exists between a political party and an interest organization when they are linked by statutes. In Denmark, such a connection was found between the Social Democrats and other organizations within the labour movement but has never existed between the other three parties and interest groups.⁷

The Danish labour movement was founded as a single organization in 1871, but seven years later its two wings formally separated, although close cooperation continued (Dybdahl, 1969: 211). In 1920, statutory affiliations existed at leadership level between the Social Democrats and the Danish Confederation of Trade Unions (LO). Both organizations appointed two representatives to the national and executive committees of the other. These rules remained basically unchanged through the decades until they were removed in 1995 and 1996. Later, any references to each other were removed from the statutes. Over the years, the Social Democrats and LO have met in various informal joint committees, but such meetings ceased during the 1990s (Due and Madsen, 1996: 528).⁶ There are no indications of institutional ties between the other three parties and interest groups.

Personal overlaps in leadership

Table 1 shows that overlapping leadership between national committees of the political parties and those of each corresponding economic interest group once existed between all four pairs of political parties and interest groups, and that by 2004 it was all gone. However, this decline did not occur simultaneously. From 1939 and onwards overlaps were rare between the Conservatives and the DA, even if they did not cease completely (the last one was registered in 1988). Overlapping leadership between the national committee of the Social Liberals – a pretty large assembly for a small party – and the board of the DFL occurred for every selected year until 1980. Overlaps between the national committee of the Liberals and of the LR/DDL ceased after 1988. Finally, overlapping leadership existed until 1996 for all selected years between the national committees of LO and the Social Democrats, although primarily due to the statutory affiliations described above.

Table 1.

Overlapping leadership between the national committee of the four political parties in Denmark and the selected interest groups (LO/3 unions, DFL, DA, LR/DL). Percentages

	S	LO/unions	RV	DFL	KF^a	DA	V^b	LR/DDL
1923	7	5	3	15	na	na	na	na
1931	5	3	6	29	10	5	na	na
1939	6	3	3	13	0	0	na	na
1949	7	5	1	7	0	0	3	2
1957	4	3	1	6	3	2	5	4
1965	19	6	1	6	0	0	2	2
1973	6	2	1	6	1	2	2	2
1980	10	4	1	3	0	0	2	2
1988	12	4	0	0	1	2	2	2
1996	6	2	0	0	0	0	0	0
2004	0	0	0	0	0	0	0	0
N^c	37–73	84–167	66–114	13–38	29–83	19–60^d	40–61	42–55

^aKF formed a national committee in 1930.

^bV formed a national committee in 1948.

^cN represents the size of the respective bodies in this and the following tables.

^dThe N-level was 19–20 in 1996 and 2004, other years 55–60.

Source: Accounts based on minutes, calendars, etc., from the parties and organizations collected from central offices, archives and libraries by the author.

Table 2 presents overlapping leadership between the parliamentary groups of the four parties and the national committees of the corresponding interest groups. The trend above is apparently confirmed. There used to be numerous overlaps but they were all gone by 2004. Regarding the Social Liberals and the Conservatives, no overlaps were registered after 1965. Regarding the Liberals, no overlap was registered after 1973. Overlapping leadership between the Social Democrats' parliamentary group and the trade union leadership was found for every selected year from 1923 until 1996, again partly due to the statutory affiliations mentioned above.

Table 2.

Overlapping leadership between the parliamentary groups of four political parties in Denmark (S, RV, KF, V) and the national committees of the selected interest groups (LO/3 unions, DFL, DA, LR/DL). Percentages

	S	LO/unions	RV	DFL	KF	DA	V	LR/DDL
1923	6	5	8	15	10	7	6	12
1931	8	8	8	14	11	7	4	7
1939	10	10	5	7	3	2	8	7
1949	9	7	0	0	3	2	9	12
1957	5	3	0	0	0	0	7	6
1965	4	3	10	6	3	2	3	2
1973	3	1	0	0	0	0	7	3
1980	6	3	0	0	0	0	0	0
1988	4	1	0	0	0	0	0	0
1996	3	1	0	0	0	0	0	0
2004	0	0	0	0	0	0	0	0
N	55–99	84–167	7–27	13–38	15–45	19–60	22–83	42–55

Source: Accounts based on minutes, calendars, etc., from the Danish parliament (Folketinget) and the organizations collected from central offices, archives and libraries by the author. See also notes to Table 1.

Personal leadership transfer between political parties and organizations

Table 3 shows that the proportion of Social Democratic parliamentarians with a background in LO has fallen only slightly. Although the highest percentages were found at the beginning of the period, the figures have remained remarkably steady over three-quarters of a century. The same numbers regarding the Conservative People’s Party and the Liberals show a greater degree of change. The shares were more or less steady until the 1960s, and then the level drops somewhat. In the Social Liberals' parliamentary group, members with a personal background in smallholders' organizations have disappeared altogether.

Table 3.

Percentage of parliamentary party group presently or previously having held a leading position in traditional economic interest groups

	S/trade unions	RV/smallholders' associations	KF/employer or trade associations	V/agricultural associations
1923	40	19	31	22
1931	42	21	29	21
1939	33	14	21	19
1949	31	18	27	23
1957	35	29	20	11
1965	34	30	22	14
1973	34	4	19	20
1980	31	0	9	18
1988	25	10	23	5
1996	35	0	15	12
2004	32	0	13	10
N	55–99	7–27	15–45	22–83

Source: Based on parliamentary biographies issued in 1950 (Elberling, 1950) and then after each election (Folketinget efter valget, various issues).

Table 4 shows that the share of organization leaders with a background in an elected political office declined markedly between 1932 and 1999. Political parties have also declined as a recruitment channel for interest group leaders.

Table 4.

Share of interest group leaders with a political background. Percentages

Background in:	Employee associations			Business or farmers' associations
Background in:	Parliament	Central party org.	N	Parliament	Central party org.	N
1932	15	15	13	38	13	16
1963	14	0	22	18	5	22
1999	2	0	47	5	6	106

Source: Christiansen et al (2001: 68).

Overall decline of relations

Long-term trends in the variables presented above point to a general de-integration of organizational relations between the four ‘old parties' in Denmark and four major socio-economic interest groups. The strongest integration was found between the Social Democrats and the trade unions in the LO, which used to have statutory affiliations, overlapping membership of their governing bodies as well as widespread personal transfers between the traditional interest groups and the parties. The removal of statutory affiliations in 1995/1996 marked a major change in this relationship. The three non-socialist parties all had overlapping membership in the 1920s and a high share of leaders in either party or interest organization had a personal background in the other.

In the 1920s, the outset of the analysis, all four old parties were to a certain extent integrated with the major economic interest group that represented their main group of voters. If this was a ‘Golden Age’, it is certainly over. However, contrary to what much structural theory predicts, it ended at different times for the different partners and not simply from the 1960s onwards. The data indicate that close integration at leadership level between the Conservatives and the business associations ended as soon as the early 1930s. Dismantling of the relationship between the Social Liberals and the smallholders' associations was a continuous process from the 1960s until the end of the 1980s; likewise for the Liberals and the agricultural associations but with 1973 as the turning point. With the exception of the Conservative People’s Party and the trade associations, relations remained relatively stable from the 1920s until the 1960s. For the Social Democrats, close organizational integration did not end until the 1990s.

Explaining decline empirically with the exchange model

This section applies the variables of the exchange model to explain the variation in timing found above. If we assume that parties evaluate ties in terms of how likely they are to help them achieve their goals – and given the time pattern of change outlined above – we should expect benefits to have been markedly reduced and costs to have exceeded them from the 1930s when it comes to the Conservatives, from around the 1960s until the 1980s when it comes to the Liberals and Social Liberals and in the 1990s when it comes to the Social Democrats.

Interest group influence through political parties

The value of a political party for an interest group partly depends on how influential it has been over public policy. Even in Denmark, which from 1920 to 2004 had minority governments for 57 out of 80 years,⁸ cabinet office is an important proxy for political influence. Since 1920 the Social Democrats have been in office for 48 years (out of 80), the Liberals for 32 years, the Social Liberals for 31 years and the Conservatives for 20 years (14 years after 1982). The combined vote-share of the old parties has declined from above 95 percent in the 1920s to around 60–70 percent after the ‘earthquake election’ in 1973. Nevertheless, the old parties have continued to dominate government, although less so for the Social Democrats, who after 1982 have only held cabinet seats one-third of the time.

Interest groups and delivery of members by parties

The number of members the parties provide to interest groups can only be measured indirectly. The Social Democrats used to require its members to be member of a union if applicable, but not anymore. What do labour market statistics tell us? Union density in Denmark is comparatively high at around 70 percent of the wage earners (Visser, 2006: 45). It means that LO has a privileged position to defend. For most of the twentieth century, LO’s membership did not seem to be affected by its links to the Social Democrats. Particularly during the past decade, however, surveys have indicated and the LO leadership has openly admitted in the press that close bonds to the Social Democrats could cost members. Furthermore, close bonds may be an obstacle to a potential merger with the other confederations. Farmers' and smallholders' associations also seem to have organized fairly high shares among their rapidly shrinking pool of potential members. DA has always organized companies rather than personal members, so numbers cannot be compared directly, but its share has been lower than for the other major organizations but on the rise, especially because of fewer smaller, non-organized, companies.

Parties and interest group voters

Changes in social composition and decline in class voting are two important developments in the electorate that may have had an important impact on relations between political parties and interest groups. The social composition of the population has changed profoundly, as Table 5 shows. The share of workers first rose and then stagnated, whereas the other groups have declined steadily. Farmers now have very limited electoral significance. Smallholders, who traditionally owned less than 10 hectares of land each, only constitute a tiny share of the remaining farmers compared to almost half of the farmers around 1920 (Skrubbeltrang, 1954; Statistisk Årbog, several years). Meanwhile, partly due to the growth of the public sector, the group of functionaries has increased and now covers more than half of the population. This group has never developed close attachments to one particular party.

Table 5.

Share of working population of selected groups in Denmark 1920, 1960 and 1998. Percentages

	Urban workers	Farmers	Business owners
1920	24	28	20
1960	37	16	16
1998	35	3	7

Sources: (1987); Statistisk Årbog (1997).

Class voting has also declined. Ecological estimates by Thomsen (1987) and election survey data reported in Table 6 indicate that workers' class voting for the Social Democrats has declined in particular. In the 1920s, farmers and smallholders were largely divided between the Liberals and the Social Liberals by around four to one. This remained the case for a long time. Farmers still overwhelmingly support the Liberals, with more than 70 percent of their votes going to that party. The estimates indicate that the Social Liberals were never really a class party. The same was true of the Conservatives, whose share of business-owners, many of them owners of small shops, was less than 40 percent.

Table 6.

Support for traditional party among occupational groups. Percentages. Ecological estimates 1920 and 1960. Survey data 1998

	Workers voting for S	Farmers voting for RV	Business owners voting for KF	Farmers voting for V
1920	79	17	37	79
1960	69	11	37	82
1998	42	(0)	27	71

Sources: Thomsen (1987); Andersen (1999: 81).

In the 1920s, the two largest parties, the Social Democrats and the Liberals, clearly had a more distinct class-based support than the other two parties (cf. Bille, 1997). Unlike farmers, the share of urban workers among the total electorate has remained high and as such electorally attractive. These differences correspond with organizational integration between Liberals and farmer’s associations declining before those of Social Democrats and trade unions.

The smallholders lost relevance as an electoral group during the 1960s and 1970s in general and, thus, also for the Social Liberals. The share of business owners has also declined, although less dramatically than the number of farmers. Hence, also in the case of the Conservatives, the development of old core voters largely corresponds to the declining organizational integration in the long run. True, a major decline in relations between the Conservatives and the business associations took place as early as the 1930s, when there were still a substantial number of business owners around. But if we look at the role of the business associations, especially in industry, we can see that they were strong on money but relatively weak on votes. After the foundation of the party in 1916 as a merger of various conservative groupings, two wings emerged a ‘large-scale industry’ faction, which took the upper hand during the 1920s over the ‘middle estate’ faction appealing to electorally diverse groups of functionaries and small-scale business owners, but soon lost its grip (Fink, 2000). Changes at leadership level in the early 1930s did seem to entail a deliberate strategy by the winning ‘middle estate’ faction to get rid of the business elite from the leading positions within the party. So, unlike the other parties, from its very beginning the Conservative Party was inherently internally divided over its attachment to organized business groups with low potential to ‘deliver’ votes. As a consequence, the declining share of business owners could be expected to have a smaller effect on the organizational integration in this case than in the three others.

Parties and financial resources

Financial donations represent another of the benefits interest groups may provide to political parties. Since 1987 there has been direct public support to member organizations of Danish political parties. This support was quadrupled in 1996. Increasingly generous public funding has apparently made the political parties less dependent on interest groups. The Social Democrats and the Conservatives in particular relied heavily upon such benefits, but no longer.

Table 7 presents Danish party accounts compounding income values and public support. Until the 1980s the Social Democrats received an increasing share of their income from the other parts of the labour movement. LO has now stopped its direct party support and the same applies to most of its major member unions at national level.

Table 7.

Party income (Central Office) and from firms, associations and trusts (IG) and from public funds (Publ.) in Denmark 1923–2007. Percentages. Average annual total party income (DKK in 1,000, 2008 prices)

	Social Democrats			Social Liberals			Conservatives			Liberals
	IG	Publ.	DKK	IG	Publ.	DKK	IG	Publ.	DKK	IG	Publ.	DKK
1923–30	13	0	8,107	na	0	na	na	0	na	na	0	na
1931–38	15	0	10,132	27	0	1,556	na	0	na	na	0	na
1939–48	36	0	10,870	na	0	na	na	0	na	19	0	8,242
1949–56	27	0	14,929	33	0	3,056	na	0	na	23	0	10,102
1957–64	48	0	26,885	0	0	2,478	na	0	na	na	0	10,172
1965–72	60	0	41,900	4	0	4,341	na	0	na	na	0	na
1973–79	60	0	36,067	1	0	4,027	na	0	na	na	0	na
1980–87	71	0	48,257	0	0	2,555	na	0	na	na	0	na
1988–95	56	21	43,180	4	30	3,242	58	18	10,505	39	13	15,406
1996–03	38	48	59,136	5	73	5,469	30	55	17,465	16	59	37,900
2004–07	11	50	46,656	4	76	9,403	53	38	22,675	16	59	44,489

Sources: Bille (1997); various party accounts from archives and www.ft.dk. Some years missing.

In the Social Liberals' accounts there are no indications whatsoever of contributions from interest groups, among these the smallholders' associations, until the end of the 1990s; contributions before the 1960s came from newspapers. The party had a long-standing declared policy position that political parties and organizations should have separate economic affairs (Meyer, 1965: 156). A former party leader of the Conservatives noted in his memoirs that in the 1960s the party strengthened its finances by receiving money from business groups, among them the DA (Møller, 1974: 23). After a substantial decrease in this source of income during the 1990s the party – in office again from 2001 – enjoyed strong support from business. According to the Liberals' party accounts in the 1990s, at least some contributions came from the agricultural associations but also from employer associations and private companies. At the end of the 1990s these kinds of contributions made up only a modest share of party income.

Discussion of findings

Table 8 sums up and compares the timing of the decline in the organizational integration between political parties and interest groups with the development of the resources each side has been able to derive from the other. To what extent is there a concurrence between the two?

Table 8.

Timing of significant developments in organizational integration between four political parties and interest groups in Denmark since 1920

Dependent variable	S/LO	R/DFL	V/DL	K/DA
Strongest detachment	(1970s)–1990s	1960s(–1980s)	1960s(–1980s)	1930s(–1980s)
Independent variables
Significant loss of policy influence	1980s	1970s	1970s	1930s
Reduced share of membership pool for interest group	Increase until 2000s, then decline	No reduction	No reduction	No reduction
Voter group decline	No decline	Since 1960s	Since 1960s	Entire period
Class vote decline	Since 1970s	Small decline	No decline	Small decline
Reduced importance of economic support	1990s	No dependence	1990s: medium dependence	1990s

We can see, as documented above, that there is variation across parties and that this corresponds reasonably well with the timing of decline in organizational integration, although there might be a relatively short time-lag in certain cases. Organizational integration between LO and the Social Democrats remained attractive until they were out of office on a permanent basis in the 1980s and the organization could not ‘deliver’ in the form of voters. The other parties faced reduced support from core groups much earlier on. The capacity of interest groups and political parties respectively to deliver votes and policy influence appear to be particularly important factors explaining the development in relations, although money and members cannot be ruled out either.

Hence, the exchange model seems fairly well suited to explaining when decline in organizational integration between particular parties and interest groups has taken place in Denmark. Exogenous, structural developments do not determine links between parties and interest organizations even though they may affect the size of costs and benefits involved in particular party–interest group relationships. Danish parties have reacted differently to similar general external developments.

Conclusion

After the Danish four-party system was fully established around 1920, the parties were to a certain degree organizationally integrated with major economic interest groups through personal overlaps and transfers between leadership. Social Democrats and trade unions were also linked by formal statutes. Since then, a process of detachment took place for all parties. What used to be an important linkage to civil society has been severed. This would appear to be in accordance with the main proposition in Katz and Mair’s (1995) cartel party thesis. This is also what we expect from the literature on general structural changes affecting political parties. However, the data presented in this article allow us more nuance and precision when we try to explain such developments. The conclusion is that a simple exchange model accounts for much of the empirical variation found over time in the de-integration of organizational relations between four pairs of political parties and interest groups in Denmark from 1920 until today. Apparently, the exchange model has explanatory power beyond Social Democrats and trade unions studied byAllern et al (2007).

The exchange model is here primarily meant to capture long-term developments and their relative timing of change. As presented in this article the model is rooted in assumptions of rational institutionalism and blind to the fact that political parties and interest groups are not unitary actors but consist of human beings and factions, which may take conflicting views on various matters. Allern et al (2007) and Allern (2010) argue that some variation may also be explained by historical or normative strands of institutionalism, pointing out that institutions, once established, are hard to change – something that may result in the absence or impediment of changes in spite of structural developments or the fact that the actor’s perception of goals would itself be affected by the party–interest group relationship. The lag between explanatory factors and the dependant variable found in the article may support more historical institutionalist claims stressing impediments to general structural developments. There may also be historical ebbs and flows due to particular personalities or even different perceptions of the right course of action.

Just as importantly, development may not flow only in one direction. In recent years, for instance, a rapprochement between the Social Democrats and the LO involving more organizational integration appears to have taken place in Sweden and Norway, and more recently perhaps also in Denmark. The Danish case also showed, especially with regard to the Conservatives in the 1920s and 1930s, that views and powers of various factions did affect the development in organizational integration with business associations.

The exchange model has been tested by Allern et al (2007) across countries for Social Democrats and trade unions in the Scandinavian countries. This study extends its use across parties belonging to different ideological families within one of those countries and the findings largely support it. Even though the analysis could be improved by including data closer to the actors and further away from structural conditions, it strengthens the generally formulated exchange model. Future research should apply the model across both countries and party families/interest groups. Further testing could also usefully highlight non-economic cleavages such as culture and religion, especially given that the latter is known to be important in a number of countries with a strong Catholic church (Lijphart, 1968; Warner, 2000).

Footnotes

Acknowledgements

Thanks to the Danish State Archives, the Library and Archive of the Workers' Movement in Denmark as well as the parties and organizations for giving me access to their material. Thanks to colleagues and participants in the panels that discussed former versions of the article in Aarhus, Lisbon and Oslo, and to the anonymous referees. Errors or omissions remain the sole responsibility of the author.

Notes

References

Allern

(2010) Political Parties and Interest Groups in Norway. Colchester: ECPR Press.

Allern

Aylott

Christiansen

(2007) Social democrats and trade unions in Scandinavia. The decline and persistence of institutional relationships. European Journal of Political Research 46: 607–35.

Andersen

(1999) Sociale klasser og politiske partier. In: Andersen

Borre

Goulandersen

Nielsen

(eds) Vælgere med omtanke. En analyse af folketingsvalget 1998. Århus: Systime, 79–89.

Bille

(1997) Partier i forandring. Odense: University Press of Southern Denmark.

Bregnsbo

(1975) Interessegrupper. Copenhagen: Berlingske Forlag.

Christiansen

Møller

Togeby

(2001) Den danske elite. Copenhagen: Hans Reitzels Forlag.

Due

Steen Madsen

(1996) Socialdemokratiet og fagbevægelsen i en opbrudsperiode. In: Callesen

. (eds) Bidrag til Socialdemokratiets historie 1971–1996. Copenhagen: Fremad, 485–534.

Dybdahl

(1969) Partier og erhverv 1. Studier i partiorganisation og byerhvervenes politiske aktivitet 1880–1913. Århus: Erhvervsarkivet.

Elberling

(1950) Rigsdagens Medlemmer gennem 100 Aar I–III. Copenhagen: J. H. Schultz Forlag.

10.

Elklit

(1988) Fra åben til hemmelig afstemning – Aspekter af et partisystems udvikling. Århus: Politica.

11.

Fink

(2000) Storindustri eller middelstand. Det ideologiske opgør i Det konservative Folkeparti 1918–1920. Århus: Aarhus University Press.

12.

Folketinget efter valget. Various issues, 1990–2001.

13.

Folketinget. Håndbog. Various issues, 1956–1988.

14.

Fuchs

Klingemann

H-D

(1995) Citizen and the State: A changing relationship? In: Klingemann

H-D

Fuchs

(eds) Citizen and the State. Oxford: Oxford University Press, 1–23.

15.

Gerring

(2007) The case study: What it is and what it does. In: Boix

Stokes

(eds) The Oxford Handbook of Comparative Politics. Oxford: Oxford University Press, 90–122.

16.

Inglehart

(1987) Value change in industrial societies. American Political Science Review 81: 1289–303.

17.

Katz

Mair

(1995) Changing models of party organization and party democracy: The emergence of the cartel party. Party Politics 1: 5–28.

18.

Lehmbruch

(1984) Concertation and the structure of corporatist networks. In: Goldthorpe

(ed.) Order and Conflict in Contemporary Capitalism. Studies in the Political Economy of Western Nations. Oxford: Clarendon Press, 60–80.

19.

Lijphart

(1968) The Politics of Accommodation. Pluralism and Democracy in the Netherlands. Berkeley. CA: University of California Press.

20.

Lipset

Rokkan

(1967) Cleavage structures, party systems, and voter alignments. In: Lipset

Rokkan

(eds) Party Systems and Voter Alignments. Cross-National Perspectives. New York: Free Press, 1–64.

21.

Meyer

(1965) Politiske partier. Copenhagen: Nyt Nordisk Forlag – Arnold Busck.

22.

Møller

(1974) Politik på vrangen. Copenhagen: Stig Vendelkærs Forlag.

23.

Quinn

(2002) Block voting in the Labour Party: A political exchange model. Party Politics 8: 207–26.

24.

Rasmussen

(1971) Komparativ politik 1. Copenhagen: Gyldendal.

25.

Skrubbeltrang

(1954) Den danske husmand. Husmænd og husmandsbevægelse gennem tiderne II. Copenhagen: Det Danske Forlag.

26.

Statistisk Årbog, various issues.

27.

Strøm

(1990) A behavioural theory of competitive political parties. American Journal of Political Science 34: 565–98.

28.

Sundberg

(ed.) (2001) Partier och intresseorganisationer i Norden. Copenhagen: Nordic Council of Ministers.

29.

Sundberg

(2003) Parties as Organized Actors: The Transformation of the Scandinavian Three-Front Parties. Helsinki: Finnish Society of Sciences and Letters .

30.

Svå sand

Strøm

Rasch

(1997) Change and adaptation in party organization. In: Strøm

Svåsand

(eds) Challenges to Political Parties. The Case of Norway. Ann Arbor, MI: University of Michigan Press, 91–123.

31.

Thomas

(2001) Political Parties and Interest Groups: Shaping Democratic Governance. Boulder, CO: Lynne Reinner.

32.

Thomsen

(1992) Det politiske rodnet In: Andersen

Nielsen

Thomsen

Westerståhl

(eds) Vi og vore politikere. En kortlægning af forholdet mellem befolkningen og politikerne i dagens Danmark. Copenhagen: Spektrumv, 26–97.

33.

Thomsen

(1987) Danish Elections 1920–79. A Logit Approach to Ecological Analysis and Inference. Aarhus: Politica.

34.

Visser

(2006) Union membership statistics in 24 countries. Monthly Labor Review 129: 38–49.

35.

Warner

(2000) Confessions of an Interest Group. The Catholic Church and Political Parties in Europe. Princeton, NJ: Princeton University Press.