New Parties in Government: Party Organisation and the Costs of Public Office

Abstract

Previous studies suggest, and common wisdom holds, that government participation is detrimental for new parties. This paper argues that the opposite is true. Drawing on a large-N analysis (111 parties in 16 countries) in combination with two case studies, it demonstrates that new parties generally benefit organisationally from supporting or entering a government coalition. Compared to established parties, new parties have the advantage that their leadership is more able to allocate effectively the spoils of office, and can change still malleable rudimentary party structures so as to respond to intra-organisational demands, as well as the functional demands of holding office. The authors conclude by setting their finding in wider perspective and elaborate on its implications for contemporary West European politics.

Acknowledgements

We are grateful to the two referees and the editors of the journal for their helpful suggestions to improve the article. Part of the research was conducted during a Marie Curie Fellowship held by N. Bolleyer at Leiden University (IEF project number 236894). The research was further supported by the Economic and Social Research Council (RES-239-25-0032).

Notes

 1. We refer to the Austrian FPÖ, the Dutch List Pim Fortuyn and New Democracy in Sweden. While the first two parties were formal coalition partners, New Democracy had a pivotal position in parliament and functioned as support party for the minority government in office.

 2. This line of argument obviously presupposes that party leaders do not behave in a completely myopic, short-term manner. We consider this short-term focus unlikely. Even if the leadership of a party sees the organisation primarily as vehicle for boosting their individual careers, this usually implies a time horizon that requires the party to perform reasonably well for at least a few years. An acceptable performance in public office – the party unit that receives most media attention and is most visible to the electorate – is an important part of this.

 3. Janda (1983: 326–8) proposes four concepts and a range of indicators to capture each to describe a party’s ‘internal organisation’. These concepts (e.g. degree of organisation or coherence) cannot easily be assigned to the two dimensions – routinisation and value infusion – we intend to capture in the two in-depth analyses. Nor can their indicators, which are linked to a focus on the comparative assessment of established parties, and their purpose to compile databases allow for medium- and large-N studies. Similar issues arise with the indicators developed by Gibson et al. (1983) for the analysis of party strength in the US.

 4. See for a similar perspective Harmel and Svåsand (1993), who specify the demands towards the party leadership of new parties in different phases of party institutionalisation.

 5. ‘Promoter parties’, whose major objective is to bring attention to a particular issue but which do not succeed electorally, are omitted (Harmel and Robertson 1985: 517).

 6. The Christian Democratic Appeal (CDA) is considered a new party, as it was a merger of three parties that had not developed organisational ties among themselves until only a few years before the merger in 1977 (Kroeger and Stam 1998; Napel 1992; van Kersbergen 1993). It could thus not build on a working organisational infrastructure.

 7. We consider both formal participation in government coalitions (i.e. involving the taking over of ministries) and entering a support agreement as a form of government participation. While the pressure imposed on a newcomer is lower when ‘only’ functioning as support party (since the latter is not in charge of any ministries), the pressure to function as a parliamentary group to deliver support, to show expertise in the law-making process and to actively cooperate with mainstream parties in government is directly comparable. Clearly, this involves a move beyond mere protest status and the party will be evaluated in terms of its capacity to use its governmental access to realise its goals.

 8. Note that the wording of the question (see Appendix) is too vague to measure party institutionalisation as a concept, let alone capture its different components, which is why we use the quantitative analysis only to provide a first overview.

 9. See the Appendix for the relevant results of both surveys.

10. For several reasons, the results from expert surveys for the measurement of the level of party organisation are advantageous. Alternative ways to measure the relevant party characteristics include secondary reading, mass surveys and elite studies. Yet expert survey data have at least four advantages over data derived from these sources (see Mair 2001: 17, 24, for a discussion of the pros and cons of using experts’ judgements as opposed to using other ways of collecting data).

11. The definition of ‘new party’ leads us to code all parties in Greece, Portugal and Spain as new, because their party systems date back to later than 1965. Leaving these three countries out of the analysis reduces the number of cases to 96 but does not substantially change our conclusions. Results are available upon request from the authors.

12. In accordance with the relevant literature (e.g. De Winter and Swyngedouw 1999), we treat Belgium as consisting of two different political systems, as Flanders and Wallonia have party systems that are completely separate. The 16 countries thus add up to 17 political systems.

13. Among the six new parties that entered or supported government between 2000 and 2004, the CCD-CDU in Italy and the Dutch CDA can be classified as ‘alliances’ (Mair 1999). As mentioned above, the CDA is actually a merger of three ‘old’ parties. Excluding the CDA, the CCD-CDU or both cases from the analysis does not alter our finding in any substantial way, however. Indeed, when both parties are left out, the coefficient of the interaction of government access and new party becomes larger (b = 1.42, p < 0.05, two-tailed).

14. Harmel and Svåsand (1993: 71) introduced the concept in their study of the Danish and the Norwegian Progress Party.

15. These interviews were conducted face-to-face in 2009, lasted on average 30 minutes and were taped. Given the sensitivity of intra-party matters, the anonymity of interviewees had to be guaranteed.

16. Since winning 22 seats in 2001, the DF has repeatedly functioned as a support party and did so successfully: in the 2005 and 2007 general elections the party increased its electoral support to 24 and 25 seats respectively.

17. This contrasts with the Liberals with 3–4 expulsions, the Conservatives with 3–4, the Social Democrats with 2–3 and the Unity List with 2. See http://www.dr.dk/Nyheder/Politik/2006/10/06/0941.33.htm

18. This was further specified in the 2006 constitution.

19. Still, in 2000 the party had 6100 members, as many as the Socialist People’s Party and the Christian People’s Party (Pedersen 2006).

20. Internal documentation provided by the DF leadership group in 2009.

21. Of the 16 candidates born after 1970, 12 had a degree, implying that the education level of DF candidates is likely to increase (Art 2006: 23–4, 28).

22. Even Morten Messerschmidt, who became a DF MP at the age of 24, worked for the party for eight years before his election. He became a member in 1997 and built up a youth organisation in his home constituency before running for office and was elected in 2005 (Art 2011).

23. Courses organised by the party cover practical skills like making a budget, constructing a website, public speaking and forming a relationship with the local media (Art 2011: 155).

24. In contrast, more than three-quarters of LN voters were blue-collar or ordinary office workers, housewives or pensioners. Subsequent studies have shown that the LN electorate is the least educated of all Italian parties, with almost half failing to complete middle school (Ruzza and Fella 2009).

25. Corriere della Sera, 24 June 2010, Interview with Interior Minister Roberto Maroni.

26. See for a cross-national study of the long-term evolution of organisationally new parties in advanced democracies looking at the full range of ‘new party families’ (Bolleyer 2013).