Small parties: From party pledges to government policy

Abstract

When estimating a party's capacity for goal co-ordination, scholars need not only consider contextual constraints but also the party's properties, since these directly affect its strategic choices. For small parties which are crucial in virtually all multi-party systems the co-ordination of votes, office and policy is much more difficult than for numerically strong actors. Since the conceptual tools to assess the weight of small coalition partners – weight defined as the capacity to defend and realise core policy commitments – and to systematise intra-coalitional processes in general are absent, this article proposes a typology to account for small parties' weight. This typology is defined by the two criteria of ‘qualified pivotality’ and ‘centrality’, each of which is assumed to create a particular set of strategic advantages. Based on the latter, the approach allows small parties' impact to be compared, first, with reference to their positions within the respective parliamentary party system, and, second, with reference to the type of coalition that is likely to be formed. Based on the separate but parallel assessment of ‘formation weight’ and ‘coalition weight’, the typology reveals under which conditions the same properties of a small party may be advantageous during the coalition formation process, but disadvantageous during the subsequent period of coalition government.

Acknowledgements

This article is based on my MA thesis submitted at the University of Mannheim. My major thanks go to Peter Graf Kielmansegg, my teacher and supervisor during my time in Mannheim. Many other people helped to improve the manuscript: I am grateful to André Bächtiger, Evelyn Bytzek, Jonathan Hopkin, Matthias Lehnert, Otto Leirbukt, Peter Mair, Bernhard Miller and Diana Panke as well as the two anonymous reviewers. All remaining errors are mine.

Notes

1. Throughout this article, ‘party system’ refers to the parliamentary party system only.

2. The major divide in this strand is between policy-blind approaches that refer to size as the main determinant (Riker 1962) on the one hand, and policy-based approaches that refer to the ideological compatibility of political coalitions on the other (de Swaan 1973). Several efforts have been made to combine both factors. Still, one factor usually dominates the other.

3. Parties are understood as government parties if they hold ministerial posts. Accordingly, parties are understood as part of the opposition if they do not do so also when they support the government in the passage of legislative proposals.

4. Anti-system parties are an exception. Yet they represent only a small minority.

5. Either heterogeneous coalitions will not form or will be chronically unstable.

6. As an exception, Bartolini (1998) does so in one of his contributions to party system change. He points out the necessity to distinguish coalition potential from governmental power. When assessing governmental power he differentiates between government participation (yes/no), the coalition format and the intra-coalitional status of a party (leading, equal or junior position) and ranks parties' share of power according to these factors' respective configurations. Yet he does not systematically consider actor-specific differences in the capacity to translate coalition potential in governmental power. Instead, factors which intervene between coalition and governmental power are regarded as idiosyncratic and short-term elements (Bartolini 1998: 55).

7. The table only includes approaches in which ‘size’ plays a crucial role for power distribution. On the whole, theories that assume policy-oriented actors do not deal with internal power distribution because they predict the formation of an ideologically homogenous, hence stable coalitions which ex ante implicitly excludes the topic of intra-coalitional power distribution. However, there are exceptions, e.g. Budge and Keman (1990).

8. A redefinition of the concept of pivotality follows below.

9. Agrarians are an exception: they would not participate without the ministry of agriculture and fisheries (Budge and Keman 1990: 159).

10. Note that in order to include systems in which minority governments are common, the following categories do not presuppose any stable executive support in the legislature but can also refer to legislative coalitions. In democracies with majority cabinets, it suffices to specify the weight of actors with reference to exclusively executive coalitions, since stable legislative support is assumed to be a consequence; see, for a comparative analysis of party discipline, Sieberer (2006). In contrast, in minority parliamentarism, legislative support has to be built up anew each time when a draft is introduced, since legislative coalitions can be composed of different fractions from bill to bill (Bolleyer 2001: 1530–1). Accordingly, an actor does not need to be part of a stable government coalition in order to possess ‘coalition power’. Although the policy impact of a support party is less stable over time, hence less secure, it gains freedom from being blamed for unpopular government policies.

11. Most of the time government participation leads to electoral losses for the respective parties. An examination of coalition governments in 13 countries has shown that 62 per cent of the governing parties lost votes in the following elections (Müller and Strøm 2003).

12. An additional proxy could be the number of procedural rules for conflict management in the coalition treaty which is of particular relevance for small partners regarding issues that could not be settled during formation or that become contested in government although there has been agreement during formation. Empirical research indicates that coalition agreements have institutionalised as a part of government formation in most Western European countries with coalition governments. Only in countries with frequent minority cabinets are agreements less established. Albeit varying considerably in length, and regarding their content, policies form usually over 90 per cent of the commitments. Most crucially, due to time constraints party leaders tend to concentrate on issues that divide them. In Belgium and Dutch agreements, explicit commitments were frequent, where differences appeared easy to split, while implicit compromises were made when party principles such as moral issues were involved (Timmermans 2003: 13–14; 133–4).

13. For a detailed discussion of possibilities for operationalisation see Leirbukt (2004).

14. A direct change of the government coalition is only possible when there is no rule that demands that government dissolution be followed immediately by new elections.

15. Parties do not only strive for electoral success but also for the support of their members, who contribute to them financially and are eager to see party policy realised.

16. Accordingly, Warwick found that large parties tend to be assigned the formateur status since the closer a party is to majority status, the less need it has for partners (Warwick 1996: 474). Since small parties usually cannot profit from this advantage, the role of the formateur is not further discussed since the typology focuses on differences between types of small parties.

17. Empirical findings hint in the same direction: Studies on portfolio allocation suggest that small parties are proportionally ‘overpaid’, yet only if the number of coalition partners is small. That is, with an increasing number of coalition partners, the observed overpayment in ministries vanishes (Browne and Franklin 1973; Browne and Frendreis 1980).

18. Empirical findings show that big parties are not oriented towards minimising the seats of a future coalition as some coalition theories propose (Riker 1962). Otherwise small actors would need an identical seat share in order to count as equivalent.

19. Fragmented party systems with strongly bipolar dynamics can decrease complexity considerably since parties clearly associate with a bloc. Yet the pressure to take sides affects centrally located parties as well. Parties cannot exploit their middle position because it is constitutive for bipolar systems that parties need to associate with one of the two blocs.

20. Roozendaal emphasised the advantage of occupying a central position. At the same time, he criticised the concept of the central player because it also identified small parties as central, which were – according to him – too weak to represent ‘real centre parties’ (Roozendaal 1990: 331). Therefore he extended the concept by the criterion of numerical strength which excluded any small parties ex ante.

21. Again, a party can also be quite secure to enter government when having only one coalition option, as is the case with pivotal/non-central actors.

22. Another factor influencing internal dynamics is the degree of ministerial autonomy indicating the degree of collectivity (or segmentation) of cabinet decision-making (Thiébault 1993). This aspect indirectly affects the relevance of collegiality. One might argue that the more decisions are taken by individual ministers and are not taken into cabinet the less a small party needs to fear the bigger partners' dominance and the less important is the degree of collegiality. However, according to experts, a division of labour can be found with regard to routine decisions but barely exists on major policy issues (Strøm and Müller 1999b: 274–5).

23. This implies that small parties are not considered to be less responsible for the maintenance of the coalition than their big partners.

24. Only those parties are considered that held on average 2 per cent of the seats over the elections under consideration. Exceptions are made in cases where parties have entered government despite minimal size or punctual occurrence.

25. Since the works using a spatial conception of centrality refer to a one-dimensional policy space, small parties can be only indicated as central on the second dimension when occupying a median position as indicated by Müller, Strøm and their associates (Müller and Strøm 2003).