Erik Moberg ã:
A Theory of Democratic Politics


In the preceding parts I have said that the constellation of main actors is dependent on the number of political parties, and their discipline and cohesion, in particular as manifested by the voting patterns in the legislature. Accordingly it is important to consider mechanisms which shape parties and party systems in these respects. In particular mechanisms with a constitutional basis, if there are any, are relevant in this context. Here, in part 6, I shall present six hypotheses about such mechanisms - the first five are related to the constitutional traits identified in the preceding part 5. Then, in the subparts 6.1-6.6, I will discuss the hypotheses in more detail. This discussion will however be confined to the hypotheses' logical foundations, and their places in the logical structure, or theory, developed here. The issue about the hypotheses' empirical truth will, on the whole, not be dealt with.

This hypothesis is often referred to as Duverger's law. The French political scientist Maurice Duverger was certainly not the first one to entertain the idea, but he was the first to give it a sharp formulation and, simultaneously, to maintain its status as a scientifically valid generalisation, and he also collected and systematically arranged a lot of empirical information in order to prove its truth (Riker, 1986, p 26). The hypothesis has been, and continues to be, extensively and explicitly discussed in political science, and it remains controversial. These three last hypotheses, as we see, deal with the discipline of political parties - and, again, it is in particular the discipline in legislative voting which is at issue. These hypotheses are not at all discussed in the same systematic way as Duverger's law. Rather, there are just occasional references, in passing as it were, to the hypotheses or similar ideas. Furthermore the distinction made here between means and incentives is never, to my knowledge, done explicitly. Sometimes the distinction is not even made implicitly and the idea expressed is rather that parliamentarism leads to discipline, or something like that. A few examples may illustrate this.

Lipset, for instance, quoting a paper of his own from 1976, talks about "the tight national party discipline imposed by a parliamentary as compared with a presidential system" (1990, p 199). Some authors also hold that the cohesion of the parties in a parliamentary system is, indeed, so strong that the parties can be treated as individual, or unitary, actors. In a discussion mainly devoted to other matters Hinich & Munger (1994, p 134) thus suddenly claim that "The same model holds for parliamentary systems, where parties, rather than individual candidates, are the main actors in elections." Similarly Laver & Shepsle (1994), although they admit (p 309) that they "have not gone into the mechanisms of party discipline" emphasize (p 301) "the role of the parliamentary-party machine in enforcing party discipline, and hence in enhancing the party to function as a single monolithic actor."

But even if most of the authors, who talk about the matter at all, thus seem to agree that parliamentarism enhances party discipline, there are also exceptions. Sartori, for instance, bluntly states (1994, p 95, his italics) that "... party solidification and discipline (in parliamentary voting) has never been a feedback of parliamentary government. If a system is assembly-based, atomized, unruly, magmatic, on its own intertia it will remain as it is. I cannot think of any party system that has evolved into a veritable ‘system’ made of strong, organization-based mass parties on the basis of internal parliamentary learning."

I do take side with those who claim that parliamentarism encourages discipline. In order to get deeper into the relevant mechanisms I have however found it productive to make a distinction between the incentives for discipline, and the means for enforcing it. Within this perspective discipline does not come about unless both incentives are present, and means available. The incentives are the subject matter of the second hypothesis, and the means are dealt with in the third and fourth hypotheses.

The two last hypotheses, number five and six, deal, without making the distinction between incentives and means, with mechanisms impeding party discipline.