Benny Geys, "District Magnitude, Social Heterogeneity and Local Party System Fragmentation," Party Politics, 12 (March, 2006), 281-297.

First Paragraph:
The question about what shapes the party system has attracted a great deal of scholarly attention. Most basically, two approaches have been followed to answer it. The 'institutional' approach emphasizes the role of electoral laws (cf. Duverger, 1954), while the 'sociological' approach highlights the effect of socio-economic differences or 'cleavages' in the population (cf. Lipset and Rokkan, 1967). These elements have generally been considered to be more or less isolated from one another. However, recently, a number of scholars have put forward the view that both elements may well interact with one another (Amorim-Neto and Cox, 1997; Ordeshook and Shvetsova, 1994). The argument is that 'a polity will have many parties only if it both has many cleavages and has a permissive enough electoral system' (Amorim- Neto and Cox, 1997: 155, emphasis in original). Cross-country evidence provided in these studies is generally supportive of this interactive model.

Figures and Tables:
Table 1. Summary statistics: average values (standard deviation)
Table 2. Determinants of the number of parties - Lijphart approach
Table 3. Determinants of the number of parties - Rae approach
Appendix A. Determinants of the 'effective' number of parties

Last Paragraph:
This article addresses these issues with Belgian municipality-level data from municipal elections over the period 1982-2000. We have included four measures of socio-economic diversity and tested for independent and interaction effects with electoral laws (i.e. district magnitude). The main conclusion is that even in a sample consisting only of moderately large multimember districts the interaction effects significantly add to the explanatory power of the model. This conclusion is not affected by our handling of the data or the exact operationalization of the dependent variable. The interpretation is that the influence of socio-economic heterogeneity on the number of electoral parties is significantly affected by the district magnitude, and vice versa. Further research should investigate whether this difference from Ordeshook and Shvetsova's (1994) finding is explained by the use of constituency-level data (e.g. by testing the model with national election results at the district level, as in Vander Weyden, 2004).