Return to: Search Page or to: Table of Contents Vol. 20, Issue 4

Dennis C Spies and André Kaiser, "Does the mode of candidate selection affect the representativeness of parties?" Party Politics, 20 (July, 2014), 576-590. [Available at ]

First paragraph:
The representation of voter preferences by political decision-makers is at the heart of studies on democratic political systems (Dahl, 1971; Downs, 1957; Iversen, 1994; McDonald and Budge, 2005; Miller and Stokes, 1963; Schofield and Sened, 2005; Stimson et al., 1995; Wlezien, 2004). While most of the literature empirically addressing the linkage between public preferences and political output focuses on the United States, it has recently been supplemented by a growing set of cross-national studies analysing the impact of different institutional arrangements on the degree of representation (Adams et al., 2004; Blais and Bodet, 2006; Ezrow and Adams, 2009; Ezrow et al., 2011; Golder and Stramski, 2010; Hobolt and Klemmensen, 2008; Soroka and Wlezien, 2004).

Figures and Tables:
Table 1. Bivariate correlations of different operationalizations of representation.
Figure 1. Cross-sectional and dynamic representation.
Table 2. Results for cross-sectional representation
Table 3. Results for dynamic representation.
Appendix: List of parties and modes for inclusion and centralization

Last Paragraph:
In contrast to cross-sectional representation, we find candidate-selection procedures have no influence on the degree of dynamic representation. However, our analysis of the relationship between the two concepts shows that both concepts are not only very weakly correlated but are also contradictory. Parties that are responsive to changing preferences of their voters over time represent their voters less well on election day. While this finding is logically puzzling, we think it is due to the way dynamic representation is operationalized, measuring the change in parties' and voters' positions between two consecutive elections. Parties might adjust their positions over a shorter or longer period of time, depending, for example, on their government participation or external pressures. Therefore, and while we acknowledge and support recent efforts to complement existing representation concepts with a time perspective, we question the dominant operationalization of dynamic representation. We conclude that the results of studies referring to either cross-sectional or dynamic representation are not comparable and suggest that scholars pay more attention to this.

Last updated June 2014