Wouter van der Brug, "Voters' Perceptions and Party Dynamics," Party Politics, 5 (April 1999), 147-169.

First Paragraph:
In contemporary democracies elections are the prime instruments of political representation. The notion of representation implies that public policies should reflect the preferences of a majority of the citizens of a country. Given the indirect way in which elections link voters' preferences to public policies, various scholars have struggled with determining under which conditions meaningful representation can exist. Modern democratic theory -- in particular the responsible party model -- attributes a key role to political parties in electoral processes (APSA, 1950; Luttbeg, 1968; Thomassen, 1994). The responsible party model defines a number of conditions under which policies will reflect mass opinions. This article concerns one of these conditions: that voters should have accurate perceptions of parties' policy positions. It assesses the extent to which this is the case in the Netherlands and whether it has changed during the last 3 decades. For this purpose, voters' perceptions of parties' left-right positions will be compared with parties' 'actual' policy positions. The reason this study focusses on left-right perceptions follows from recent findings of electoral research.

Figures and Tables:
Table 1: Overview of the election years included in each of the 10 analyses
Table 2: Overview of the comparisons between party programs and left-right perceptions
Figure 1: Plot of party manifestos in 1952, 1956 and 1959
Table 3: Multiple regression of issue categories on coordinates in two-dimensional MDS-solutions. Normalized regression coefficients and explained variance
Table 4: Multiple regression of perceived party positions in terms of 'left-right' over coordinates of two-dimesional MDS-solutions. Normalized regression coefficients and explained variance
Figure 2: Plot of party manifestos in 1967, 1971, 1972 and 1977, and the direction of the left-right dimension
Figure 3: Plot of party manifestos in 1971, 1972, 1977 and 1981, and the direction of the left-right dimension
Figure 4: Plot of party manifestos in 1977, 1981 and 1986, and the direction of the left-right dimension
Figure 5: Plot of party manifestos in 1981, 1986 and 1989, and the direction of the left-right dimension
Figure 6: Plot of party manifestos in 1976, 1989 and 1994, and the direction of the left-right dimension
Appendix: Distance, dimensionality of MDS solutions and vectors.

Last Paragraph:
This being said, it must also be noted that the most recent developments observed in this paper provide reason to suspect that in the near future Dutch voters may experience more difficulty in forming adequate perceptions of party policies. The left-right dimension has been, and still is, an important frame of reference for Dutch politicians and voters (e.g. Van der Eijk, 1995; Tillie, 1995). However, in the eyes of the electorate parties have been converging on the left-right dimension since 1986 (Van der Brug, 1997), while the linkage between left-right perceptions and the contents of party programs has been progressively weakening since the 1970s. Moreover, the analyses of party manifestos suggest that important changes are taking place in the contents of the ideological dimensions that structure party conflict in the Netherlands. These developments suggest a decreasing prominence of left-right issues for structuring party behavior. Inadequate perceptions are likely to develop when a party system realigns, so that a traditional dimension of conflict becomes obsolete while voters are not yet fully acquainted with newer dimensions of conflict. The fact that Dutch voters did hold accurate perceptions in the past does not therefore imply that this will invariably be the case in the future.