Return to: Search Page or to: Table of Contents Vol. 20, Issue 2

R Michael Alvarez, Ines Levin, Peter Mair, and Alexander Trechsel, "Party preferences in the digital age: The impact of voting advice applications," Party Politics, 20 (March 2014), 227-236. [Available at ]

First paragraph:
Since the 1970s, a growing body of literature in political science has addressed the 'crisis of democracy' (see, e.g., Crozier et al., 1975; Dalton, 2004; Pharr et al., 2000). The reasons given for the 'crisis' are manifold, though most refer to declining levels of trust among citizens, erosion of turnout, disaffection from party membership, decreasing levels of party identification, low levels of party cohesion, the end of party government, the rise of guardian institutions, etc. A number of studies point to changes on the political supply side, in particular the changing nature, role and importance of political parties. Generally, or so it is argued in this growing body of literature, there has been a weakening of traditional cleavages, with policy positions of parties and candidates becoming more opaque, parties compete in a complex and multidimensional issue space, and there is party system fragmentation and electoral volatility. All of this, in turn, makes decision-making difficult for many voters; the usual issues and cleavages that might have guided their decisions in the past are not necessarily relevant or useful in contemporary politics. This has many potential consequences of most concern; it is possible that the growing levels of disaffection from electoral politics seen in many European nations may stem from these trends (Torcal and Montero, 2006).

Figures and Tables:
Figure 1. Sequencing of the EU Profiler.
Table 1. Congruence between PTV-t1 and best matching party in the EU Profiler.
Figure 2. Relating congruence to the number of parties.
Figure 3. Distribution of representative deficit.
Figure 4. Multi-level logit model, first-level coefficients, average country.
Figure 5. Switching probability and representative deficit.
Table 2. Baseline choice probabilities and marginal effects, average country.

Last Paragraph:
These data are by no means representative of the electorate in Europe. A strong process of self-selection precedes the use of any VAA, and one could argue that it is precisely those voters who are lost in partisan terms that would be overrepresented among the users. At the same time, our results reveal the presence of context-dependent variation, which shows that even if self-selection takes place it produces different levels of congruence and somewhat relativizes the very dark picture of the gradual disconnect between parties and voters. It goes without saying that this bivariate model still needs more careful specifications; it was included here for the sake of understanding of the country-level variation that had to be taken into account when trying to explain the effects of the EU Profiler on party preferences in t2. Clearly, much more research, methodologically and substantively, is necessary in order for social scientists to better understand the important new phenomenon of VAAs and how they affected the preferences and behaviour of European voters.

Last updated March 2014