Specific and Diffuse Support: Different Effects upon Changes in Voting Behavior?

As the importance of the influence of socio-cultural cleavages on voting decisions has waned and party identification decreased in many Western Democracies, explanations of voting behavior based on support for (or trust in) political actors and institutions have become more pronounced. Extant research found significant effects of indicators that can be classified among the concept of political support/trust on individual-level voting behavior (e.g. Dalton and Weldon 2005; Dejaeghere and Dassonneville 2016; Söderlund 2008; Zelle 1995). But not all findings point in the same direction. Other studies did not find significant effects of political trust (e.g. Dassonneville et al., 2015; Dassonneville, 2012; Walgrave et al., 2010). This paper argues that those mixed results can most likely be explained by the fact that those studies used different items of political trust with regard to the political objects towards which trust is directed. Starting from Easton’s (1975) distinction between specific and diffuse support I investigate how different aspects of political trust systematically impact vote shifting between parties and vote shifting into non-substantial voting choices (Don’t Know, Blank Voting). Easton’s classical argument would state that the more ‘specific’ types of trust - directed to particular officeholders based upon their performances - would have an independent and much stronger effect upon substantial voting decisions than levels of the more ‘diffuse’ types of trust. ‘Diffuse support’ – the type of support directed towards the more abstract institutions in a polity and towards the political regime itself - is interpreted as a non-performance based long-term attitude establishing political legitimacy (e.g., Dalton 2004: 23). Levels of diffuse support towards the political system are not believed to strongly affect levels of party-switching. Yet, diffuse support can be expected to be a good predictor of vote abstaining. In order to systematically test whether more specific and more diffuse types of support can be disentangled as independent causal forces with different effects upon different changes in voting behavior, I apply a panel dataset from the Netherlands that is currently collected around the 2017 parliamentary election campaign that covers 3 pre-election waves and one post-election wave. The data contains items indicating ‘specific’ (evaluations of party leaders, evaluations of party issue positions) as well as by the author developed items of ‘diffuse’ political support (satisfaction with the way democracy performs, fairness of the democratic process, efficiency of the democratic process, transparency of the democratic process). As voting behavior at the previous elections is also known, it is also investigated whether effects are different in the inter-election period and the campaign period as different incentives for stable and volatile vote intentions might be at work during both periods. Using the panel structure of the data, both the effects of differences in trust levels between persons as well as changes in trust levels within persons upon shifts in vote intentions and actual vote shifting will be analyzed. The effects will further be controlled for socio/demographic characteristics and basic political attitudes of respondents.