Representation by Government, Political System Perceptions and Social Capital

While there exists a flourishing literature on the extent to which citizens are well-represented by their governments and what causes congruence and responsivity of the government, comparatively less focus has been placed on the consequences of being represented on individual citizens. Studies in this field have primarily analyzed how citizens’ satisfaction with democracy is influenced by how well they are represented by their government, either in the form of having voted for one of the government parties in the preceding election or through a low distance in their ideological position respective to the mean ideological position of the government. Given the government’s key influence on the policy-making process, citizens who are concerned with policy outcomes should be more satisfied with the functioning of democracy if their political views are well represented by the government. Most of these studies, however, rely on cross-sectional data or at most on two-wave panels surveying respondents directly before and after an election. While two-wave panels improve over cross-sectional data by assessing intra-individual changes and thus controlling for individual-level confounders, they still do not allow for distinguishing whether the effect is temporary or lasting. In this study, I aim to improve on existing studies by making use of the Dutch Longitudinal Internet Studies for the Social sciences (LISS) panel, a panel survey based on a true probability sample of households collected annually since 2007. In this time span, there has been satisfactory variance in the ideology of Dutch governments, ranging from centre-left, right-wing, grand coalition as well as centre-right governments. In order to measure how well individuals are represented by their government, I employ three different measures: a) intending to vote for one of the government parties if an election were held, b) average sympathy with the parties in government and c) average policy distance in four policy areas (income redistribution, multiculturalism, moral policy and European integration) using Chapel Hill Expert Survey data for the position of government parties. Using fixed effects regression models, I show that being well represented by the government has wide-ranging and lasting positive effects not only on the satisfaction with democracy and political trust, but also on an individual’s external political efficacy as well as their social capital in the form of voluntary engagement and generalized social trust. Overall, my results contribute to extant literature in three ways: First, employing panel data allows for more robust inference than existing cross-sectional surveys. Second, I use three different, independent measures to capture how well citizens are represented by their government and find very similar results across all three measures, showing that the results are not contingent on specific measures of representation. Third, I go beyond existing studies by showing that being well represented by the government does not only affect the satisfaction with democracy, which most studies in this field have looked at, but has wide-ranging effects on political trust, external political efficacy and social capital as well.