Income Inequality and Policy Responsiveness Perceptions - A Genetic Matching Approach

Previous research on policy responsiveness points to income inequality as one of the most consistent predictors of policy responsiveness suggesting that policy outcomes strongly reflect the preferences of the “rich” (Bartels 2002, Gilens 2005, Hayes 2013). However, the link between income inequality and policy responsiveness is not fully understood. Firstly, some suggest that the lower income citizens often participate in politics at much lower rates than others (Verba, Schlozman, and Brady 1995). Secondly, studies also show that there are gaps in the political knowledge and interest between rich and poor (Delli Carpini and Keeter 1996). Thirdly, it is also not understood if these differences translate themselves at the individual level in terms of perceptions of how the government listens. This paper provides an attempt to disentangle the effects of participation rates, political interest, ideological preferences, and other related variables from the effect of income on perceptions of policy responsiveness. Additionally, the measure of policy responsiveness chosen is subjective using citizens' satisfaction on how policy shifts follow their will. Thus, the question that this paper addresses is whether people in different economic situations consider that policies are more responsive to them? The first part of this paper responds to some of the criticism raised against survey-based measures of responsiveness. The second part presents results on income inequality’s effect on perceptions of policy responsiveness using data collected in Hungary in 2012 by the ESS. The paper uses a matching method that helps in modeling observational data to look closer to one that would result from a randomized experiment. The effect of income inequality is isolated from various confounding factors such as participation rates, interest in politics, and left-right ideological positioning. If before matching, t-test results suggest that self-perceived high income earners assess the responsiveness of governments to their opinions of policy differently than the rest of the sample, after matching such an effect is not supported. Moreover, this null result remains fairly unchanged even after various robustness tests using both different measures of income earning and different thresholds for identifying the “rich”. On the one hand, one interpretation would consider these results as going against the theoretical expectation that economic inequalities translate themselves at the representational level. On the other hand, since the measure of policy responsiveness used in this paper is subjective, we argue that a different interpretation of these null findings should be given. Even if economic inequalities produce inequality at the representational level in terms of policy responsiveness, this inequality is not perceived at the individual level. While policy might indeed follow more closely the opinion of the “rich”, this bias is not perceived by the two groups when controlling for political and demographic variables. Additionally, one could also argue that even if there is an association between income and policy responsiveness, it is not perceptions of income per se that drive differences in perceptions of policy responsiveness, but rather political and demographic variables associated with it.