‘But I Thought: Politicians’ Beliefs on Income Distribution and Their Effects on Redistributive Preferences

About this Session


Wed. 10.04. 14:45



Authors – Nathanael Gratias Sumaktoyo , Christian Breunig , Wolfgang Gaissmaier

Abstract :

Income and wealth inequality within and across countries is a widely recognized and substantial problem, with many undesired consequences. Understanding why individuals support or oppose redistributive policies is a normatively important and practically relevant research endeavor. Studies consequently have documented various factors that shape redistributive preferences. We contribute to this literature in three respects.

First, we add to the nascent but growing number of studies that examine redistributive preferences of political elites. Second, we measure the relative influence of two antecedents and how they are related: ideological predispositions and perceptions about societal conditions. Third, we operationalize beliefs about societal conditions in two ways: beliefs about societal affluence and societal inequality. This improves existing studies that have focused solely on misperceptions of affluence.

We interviewed 535 members of parliaments (MPs) in eight German states. We asked the MPs to estimate the percentage of the population that fell into six monthly household income brackets: (1) less than €1,000, (2) between €1,001 and €2,000, (3) between €2,001 and €3,000, (4) between €3,001 and €4,000, (5) between €4,001 and €5,000, and (6) more than €5,001. From this provided distribution, we calculated for each MP their estimated mean population income (perception of affluence) and Gini coefficient (perception of inequality).

We find that conservative MPs tend to predict higher population mean income and lower income inequality than liberal MPs. When it comes to how these beliefs relate to the MPs’ redistributive preferences, our path modeling reveals two significant patterns. First, beliefs about income inequality are a more consistent and powerful predictor of redistributive preferences than beliefs about the level of income. Second, the estimated effects of perceptions about societal conditions are not reducible to ideological differences. That is, they significantly predict redistributive preferences, even after ideological predispositions are accounted for.

These results underline the importance of not only understanding politicians’ ideological beliefs, but also their factual beliefs. And in terms of those factual beliefs, the perceived level of overall income (how affluent the society is) seems to matter less for redistributive preferences compared to how the income is perceived to be distributed (how unequal the society is). The distributive dimension therefore is of high practical relevance, as factual beliefs likely are more responsive to information and thus can be changed, whereas this is very difficult for ideological beliefs. More specifically, providing information about societal conditions and correcting existing misperceptions might influence politicians’ stances on redistribution.