Inequality, transaction costs and voter turnout: evidence from Canadian provinces and Indian states

This article tests for the presence of a regularity in the relationship between income inequality and voter turnout in two countries with common Westminster parliamentary systems. We begin by using data from a panel of Canadian provinces to assess two contrasting monotonic hypotheses: conflict theory that predicts a positive monotonic relationship (inequality promoting conflict and greater electoral participation) against relative power theory that predicts a negative monotonic relationship (inequality leading to political alienation and electoral disengagement). Nesting these hypotheses within a rational choice model of voter turnout, we find that neither hypothesis explains the data convincingly while a search across fractional polynomials finds that the relationship is better described as non-monotonic with an inverted U shape. The generality of this finding is assessed by rerunning the analysis on a panel of 14 Indian states. The commonality of results across countries with similar political structures but widely different demographics and stages of development is striking and consistent with the hypothesis that conflict theory operates at low levels of income inequality before growing inequality leads to voter alienation and lower voter turnout. In the Canadian case the tipping point arises at an income Gini of 0.32 while the Indian case peaks at consumption Gini of 0.34.