Social networks and experienced inequality

Abstract

Traditional measures of inequality, such as the Gini coefficient, involve pairwise comparisons across all members of a given population. But, most people possess information about, and therefore experience inequality in comparison to, only a subset of the population. In this paper, we provide simple axioms to describe inequality as experienced in social networks. We propose an index to measure aggregate experienced inequality that is consistent with these axioms. We then compute the Gini coefficient and ‘experienced inequality’ in 75 villages in Karnataka, India. We show that for a given wealth distribution, the social network could either accentuate or diminish experienced inequality. We show, first analytically and then empirically, how this can happen with respect to two network properties. Firstly, wealth-based homophily is negatively associated with experienced inequality. Secondly, caste-based homophily is negatively associated with experienced inequality when within-caste inequality is less than the overall Gini coefficient (and positively associated when the opposite is true).