Validity of the Meltzer and Richard hypothesis under captured democracy and policy regime hypotheses

The well-known Meltzer and Richard (1981) hypothesis proposes that median voters are more likely to support redistribution policies that reduce income inequality when the gap between pre-tax income and mean income increases. Despite the academic influence this hypothesis has had, empirical studies demonstrate that the relationships between democracy, redistribution, and inequality are far more complex than Meltzer and Richard assume. Based on previous theoretical research, this study proposes two hypotheses to explain why the equalization effects of democracy may not work: the captured democracy hypothesis and the policy regime hypothesis. Empirical testing with data for 164 countries between 1990 and 2018 suggests the following results. First, the degree of redistribution of democratic countries is much higher than that of nondemocratic countries. However, the variability of income redistribution in democratic countries is higher than that of non-democratic countries. Second, the estimation results do not support the Meltzer and Richard hypothesis for any of three democracy-related indices. This implies that democracy itself is not a sufficient precondition for redistribution. Third, the empirical tests of the two hypotheses on democratic countries support that the equalizing effects of democracy are weaker in societies where political influence depends on socioeconomic position and in societies where neoliberal policies are dominant.