Abstract.
This study shows how redistribution takes place as an outcome of the optimizing behaviour of dictators in non-democratic countries. Tinpots are clearly different than totalitarians in this respect, and non-democracies differ from democracies. A robust finding is that a more autocratic totalitarian dictator redistributes more than a less autocratic one, the opposite in tinpots where the expenditure share of redistribution increases consistently with freedom. The results place an interesting restriction on the sign of the coefficient estimate of freedom in equations explaining redistribution in the types of regimes. It is negative in the regression for totalitarians but positive in tinpots (and democracy).This indicates that political freedom depresses redistribution at low level of freedom but enhances it when a moderate level of freedom is already attained. In other words, redistribution is nonlinearly related to political freedom with a U-shaped pattern.This pattern is little sensitive to the choice of freedom index (Polity IV or Gastil) in regime classification.
This study reveals a feedback relationship between income inequality and redistribution, which differs by regime types. Income inequality induces a decrease in redistribution in totalitarians but an increase in tinpots that in turn leads to a rise in inequality in the former and a fall in the latter. Thus income inequality first increases and later decreases during the process of democratization. In other words, income inequality is related to freedom with an inverted U-shaped pattern. The results further show that inter-country differences in political institutions, costs of government services, socio-demographic characteristics of the people and their culture can lead to variation in the share of redistribution across regimes, without affecting its pattern of relationship with freedom.