As various statistical tables can show, the poorest countries in the world are dictatorships.
All the famines have happened under autocratic rule (Dreze and Sen 1989). Neverthe-
less, the relationship still can go every which way. Second, they found that growth rates
are either very low or very high in dictatorships and in between in democracies. This
observed large variance in economic performance casts doubt on the effect of dictator-
ship on economic development. Dictatorial regimes can exhibit economic miracles and
disasters.6 For instance, since the economic reforms of 1978, China’s GDP per capita in
constant US dollars has doubled every 8-9 years. Finally, their empirical results reveal
that dictatorships are likely to democratize in middle-income countries and remain stable
in low- and high-income countries, thus invalidating endogenous democratization.7
Despite their inconclusive results, these empirical studies offer two preliminary findings.
First, the economic dimension is relevant for the analysis of political regimes. Second,
economic welfare is not neutral for the stability and instability of political regimes.
The present paper attempts to account for the large variance of economic performance
under dictatorship and its political stability. This will depend ultimately on the distance
between the preferences of the ruled in terms of economic welfare and social order, and the
effect of the dictator’s own preferences on these two variables. We think that economic
welfare and social order are the contemporary relevant factors of legitimacy and, hence,
stability of political regimes. By social order it is meant security of persons and their
property. As mentioned before, the economic factor is relatively recent and backed by
6See also examples in Temple (1999) (table 2).
7Boix and Stokes (2003) challenge this econometric result.