Income, growth, and democracy looking for the main causal directions in the nexus

The development of the political system of countries is noisy, but in the longer run a strong relation to the economy emerges in the cross-country data for income, growth, and the main democracy indices. Two main theories explain these relations: (α) starts from the strong correlation between income and democracy, seeing income as the causal variable. It is the democratic transition, which is the political part of the theory of the grand transition. (β) starts from the much weaker correlation between democracy and economic growth, seeing democracy as the causal variable. This is a part of the primacy-of-institutions theory, where the political system is a key institution. The discussion needs (λ) a link-relation between growth and income. It connects the (α) and (β) theories, so that one may explain the other. The analysis looks at all six possible univariate relations between the three variables using kernel regressions on a large, unified data set. This method gives a clear picture. The strong α-relation can indeed explain the weak β-relation as spurious, but the weak β-relation predicts that the α-relation is very weak. Thus, (α) encompasses (β), but not vice versa.