Conclusions
Lipset’s renowned quotation (1959, 75) suggests a gradual democratization
process associated with greater socio-economic development. Thus, the empirical
strategy of panel data with fixed effects, which assesses the within-country
variation between relative wealth or education and democracy, captures econometrically
the spirit of Lipset’s hypothesis.
Applying fixed-effects OLS and Arellano-Bond methods to our post-World
War II data set, using five-year high-frequency panels and conditioning on a proxy
for capitalism, we obtain results qualitatively similar to those of AJRY to the extent
that increasing both income and education do not induce greater political
development. However, to account for weak instruments and endogeneity bias,
we use a System GMM estimator advanced by Blundell and Bond, and we find
that income and education predict democracy in five-year panels conditioning on
economic freedom. Intuitively, as countries become relatively wealthier and their
populations relatively more educated, the likelihood of these countries becoming
relatively more democratic increases. We also find using OLS that democracy
changes over a 35 year period are predicted by income and human capital in 1975.
Thus, using our sample and the frequently used method of OLS, we corroborate