Political institutions and economic development over more than a century.

• This paper investigates the relation between democracy and income per capita over the period 1870–2010. • I use both cross-sectional and panel fixed effects regressions to explore how this relation changed over time. • Results show that democracy positively affects income per capita in the period post-1950, especially after 1990. • Yet, democracy has no significant effect in the period prior to 1950. • In the years before World War II, income per capita is mostly explained by prior income levels and education. Economists and political scientists have long investigated the effect of political institutions on economic development, mainly focusing on cross-sectional analyses for the years after World War II. This paper takes a historical perspective and studies whether this effect can be traced back to 1870 and how it changed over time. Using both cross-sectional and panel fixed effects regressions, I show that democracy positively affects income per capita in the post-1950 period. The effect increases as time passes and becomes especially larger after 1990, i.e. with the great democratization wave induced by the Fall of the Iron Curtain. Yet, I find no effect for the period before 1950, when the variation in income per capita is mostly explained by prior income levels, and in particular, by education. These findings are robust to accounting for income dynamics as well as to using different data sources to measure political institutions. [ABSTRACT FROM AUTHOR]