During recent years, a number of studies have argued that variation in human capital endowment across European countries has important political implications for electoral and democratic reforms adopted during the first decades of the 20th century. The existing tests of these competing explanations have generated inconclusive results. One important shortcoming of existing studies is the absence of systematic data measuring the core explanatory concept proposed by these theories, namely variation in 'co-specific investments in skills'. In this paper, I take advantage of a new dataset that measures the ratio of skilled to unskilled workers in 200 occupations across German localities to examine the effect of differences in levels of human capital for electoral reforms adopted by subnational parliaments in Germany. The analysis combines a panel analysis of the adoption of electoral reforms with an analysis of the economic and political determinants of individual support of electoral reforms in several German regions.
