Fifty years ago, Lipset (1959) linked the capitalist economic development with the expansion of democracy. Using the concept of regimes of accumulation (Boyer, 2007) this paper attempts to offer a different perspective on this relationship. First, here I argue that only some regimes of capitalist accumulation were associated with democratic stability in the period 1945-2001; and only because of their special patterns of income distribution, and their stable rates of GDP growth. Further, the effects of income distribution on democratic stability are evaluated here in new ways. First, because the Median Voter Theorem is not used here (Acemoglu and Robinson, 2006; Boix, 2003). And second, because instead of using the Gini index as an independent variable, here I constructed a measure of income polarization (IED) that predict democratic stability more efficiently than the Gini index.