Contemporary Democracies Revisited

Abstract

Few have questioned how democracies as a group differ within and among themselves. The most important study in this area of inquiry is Powell's (1982) Contemporary Democracies. Unfortunately, some of his results may be both inefficient and biased due to the use of what we now understand to be an inappropriate method. This study applies more appropriate event count models to Powell's data in hopes of gaining new insights into the relationship between political violence and elements of democracy. Evidence to support Collective Action explanations of political violence was found. Strong support for the argument that presidencies can be detrimental to the state and that representational electoral systems and constitutions, especially consociational constitutions, outperform majoritarian systems was also supported. Environmental factors are important, but constitutional variables, discounted to some degree by Powell, were also found to have extremely important and significant effects on the degrees of violence in democracies.