Direct Democracy and Government Spending

There is a long list of contributions arguing that direct democratic institutions go together with lower government spending. While this finding is hardly contested for the late 20th and early 21st century, it might not have always been like this. Most notably, Matsusaka has shown that before WWII, cities in the US with direct democratic elements tended to have higher spending. We examine this "Matsusaka paradox" by differentiating between the effects of direct democratic institutions allowing to block legislation and institutions enabling actors outside the legislative and executive to propose new laws. In addition, we analyse how the representative system, in which direct democracy is embedded, shapes these instruments' effect on state spending. Based on almost 200 years of fiscal and political data from Swiss cantons, we exploit the different moments of introduction as well as the varying degree to which these institutions existed to examine the effect of direct democracy on government spending. We show that the distinction between blocking and enabling institutions in interaction with the representative systems helps to account for the Matsusaka paradox.