Partisan Campaigning and Initiative Petition Signing in Direct Democracies

Abstract

This paper investigates whether popular initiatives signed by a larger share of the population have higher acceptance rates at the subsequent vote. The main analysis is based on all Swiss federal initiatives voted between 1978 and 2000 with a panel of aggregate voting data at cantonal level. The results suggest that petition signing is positively and significantly related to acceptance rates at ballot. I address potential omitted variable bias from underlying preferences which might be driving both signatures and acceptance rates in three ways. First, the panel structure of the data allows to control for time-constant preferences via fixed effects. Second, results are robust to various proxies for voter preferences. Third, using the doubling of the signature requirement in 1978 as an instrumental variable confirms the above result. The findings imply that petition signing can serve as an effective partisan campaigning tool.