AMERICAN POLITICS RESEARCH / NOVEMBER 2001Tolbert et al. / BALLOT INITIATIVES ON VOTER TURNOUT THE EFFECTS OF BALLOT INITIATIVES ON VOTER TURNOUT IN THE AMERICAN STATES

With few exceptions, voter turnout continues to decline in the United States. Although normative theorists, journalists, and defenders of participatory democracy frequently suggest that citizen-initiated ballot measures can increase voter turnout, previous research has not supported this claim. Yet, in the past 25 years, usage of direct democracy has exploded in the United States. Using pooled time series data for the 50 states over a 26-year period (1970-1996), we find that the presence and usage of the initiative process is associated with higher voter turnout in both presidential and midterm elections. The disparity in turnout rates between initiative and noninitiative states has been increasing over time, estimated at 7 % to 9 % higher in midterm and 3 % to 4.5 % higher in presidential elections in the 1990s. Our analysis suggests that the initiative process can and does play a positive role in increasing electoral participation. As the number and importance of citizen-initiated ballot measures continue to grow in the American states, scholars are increasingly turning their attention to this particular brand of plebiscitory democ-racy. There has been a growing use of initiatives and referenda in the United States and European countries to decide policy issues (Bowler, Donovan, & Tolbert, 1998; Butler & Ranney, 1994). In some Ameri-can states, such as California, direct democracy has become a preferred mechanism of governing as the state’s most important policies— taxes, spending, education, health, environment, affirmative action, immigration, welfare—are now decided via initiatives and referenda (Schrag, 1998). As we shift into an information-technology age, worldwide trends highlight the increased use of referenda and more participatory models of governing (Mendolsohn & Parkin, 2001