In recent years, critics have argued that the California attorney general’s role in formulating ballot language opens the door to biased titles and summaries for statewide propositions. This paper examines the debate over the propriety of the attorney general’s role, using the largely unexamined case study of Proposition 58 (2016) to gain insight about the attorney general’s influence. In November 2016, voters in California – many in counties that supported Donald Trump – approved Prop. 58 with 73.5 percent of the vote. This proposition removed many restrictions on bilingual education programs in public schools. Eighteen years earlier, 61 percent of Californians had approved an initiative to restrict bilingual education in the state. This paper uses a combination of qualitative and quantitative analysis to examine potential reasons for this swing in voting behavior on bilingual education policy, concluding that the attorney general’s ballot language most compellingly explains the bulk of this shift. As such, the paper suggests adding an extra check on the attorney general’s office during the ballot drafting process to ensure that Californians receive clear, impartial information on election day.

What role does political knowledge play in campaigning for and participation in direct democracy? A foundational principle of democracy is citizen participation in decision-making. This foundation assumes that citizens are at least somewhat knowledgeable about government and able to make informed choices. This analysis examines the role that meaningful decisions play in direct democracy, because “for voters to make meaningful decisions, they must understand the options on which they are deciding” (Dalton 1988: 13). This analysis uses three different methodologies to investigate this relationship. First, through qualitative analysis and a mail survey of petitioners, this study explores how petitioners view and approach the public. This study finds that expectations of political knowledge affects how petitioners approach the public and how much time they spend educating the public about their initiative. Second, through statistical (multi-level regression) analysis, this study investigates the impact of the ballot language on participation in individual ballot propositions. This study finds that ballot language is a significant barrier to participation. Third, through experimental analysis, this study connects measures of political knowledge and participation on ballot propositions written by petitioners across the country. This study finds that when confronted with more difficult ballot language voters are less likely to participate. However, when controlling for political knowledge this effect is truncated. The findings of this analysis argue the elite bias of direct democracy in ballot language, accessibility, and motives of petitioners. The study of participation in direct democracy and political knowledge across American states advances the theoretical understanding of democratic participation, and furthers our understanding of the role citizen political knowledge plays in policymaking.