ABSTRACT
This chapter considers how often citizens engage in online research about ballot measures before elections and whether it actually motivates any real-world behavior. To examine whether voters engage in Internet research for ballot questions we use 2012 American National Election Study (ANES) data, and we use aggregate data on real-life Google searches. To investigate potentiality on ballot measures, we have three hypotheses. Hypothesis 1 is that the proportion of Internet searches on a state ballot measure will increase within the state as the election becomes closer. Hypothesis 2 is that the proportion of Internet searches for the topic of the ballot measure will increase at a greater rate as the election approaches in the state where the ballot measure is relevant. Hypothesis 3 is that that the more Internet searches there were for a ballot measure in the week before the election, the less likely voters were to roll-off.
