Normative evaluations of popular vote processes, such as referendums and initiatives, often pay attention to specific aspects of institutional design, such as whether a popular vote process is binding or advisory. While it is necessary to analyze these specific features, I contend that we need to do so in the context of the broader democratic system in which popular vote processes are situated because these specific design features may have distinct consequences depending on the context. I contend that taking this systemic view also makes it possible to understand the circumstances in which democratic innovations could address democratic deficits that are commonly associated with particular types of ballot measures.