The Utility Outcomes of Referenda

My research will focus on the utility outcomes of referenda. By developing a citizen- candidate model, I will examine whether a binary vote on a single issue (for example, territorial independence, the death penalty or EU membership) appeals to individuals' basest instincts, and results in overall utility that is lower than a decision taken by a representative government. The logic for this statement is as follows. The citizen-candidate model predicts that in a general election, candidates converge to the median to maximise chances of winning an election (the median voter theorem). Across a broad spectrum of issues, therefore, moderation is rewarded and overall utility maximised. In a referendum, however, the binary nature of the vote discourages (or at worst negates) moderation. In effect, the spectrum can be truncated to shift the true median to one end and a polarised view to another, resulting in an outcome more extreme than the true median. Alternatively, the spectrum can be segmented, excluding the true median from voter choice, offering only more extreme positions in a binary vote.
Research is required to ascertain whether the use of referenda thereby results in significant decreases in overall utility, as opposed to executive action by a representative government. Several methods of testing are possible. Examining referenda that were later defied or overturned by executive action alongside socioeconomic indicators could inform the relationship between public preferences and the two different voting mechanisms. This could be complemented by laboratory experiments eliciting individual preferences across a spectrum of issues as opposed to focused \votes". Such questions have numerous implications for the balance of power between government action (\executive action") and direct elicitation of public preferences (\referendum"). How each branch is used to inform policy greatly impacts short- and long-term overall utility.