Brexit and Two Roles for Referendums in the United Kingdom

Abstract

The 2016 referendum on the United Kingdom’s membership of the European Union is a textbook example of how not to conduct a referendum. Further, the Brexit referendum was not an isolated democratic failing. It reflects a deeper structural weakness in the use of referendums in the United Kingdom. For decades in the United Kingdom before Brexit, referendums were used in an ‘unplanned’ and ad hoc way when it was politically convenient. This paper will argue that the unprincipled use of referendums in the United Kingdom, at least in part, reflects an underlying disagreement about what referendums are for and their relationship with democracy. There are, broadly, two different accounts of the relationships between referendums and democracy in the United Kingdom. On the first view which I will call the ‘referendums as limits’ approach, referendums are a way for voters to act as a check on the actions of their representatives. This is the view taken by Dicey, Qvortrup, and Bogdanor. The second approach is to see referendums as a way of temporarily substituting voters for representatives. On this view, drawing on the civic republican tradition, referendums are deliberative processes of expressing popular sovereignty. I will call this latter view ‘referendums as substitutions.’ This latter view of referendums is presented in its most effective form by Tierney. It is also the view of referendums that is increasingly reflected in law in the United Kingdom. I will argue here that the ‘referendums as limits’ approach is the better view both in theory and in practise. The practical argument will focus in particular on the 2016 Brexit referendum. I will show that there were numerous failings with the Brexit referendum, but many came back to the same root: the attempt to use the referendum as a substitute for representatives rather than limits on the action of representatives. This attempt to use referendums to replace rather than guide representatives led to failings of process, content, execution, and accountability. I will show that I do not think these failings would have occurred if referendums been used as limits on the actions of representatives to approve the Maastricht and Lisbon treaties, or if the 2016 Brexit referendum was clearly connected to new legislation or proposals for reform.