Voting Behaviour in General Elections vs. Referenda: An Analysis of the UK

Due to unexpected outcomes of recent UK elections, factors that determine how voters behave in different election types are increasingly relevant to political science. It is important for referenda to be studied hand-in-hand... [ view full abstract ]

Due to unexpected outcomes of recent UK elections, factors that determine how voters behave in different election types are increasingly relevant to political science. It is important for referenda to be studied hand-in-hand with general elections to deepen political knowledge of what influences the electorate most in each election type, however, the literature rarely explicitly addresses both election types. 

In this project, therefore, I ask the question "Is voting behaviour in general elections determined by the same factors as when voting in referenda? (i.e. Do the same variables impact voting outcomes in each election type?)" with respect to the 2015 UK General Election and the 2016 EU Referendum. I predict: (1) outcomes of each election type will be affected by similar demographic variables like gender, income and ethnicity, (2) different factors will be the most significant in determining how individuals vote in each type of election: party identification for general elections and standpoint on immigration for referenda and (3) quality and quantity of information will not have a significant impact on voting in either election type. Using logistic regression analyses, my data suggest: (1) a weak relationship between voting behaviour in each election type, (2) party identification is an important indicator of general election vote whilst age does not predict referendum vote, but opinions on immigration and the economy do (3) quality of information is important to both election types, and (4) education level and opinion on party leaders have a significant influence on voting behaviour in both election types.