The idea that ‘local’ candidates enjoy an electoral advantage has been firmly established in both the theory and practice of British general elections. It has become de rigueur for candidates in UK general elections to emphasize the depth of their constituency links in electoral communications, and several empirical studies appear to provide support for a ‘localism’ effect. However, while voters state that they prefer to vote for a ‘local’ candidate, there are many incidence of candidates ‘parachuted in’ winning representation, including in marginal seats. Is the ‘localism effect’ perhaps a myth?
In this paper, we seek to test the ‘localism’ thesis with new data drawn from the 2015 General Election candidates survey. Unlike previous studies, we include the full range of candidates of all parties and test whether an independent localism effect, on both vote share and winning a seat, can still be identified once we control for a range of other factors that may influence the electoral performance of individual candidates. Moreover, we rely not only on the ‘objective’ measurement of ‘localism’ defined on the basis of home addresses of candidates but on data measuring the perception of whether a candidate truly is a ‘local’ candidate. If there is a localism effect, we hypothesise that most of that effect on electoral performance can be explained by the perception that a candidate is local, instead of their actual domicile.
