A method for measuring and decomposing electoral bias for the three-party case, illustrated by the British case

The measurement of bias in election results, whereby one or more parties are advantaged in the translation of votes into seats at the expense of others, is attracting increasing attention. So far, almost all of the analytical work - aimed at both identifying the extent of bias in an election result and establishing its causes - has focused on either two-party systems or on the largest two parties in multi-party systems. Building on the firm foundations of one such approach, this paper introduces an original procedure for analysing bias in three-party systems using a readily-appreciated metric for both evaluating the degree of bias and decomposing it into the various causal factors. This is illustrated using the example of the 2005 British general election and a comparison of the results from two-party and three-party analyses of six recent elections there.