Fendalton: A By-Election Opinion Survey

Abstract

By-elections have always presented a particular challenge to the psephologist. The motivations of the non-voters are normally difficult to interpret and as the poll goes down so do the dimensions of the problem increase. Crude voting figures tempt the researcher to assume that change largely takes the form of a greater or lesser fall in the percentage shares of the enrolment of the different parties competing. The scale of its apparent losses to abstention therefore measures the standing of a party. Such thinking is perhaps a convenient form of journalistic shorthand. Not only does it largely ignore the role of the ‘changers’; it also assumes, explicitly or by inference, some kind of rational process, almost a decision to abstain, on the part of non-voters. In both these respects by-election commentary tends to ignore the findings of opinion surveys at general elections. Clearly there is a need to go behind the voting figures of a by-election and to assess the results in the light of the actual processes of change as measured by some form of survey.