News

General election turns spotlight on voting behaviour and predictions research

14 Jun 17

General election turns spotlight on voting behaviour and predictions research

Nuffield’s role in advancing research and methodology in the field of psephology – the study and scientific analysis of elections – was recently made evident in the run-up to and wake of the 8th June general election.

The BBC produced a short explanatory video on the history of the “Swingometer” which made its first national appearance on British television in 1959. The Swingometer is a tool developed from the research of Sir David Butler (Nuffield Emeritus Fellow), which shows how movements of votes from one party to another can be used to predict the final outcomes of the election. Before David Butler, nobody had attempted to analyse elections in terms of percentages, and from there to spot trends that would help predict election outcomes before the final count.

The video is shown in this BBC News article 

Sir David Butler is often recognised for helping to establish the science of psephology, and many other Nuffield alumni and fellows continue his legacy. The College is one of three members of a consortium who manage the British Election Study (BES), which is one of the longest running election studies world-wide and the longest running social science survey in the UK. It has made a major contribution to the understanding of political attitudes and behaviour over nearly sixty years, since its first study conducted by David Butler and Donald Stokes in 1964.

Visit the British Election Studies site 

A number of Nuffield fellows and alumni have also made significant contributions to the development of a new design of exit poll, which has led to much greater accuracy in election result predictions. The team includes: former Nuffield fellows David Firth, Jouni Kuha and Stephen Fisher; Clive Payne and Neil Shephard, both Emeritus Fellows of Nuffield College; and John Curtice, a Nuffield College alumnus.

Read an explanation of the methodology and its record of achieving accurate prediction results.

In his blog, David Firth explains the significance of early accurate election predictions. “…most of the pre-election voting intention polls had predicted a substantial Conservative majority…But the exit poll prediction made it pretty clear that the Conservatives would either not achieve a majority (for which 326 seats would be needed), or at best would be returned with a very small majority such as the one they held before the election. Media commentary turned quickly to how a government might be formed in the seemingly likely event of a hung Parliament, and what the future might be for Mrs May. The financial markets moved quite substantially, too, in the moments after 10pm.” 

Understanding voting behaviour is a complex and fast-changing area of academic enquiry. Current models don’t account well for assessing the influence of new populist movements. Polls financed by businesses or specialist interest groups can undermine efforts to give fair and balanced predictions of election outcomes. Coalitions and progressive alliances are changing the game from measuring voter swing between a two-party choice and evaluating an array of personal and policy preferences. For the past fifty years, Nuffield College researchers have helped to develop and drive the science of psephology, improve the accuracy of polls and interpret changes in public opinion. There are definitely interesting times ahead for academics working in this field of research.

A selection of recent press coverage featuring election-related research, analysis and opinion from Nuffield fellows and alumni includes: