Knowledge of why people voted the way they do depended on anecdotes by party workers until the first comprehensive surveys began after the Second World War.

Lazarsfeld and his colleagues from Columbia University studied voting in local areas during the presidential election campaigns of the 1940s expecting to find that the election issues and media coverage during the campaign affected voters’ choices.

To their surprise, the campaign hardly made any difference and they concluded that voters chose who to vote for on the basis of long standing loyalties based on class, ethnicity and the social networks that people had. People frequently fitted their understanding of what the candidates were saying to their pre-existing view of who they supported.

A team at Michigan University studied the Presidential elections of the 1950s and concluded that most voters developed party loyalties at an early age and, although these could change, they only did so after some period of time.

Short term factors such as the attraction of a candidate or the national mood on the economy, could lead to a short term switch to the other party but then voters reverted back to their previous loyalty; detailed policy arguments had little effect.

This theory of partisan attachment based on social characteristics became the dominant theory of voting in the 1960s and early 1970s.

It underlies Lipset and Rokkan’s model of party systems and the main study of British voting behaviour, which was by David Butler and Donald Stokes , Political Change in Britain, 1969 (both Rokkan and Stokes spent some time at Michigan).

Previous Next