British Journal of Political Science

Research Article

Hanging on the Telephone? Doorstep and Telephone Canvassing at the British General Election of 1997

C. J.  PATTIE  a1 and R. J.  JOHNSTON  1 a2
a1 Department of Geography, University of Sheffield
a2 School of Geographical Sciences, University of Bristol


After years of neglect, a growing literature has reclaimed the constituency campaign as an important aspect of British elections. However, relatively little work has been done to disentangle which aspects of the local campaign are effective, and which are not. For much of the twentieth century, the mechanics of the local campaign were in essentials unchanged. But changing campaign technologies in the last decade offer new possibilities to party campaign managers. The 1997 British general election was the first in which parties made extensive use of telephone canvassing as well as the more traditional doorstep canvass. This article provides a comparative analysis of the effectiveness of traditional versus telephone constituency campaigns. Traditional face-to-face canvassing had a statistically significant influence on the outcome of the 1997 general election. But the telephone canvass did not.


1 The analyses reported here draw on the British Election Study's 1997 cross-section survey. The data used in this article were gathered by the Centre for Research into Elections and Social Trends (CREST) and were supplied to us by the ESRC Data Archive, University of Essex. We are grateful to both bodies. We are also indebted to Professor David Sanders and to two anonymous referees for their insightful comments on an earlier draft.