The Impact of Political Advertising in the 2001 U.K. General Election

Article excerpt

This article explores the extent to which advocacy and attack Party Election Broadcasts (PEBs) affected voters' party preferences during the British general election campaign of 2001. The analysis uses an experimental design that involved conducting "media exposure" tests on a representative sample of Greater London voters (N = 919) during the final weeks of the June 2001 election campaign. Respondents completed a pre-test questionnaire before being exposed to a variety of different media stimuli. Their political attitudes were then measured again in a post-test questionnaire. The empirical findings suggest that, in general, PEBs exerted little direct effect on voters' images of the main political parties in 2001. However, there were a series of "partial" exposure effects confined to particular sub-groups of voters. For example, for non-partisan voters, "attack" advertising appears to have been less effective than "advocacy" advertising. Indeed, in the U.K. in 2001 there were contexts in which negative campaigning was explicitly counter-productive in the sense that it appears to have actively stimulated sympathy for the target of the attack rather than strengthened the relative position of the sponsor.

In recent years scholars and practitioners have turned increasing attention towards understanding the impact of political advertising. One aspect of the debate has revolved around the issue of how far there are significant electoral rewards from either "advocacy" broadcasts, which offer a positive vision of the advertised party, or "attack" broadcasts which concentrate on criticising the opposition.

In this article, we explore the extent to which advocacy and attack Party Election Broadcasts (PEBs) affected voters' party preferences during the British general election campaign of 2001. Our analysis uses an experimental design that involved conducting media exposure tests on a representative sample of Greater London voters (N = 919) during the final weeks of the June 2001 election campaign. Respondents completed a pre-test questionnaire before being exposed to a variety of different media stimuli. Their political attitudes were then measured again in a post-test questionnaire. Our empirical findings suggest that, in general, PEBs exerted little direct effect on voters' images of the main political parties in 2001. However, there were a series of partial exposure effects confined to particular sub-groups of voters. For example, for non-partisan voters, attack advertising appears to have been less effective than advocacy advertising. Indeed, in the U.K. in 2001 there were contexts in which negative campaigning was explicitly counter-productive in the sense that it appears to have actively stimulated sympathy for the target of the attack rather than strengthening the relative position of the sponsor. Part one of this article outlines the theoretical debates and rationales that state the specific hypotheses that we test. Part two describes the experimental design that we used in order to generate the data to test these hypotheses, together with our operational measures. Part three reports our empirical findings.

THE THEORETICAL AND EMPIRICAL BACKGROUND

An extensive literature has sought to assess the effectiveness of television-based campaign advertising in American elections (Pfau and Kinski 1990; Ansolabehere and Iyengar, 1995; Lau and Sigelman, 2000; West 2001; Thurber 2000; Lau and Pomper 2002). In Britain, many studies have described the evolution and character of campaign communications (see, for example, Scammell 1995; Seymore-Ure 1996; Butler and Kavanagh 2001) and the impact of news media coverage upon electoral behavior (Miller 1991; Norris et al. 1999). A smaller body of work has focused on trends in the format and contents of party election broadcasts (see Scammell and Semetko 1995; Harrison 2001). In particular, content analysis by Hodess, Tedesco, and Kaid (2000) noted a tendency towards increased negativity evident in PEB aired during the 1997 campaign, compared with 1992. …