Academic journal article Political Research Quarterly

Measuring Issue Salience in British Elections: Competing Interpretations of "Most Important Issue"

Academic journal article Political Research Quarterly

Measuring Issue Salience in British Elections: Competing Interpretations of "Most Important Issue"

Article excerpt

This article is about responses to the "most important issue" question used in numerous election polls and surveys. Following Wlezien's work, two interpretations of the question can be sketched: (1) personal (the issue most important to the respondent) and (2) contextual (the issue that respondents perceive as topping the national political agenda). Using British Election Study data from 2005, the author shows that issues prominent in that campaign were often cited as most important by respondents who were neither particularly knowledgeable about those issues nor particularly influenced by them when voting. In sum, the contextual interpretation predominates. Hence, whatever else it is, "most important issue" is not an accurate gauge of salience effects in models of vote choice.

Keywords: issue salience; British elections; issue voting; valence politics

In this article, I subject the "most important issue" (MII) question used in the 2005 British Election Study (BES) to detailed theoretical and empirical scrutiny. The wording of the question is, "As far as you're concerned, what is the single most important issue facing the country at the present time?" Parallel questions are posed in almost all national election studies and countless opinion polls. Drawing on Wlezien's (2005) dissection of such questions, I sketch two competing interpretations of answers: first, as a respondent's report of the issue most important to him or her; second, as a respondent's perception of the issue most important in the country. These I term the personal and contextual interpretations, respectively. The two need not imply different answers, since personal and national importance are obviously related. But when they diverge, respondents must choose whether to report what matters most to them, regardless of whether it matters most nationally, or vice versa. The central concern in this article is, which do they choose? (Or, as interpretations may vary across respondents, which do more of them choose?) Wlezien's analysis was based on the dynamics of aggregate responses, but here I bring individual-level data to bear on the problem.

First, the two interpretations of the MII question are explained in more detail, and I assess their relative plausibility given the question wording. I then consider what existing evidence can (and cannot) tell us about how respondents interpret it. The next stage is to specify empirical criteria by which to adjudicate between the two interpretations. (This indirect approach is necessary since, at least in the absence of cognitive interview protocols, we cannot find out directly how respondents interpret any survey question.) After describing the data and methods used, I present an array of evidence that confirms that interpretations vary, but weighs in favor of the contextual interpretation. Finally, I assess the implications of the results, and suggest that this important survey question is not fit for the purposes for which it is often used, and echoing Wlezien's (2005, 577) call for research into alternative formulations.

1. "Most Important Issue"-Personal Versus Contextual Interpretations

The influence of issues on party choice is a matter of long-standing debate in British voting studies (Butler and Stokes 1969, chap. 8-9; Särlvik and Crewe 1983; Heath, Jowell, and Curtice 1985; Clarke et al. 2004), but few contemporary researchers would exclude issues altogether from their models of party choice. Moreover, most now take explicit account of the importance, or salience, of issues (Hutchings 2003; Whiteley et al. 2005; Green 2007). The basic premise was stated by Särlvik and Crewe (1983, 222): "One would expect a stronger relationship between issue opinions and voting among those who consider a given issue important than among those who are less concerned." Importance here signifies the weight assigned to an issue in a voter's decision calculus (Wlezien 2005, 557-59). Of course, survey respondents may be asked about the importance of issues between elections, when there is no "decision" to be made as such, and so their responses are a more abstract indication of subjective salience. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.