The post below is an example of something I’ve been toying with: providing ‘sample assignments’ for my undergraduate students. In this case, the assignment was to create a 1,000 word blog entry on the study of elections and voters in the UK, including some contemporary analysis focusing on 2015. I’ve graded myself (a nice luxury) at an 80 for this level 2 undergraduate assignment.
It was a little nerve-wracking in some senses, as the pressure was on to create a good assignment. But I think doing this, sending it to my students and letting them critique it was a very valuable learning experience.
Anyway, I hope that some of you find this idea interesting, and that you also enjoy the blog, which is a (hopefully) readable summary of a large and evolving field of enquiry.
In this post, I will provide a potted history of trends that have characterised the study of voting behaviour in the United Kingdom from the early 1950s through to the present day. This account is limited to work that has sought to explain vote choice in British general elections using surveys of voters.
I sketch a transition from socio-economic and socio-psychological, identity-based accounts of voting, through a period of debate about whether or not voters had become ‘dealigned’ from parties, to contemporary theories and models of vote choice that emphasize elite performance, perceived competence and leadership. I will conclude with a discussion of how scholars will approach the challenge of explaining vote choice in the upcoming 2015 general election.
Curtice notes that the first studies of British voters employing survey methods focused on one, or, at most, two constituencies. An appreciation of the importance of social structure, particularly social class, is discernible in all of these early studies. Voters’ party preferences appeared to be principally attributable to socio-economic background, leading to the establishment of a consensus around the ‘two-class, two party’ model of British politics. The image below, for instance, is taken Benney, Gray and Pear’s study of Greenwich in the 1950 election campaign, and indicates that socio-economic status was a strong predictor of vote choice.
Figure 1. Excerpt from Benney et al. ‘How People Vote’
The consensus of early studies around the centrality of class to British electoral behaviour was such that Peter Pulzer was able to state in 1967 that ‘class is the basis of British politics; all else is embellishment and detail’. Of course, this model did not paint a complete picture of British voting behaviour (if it did, Labour would have won every election), and some scholars sought to analyse ‘deviant’ behaviour, focusing particularly on working class Conservatism.
The United States was the venue for the first ‘modern’ academic studies of electoral behaviour using regional and eventually national surveys. These studies are described in detail elsewhere, but is important to note here that theories and methods developed to analyse American electoral behaviour have been vital to academic understanding of the behaviour of British voters. The work of scholars at the University of Michigan had a particularly profound impact. Later, the work of American academics Anthony Downs and Morris Fiorina would also prove influential in the UK, refining British scholars’ understandings of ideological and performance-based vote choice respectively.
Donald Stokes, one of the members of the Michigan research group, teamed up with Oxford’s David Butler to inaugurate the ‘British Election Study’ survey series and to co-author the seminal work on British electoral behaviour in the 1960s – Political Change in Britain. In this book, based on a series of national surveys covering the 1964, 1966, and 1970 elections, the core concept of the ‘Michigan model’: party identification, was presented as the key determinant of vote choice in the UK.
The authors argued that partisan self-image tended to be transmitted from parents to their children during ‘the impressionable years’ and that partisan self-image was relatively stable over time; reinforced by the existence of homogenous class communities and the class basis of British politics.
From the early 1970s onwards, however, this approach was undermined by growing British electoral volatility. Those examining the British Election Studies noted both a waning of the strength of social class as a predictor of vote choice (evidenced through declining scores on the Alford Index) as well as a decline in the number of party identifiers and strength of party identification in the electorate. Analysing these patterns, Särlvik and Crewe concluded that the 1970s had been a Decade of Dealignment.
While this thesis was hotly contested into the mid 1990s, it is now broadly acknowledged that class based voting and party identification were weakened, particularly from the mid 1990s onwards, when Tony Blair repositioned Labour and courted the middle class vote. In this context, the intellectual ground was fertile for new ideas.
In 2004 Clarke, Sanders, Stewart and Whitely’s Political Choice in Britain provided such an idea, arguing that the key calculation made by voters at the polls was based on ‘their perceptions of the likely competence of competing parties’ managerial teams’. Voters use a combination of their evaluations of how well-equipped parties are to ‘handle’ key issues alongside their impressions of the party leader in forming these perceptions. This approach, known as the ‘valence’ model dominates contemporary academic account of British voting behaviour.
I conclude this overview noting that the concept of party identification plays a key role in both socio-psychological and performance/leadership based accounts of British voter behaviour. The central point of disagreement between these accounts lies in the nature of partisanship – is it deep seated, emotional, and stable (as the socio-psychological account argues) or is it a shifting indicator of voters’ evaluation of parties’ managerial competence(as per the valence account)?
As Denver, Carman and Johns note, ‘one of the pleasures – and, occasionally, frustrations – of studying elections is that they keep happening’. The 2015 election cycle, coming on the back of the first full coalition government since 1945, presents several fascinating puzzles. Looking forward is less natural to the academic than looking back, and the discussion here is more political opinion than academic summary.
Firstly, the planned 2015 British Election Study promises some exciting innovations, by merging several data sets it will allow analysts to do more to separate the influence of social environment from their political evaluations as well as offering daily tracking of key variables as the campaign unfolds.
I predict that the valence framework will continue to be prevalent, with all major parties striving to convince on the competence of their leadership teams, although the clarity of these signals from the government parties may be blurred somewhat by the coalition experience. However, I anticipate that socio-economic factors will also matter, whether they are emphasised by the parties or not. The regional distribution of voting looks set to continue along the ‘North-South’ divide that we have seen in recent years. Finally, the importance of ‘events’ can never be discounted and crises, scandal, or unexpected good fortune will all have their part to play.