Another UK General Election: More Books!

The result of the 2005 UK general election, like that of its two predecessors, was a foregone conclusion to most commentators, and the main issue was the size of Labour's majority; all expected it to fall, but by how many? Unlike its two predecessors, however, the contest did not stimulate much media or general interest: the Labour government was well past its honeymoon period and the prime minister was no longer trusted by a large portion of the electorate, but the main opposition party had failed to attract much additional support. Most people and media pundits thought that New Labour should be re-elected, but with Blair getting a ‘bloody nose’.

That is what happened. Turnout was only slightly up on the 2001 all-time low, indicating no greater interest this time. The Conservative share of the vote hardly changed, whereas Labour lost about 13 percent of its 2001 share and the Liberal Democrats increased theirs by 20 percent – largely at Labour's expense. The decline in Labour's vote share meant that it lost 47 seats net in the House of Commons: the Conservatives' number of seats increased by 32, even though they won few extra votes, and the Liberal Democrats had a net gain of 10.

It was not earth shattering, therefore, but nevertheless British psephologists have written about it at least as extensively as they have after previous elections. Each recent contest has been the focus of six books from different stables, all offering their own take on the election. Four of those published in 2005 about that year's election are reviewed here. ¹

The series of Nuffield studies of British general elections extends back to 1945 – seventeen books covering every post-Second World War contest (plus a separate series on the European elections, also ‘fronted’ by David Butler). David Butler's name is on the front cover of all but the first two, but this is his final appearance and his co-author for the last ten volumes – Dennis Kavanagh – is first-named: the preface indicates that this is because he carried a larger share of the work than previously.

The British General Election of 2005 follows the format set by its predecessors. A review of the political scene since the 2001 election is followed by separate chapters on: the main parties during that time; the run-up to the election in the first quarter of 2005; the national campaign; the polls; broadcasting and press coverage; the candidates (this and the previous two chapters are by invited contributors); constituency campaigning; and an assessment of the campaign. There are also the usual appendices containing the basic statistics of the election (particularly the result in each constituency), and an analysis of the results. But the book handles and feels ‘lighter’ than its predecessors, with the three invited chapters very different from the remainder: Martin Harrison on broadcast coverage, Margaret Scammell and Martin Harrop on the press and Byron Criddle on the candidates present a wealth of empirical detail whereas most of what Kavanagh and Butler have to say is ‘instant history’, largely comprising informed judgements, a mode that has characterised their work over the last three decades: they mention interviews and people replying to questionnaires, but no details of their research methods are given.

A major feature of the Nuffield series has been the statistical analysis of the results, long undertaken by John Curtice and Michael Steed, and now joined by Stephen Fisher. These have been characterised by the clarity which their authors have brought to such high-speed analyses (given the time available for a book that appeared within six months of the election) and their ability to identify the salient features within the available material through straightforward analyses of ecological data. The current discussion begins with a crisp set of key questions (p. 236), proceeds through analyses of the contest between Labour and the Liberal Democrats in seats with large Muslim and student populations, an indication that the Conservatives were only able to tap into discontent with Labour (while yielding little ground to the Liberal Democrats) in parts of the south-east, discussions of the continuation of tactical voting, of the performance of minor parties, individual effects for candidates, and turnout, to an assessment of the importance of the electoral system in translating Labour's share of the votes into an incommensurate share of the seats, especially when compared with the Conservatives' situation. It is another masterpiece of analytical insight (though their conclusion that ‘the electoral system currently treats Labour considerably more favourably than it does the Conservatives’ [p. 252] suggests an exogenous relationship when the reality is that much of the current pro-Labour electoral bias results from the party's greater ability than the Conservatives to operate on and within the system and benefit from the potential that it offers for differential translation of votes into seats [Johnston et al., forthcoming]).

Tony King has been writing about British general elections for some 40 years, having co-authored the 1964 and 1966 Nuffield volumes with David Butler. Since 1974, with one exception, he has edited a volume of essays which has appeared within a year of the election – although with the 2005 version he, like David Butler, has taken an apparently lesser role, his Essex colleague John Bartle being the first-named editor.

Britain at the Polls 2005 is an intriguing book. Much of it follows the set pattern, leading off with good chapters on Tony Blair's second term (Thomas Quinn), the Conservative party (Philip Norton) and ‘A Restless Electorate’ (Nicholas Allen). But, because ‘very little changed’ within the Labour and Liberal Democrat parties between 2001 and 2005 ‘and what little did change had only a negligible impact on either the course of politics during those years or the eventual outcome of the election’ (p. vi), there are no chapters on them, which provided ‘an opportunity to examine more important long-term developments in British politics’. The selected ‘developments’ are very odd in the context of the election, however. David McKay's chapter on Europe as an issue in British politics, despite its intrinsic merits, focuses on something that he admits was of ‘low electoral salience’ (p. 92) in 2005 – an issue that seems to concern (some members of) the ‘chattering classes’ but not the electorate. Iain McLean's chapter on the evolution of devolution – ‘The Politics of Fractured Federalism’ – is even more distant from the book's overt subject, for he makes no attempt to connect his overview with elections to the House of Commons. ² E. J. Dionne's fascinating analysis of Blair's relationship with two US presidents (Clinton and Bush) is even more distant from the book's raison d'ĉtre, despite the last phrase in its title – ‘America's Other President: The American Left, Right and Centre on Tony Blair and the Election of 2005’ (and does not read to me – as the editors present it [p. vi] – as ‘a well-informed and acute North American perspective on the 2005 election’).

Other chapters focus on 5 May 2005, however. John Bartle presents a lengthy analysis of the Labour government's relationship with the media. And then Tony King addresses the key issue ‘Why Labour Won – Yet Again’. His qualities as an observer and analyst of British politics, set in the context of a full appreciation of recent research by political scientists, and the clarity of his writing mean that this chapter is the pick of the four books. In effect, if you want to know all that is worth knowing about the election, this is the one piece that you must read – it is a paradigm for us all in combining research depth and communication skills. Finally, Ivor Crewe looks forward to the prospects for each party at the next election, including a cogent seven-point memo that should be delivered to the Conservatives' new leader on his first day in office. (I wonder if it was, and whether he prepared another for the Lib Dems' new leader!) Like his long-time Essex colleague Tony King, he fuses research and interpretation in ways that others should seek to emulate, in a masterly overview that comes from the combination of research depth (including a wide appreciation of others' work), well-honed interpretative skills and clear prose.

Britain Decides is a relative newcomer, the third in a series of election books edited by Andrew Geddes and Jon Tonge. Presented as ‘the story of a rather dull campaign that produced some rather intriguing results’ (p. 4), it focuses on three main topics: the parties; local variations; and the key issues that dominated the inter-election period, if not the election itself. There are also chapters on feminising politics, campaign finance, the internet and the media. These are preceded by an analysis of the results from David Denver who concludes, in common with other commentators, that although each party was able to identify positive elements in the outcome, in effect none could really derive much satisfaction. This is confirmed by the chapters on the individual parties – Ed Fieldhouse and David Cutts arguing, for example, that despite their gains (based on relatively transient issues such as Iraq and student fees), for the Liberal Democrats the election was an opportunity foregone.

Local variation is treated in a somewhat unbalanced way, with separate chapters on Scotland, Wales and Northern Ireland, plus a brief chapter by a candidate (Tony Wright: Labour) in which he claims to have been so insulated in his own constituency (not watching any party election broadcasts, for example) that he had no idea how the national campaign was progressing! Nineteen one-page constituency vignettes apparently illustrate the ‘interplay of national and local factors’ (p. 5) – though nothing is done to integrate the parts into a larger whole. The country chapters are well done, however – especially that on Wales by Jonathan Bradbury, very often the poor relation in such treatments.

There are four chapters on issues – the economy, delivery of public services, foreign policy and the ‘war on terror’, and nationalism/immigration. These are excellent overviews which put the contest in context – including the exclusion of some of them from the inter-party debate, as in the lack of much challenge from the opposition parties to Labour's apparent unassailability on the economy (discussed by Martin Smith) – as well as being valuable discussions in their own right, for example David Richards' chapter on the delivery of public services.

Finally, Explaining Labour's Landslip is also the third in a series – although the first two were about landslides! It is produced by two members of MORI, one of Britain's longest-established polling organisations, and a marketing academic (Paul Baines) who was working there during the campaign. Not surprisingly, it draws very heavily on MORI's polling data, for a wide range of customers, but also deploys material from other pollsters where appropriate.

Although the book is structured into six chapters –‘Interpreting Public Opinion’, ‘The 2001–2005 Parliament’, ‘The 2005 Election Campaign’, ‘The Results’, ‘Engagement and Turnout’ and ‘What Next?’ – each is divided into a number of separate sections that in many cases are only weakly linked together. They read like a series of short essays on related topics rather than an integrated whole, perhaps necessarily so given the task of compiling the data and writing a book in such a short period. The chapter on ‘Engagement and Turnout’, for example, moves directly from a section on ‘The reputation of politicians and politics’ to one on ‘How and when children form political opinions’ without any linking narrative. MORI had done some polling for Nestlé in 2003 on the political opinions of secondary school students which ‘found some disturbing patterns’ (p. 287), and so this material was included. Interesting material in itself – and followed directly by a section on ‘Encouraging turnout with new voting methods’! In brief, the book is filled with interesting interpreted data, but these are not integrated in a way which leaves the reader other than rather overwhelmed by information.

Of course, books produced within weeks of a general election – all four of these (and the Norris/Wliezen volume) appeared within seven months of election day – cannot be expected to produce rounded appreciations based on detailed research: that will come later, using the British Election Study (BES) survey data (versions of which were released very soon after the election but are only deployed in a chapter by the principal investigators in the Norris/Wliezen volume [Whiteley et al., 2005]). Such ‘instant history’ books can only provide just that, and there are severe limits to what can be achieved in a few weeks of analysis and writing, which is why many of the chapters were clearly planned and written before the election.

So we have four books which, in different ways (to some extent: all have chapters on the media, for example, and not surprisingly they overlap considerably, even using the same data in some cases), tell us that Labour won, despite the loss of trust in Blair over Iraq and perceptions of failure to deliver meaningful public service reform. This was because of a strong economy, a main opposition party/leader which failed to win further support, a third party which made only limited gains in specific types of seat at the government's expense and an electoral system which Labour has so successfully engaged with over the last decade that only a massive haemorrhage of support would have produced defeat (unless it was very geographically concentrated in the ‘right’ marginal constituencies). People could show their dissatisfaction with Blair both by giving him a bloody nose and returning a government that had delivered economic prosperity for most. Or could they? Voting against the government for such a reason was a gamble because too many people aiming at his nose in the wrong constituencies could have reduced his majority much further – even made him (or a successor: would he have stayed on if this had been the case?) dependent on other parties to deliver his manifesto commitments. The gamblers – how many were there? where were they? were there potentially more? only survey data can tell us and they were not available in time to meet the editors‘/publishers' deadlines – had to believe what the commentators and the polls were telling them: Blair was going to win and the Conservatives could not. There are too many uncertainties, even in a system so stacked in Labour's favour, for even the most sophisticated voters to be exactly sure of what they are doing – though in the end it turned out almost exactly as perhaps they wished?

What can such books and their immediate analyses tell us about British electoral behaviour that has lasting value? In the past – before there were competitors for the Nuffield volumes – there was much more stability and predictability: the class cleavage was a strong foundation to how people voted and the electoral system was not likely to translate votes into seats in a way that massively advantaged one party over others. But all changed: the class cleavage became much weaker; fewer people could be expected both to identify strongly with one party and turn out to vote for it; more voters apparently made their minds up according to their perceptions of the party leaders and of the parties' ability to deliver on the salient issues; and the electoral system became less predictable in how it would translate votes into seats because of the changing geographies of who voted, and what, where. Given that, what do the three types of analysis deployed in these books tell us?

The first type is the informed judgement, to some extent sustained by polling data and statistical analyses of ecological data (almost all of them at the constituency scale), which characterises the Nuffield series and, to a large extent, the ‘Britain at the Polls’ series. These are examples of a long-established mode in British political studies, providing well-crafted interpretations of contemporary events set within the context of a well-developed research tradition with a stronger base in history than the US-dominated discipline of political science. And much of it is very good – as exemplified especially by Tony King's chapter. ³

The second type is ecological analysis, involving statistical manipulation of the data available immediately after an election (a mode that – apart from the appendix in the Kavanagh/Butler volume – is used by the contributors to the Norris/Wliezen volume rather than the majority of those appearing in the volumes reviewed here). This is where many of the current generation of British political scientists have cut their teeth, as exemplified by the statistical appendices to the Nuffield series and chapters such as Denver's in Britain Decides. Finally, and linked to the second, are the survey analyses, almost all of them based on the British Election Study surveys (the ‘experiment’ of a second survey was not repeated after 1983), with some additional material from polling organisations. These have become increasingly sophisticated (as illustrated by the first volume from the ‘new’ Essex team [Clarke et al., 2004]).

But there is a problem in linking the ecological and survey analyses. The former assume not only that we can infer individual behaviour from such data but also that the variables that can be included in such analyses are meaningful. Are they? Denver's analyses in Britain Decides include correlations with census variables which suggest that the traditional cleavage patterns are maintained, whereas much of the remainder of that book plus virtually everything in the others tell us that the key issues were trust in Tony Blair and voters' evaluations of the ability of the various parties to deliver on their key issues. These cannot be expressed as independent variables in constituency-scale analyses, so what are the latter telling us? – that opinions on Blair and on party performance/potential are related to socio-economic variables, in which case we face the endogeneity problem that Geoff Evans and Robert Andersen have been highlighting in recent years (see, for example, Evans and Andersen, 2004).

A partial response to this apparent paradox is provided by Ivor Crewe (in Britain at the Polls). He distinguishes between long-term structural and short-term campaign factors:

[t]he former consist of the enduring features of the political landscape, barely changing from one election to the next, which determine the underlying bedrock support for each party – their ‘normal’ vote. … The latter consist of powerful but short-lived influences on the vote which are limited to one election only but can supplement or erode a party's normal vote by a sufficient margin to make the critical difference to the result (p. 202),

In 2005 the Muslim anti-Iraq and student anti-fees, anti-Labour votes exemplify cases of the latter that can be identified in ecological and perhaps survey data (depending on sample size and geography [Johnston et al., 2006]). The latter can also slowly change the landscape. As Crewe notes, some campaign factors ‘may develop an enduring influence’: alternatively, having once shifted their allegiance some voters might find that they feel quite at home with a new party and instead of returning to their earlier affiliations/identities, stay there which, of course, is the hope of those who established the New Labour ‘project’.

How do we ‘test’ these ideas? Clearly not through one-off ecological and/or survey data analyses. What we need are longitudinal surveys which enable people's changing values, attitudes, opinions and behaviour to be tracked. As yet, we have little material to use for such research. The British Electoral Panel Studies have followed respondents through one electoral cycle (and provide the data used, for example, by Evans and Andersen [2004]) and the British Household Panel Study (now into its fifteenth wave) has some political data. But more is needed – which means a major funding commitment to a large survey over a decade or more. Without that, we can track short-term shifts, but cannot be certain whether longer-term trends – to the extent that they can be discerned from other data – are cohort effects or …? Perhaps British psephology needs to transfer some of its energy away from the short term towards more ambitious longer-term data collections and analyses (although what will that mean for the dreaded RAE?!).

What we currently have is a regular series of books providing instant interpretations of each general election result, with some statistical analyses, plus major studies based on the BES that necessarily appear some years after the election, and journal papers using BES and other data to expand the coverage. The instant history books necessarily have to appear quickly, otherwise the publishers probably will not be interested. Presumably there is a market, or they would not come back for more!

But is that enough? Only twice has there been an attempt to produce a more considered collective volume some time after an election, allowing BES and other survey data to be deployed, in a wider perspective (Evans and Norris, 1999; Heath et al., 1994). Perhaps it is time for another such venture? More importantly – and this is not the first time I have made such a point (Johnston, 1998; 2002) – is the knowledge about British elections sedimented in these volumes in any way providing us with a cumulative appreciation of how parties and voters behave? Textbooks about electoral behaviour are thin on the ground – the exception is David Denver's (2003) excellent brief introductory overview for students – and there are no monographs which draw together the material in any synthetic way (with the partial exception of Heath et al.'s (2001) overview of the series of elections for which they ran the BES and a forthcoming discussion of voting in Great Britain from the particular perspective of geographers [Johnston and Pattie, 2006]). Are we too concerned with separate election-by-election analyses, ⁴ or are there no general understandings to be drawn, or … is it time that some (at least one?!) student of British elections stood back and took stock? A lot has been written about British elections and voting behaviour over the last two decades: surely there is a larger story to be told than that which appears in books such as those reviewed here?

Or perhaps not; perhaps our elections are just ephemera – to be discussed straight afterwards and then forgotten. After all, Kavanagh and Butler end their concluding chapter (p. 202) by noting that:

In the eight weeks following polling day in 2005, political debate in Britain was dominated by various issues, the Government's scheme for ID cards, the uncertain future of the European Constitution and the Government's decision not to proceed with a referendum on it, and terrorism – none of which had received more than perfunctory mention in the campaign.