Explaining Cameron's Coalition [Hardcover]

Explaining Cameron's Coalition is the latest in the series of general election analysis by MORI's Robert Worcester and Roger Mortimore, this time joined by two other authors. The book is therefore very much the tale of the 2005-2010 Parliament and subsequent general election seen through the eyes of MORI's opinion polling, with an often pungent analysis which certainly fits Robert Worcester's happiness to point out when he got predictions right and others got them wrong.

Though there is a smattering of references to polling results from other firms, the great strength of the MORI data is that many of the questions have been asked regularly for decades, allowing the story of 2005-2010 to be put into a consistent historical context and polling results judged against previous ones that led up to victory or defeat. It also means that (as with Deborah Mattinson's excellent book, Talking to a Brick Wall: How New Labour stopped listening to the voter and why we need a new politics, based on focus groups rather than polls) it is an account of politics in which the views of the public dominate rather than the machinations and words of politicians, who usually take centre stage in post-election accounts.

The book is bulging with facts that make it hard to summarise them beyond "go read the book", though a few do particularly stand out. The authors conclude that "the nature of electoral support in Britain has changed, probably permanently ... the culmination of years of steady change ... British voters are ... less tribal ... and less polarised". Yet geographic division, especially the decline of the Conservative Party in Scotland, has hardened even as other divisions have softened.

Somewhat paradoxically the authors also very successfully model vote share in individual seats based on 17 different characteristics drawn from 2001 census. Factors such as the number of two or more car households are very influential in explaining the Conservative Party vote share, whilst factors such as the proportion of single-parent families do the same for Labour. Some factors do seem to divide even if the old patterns no longer have the same power.

In addition, "the old habit, whereby a predominant belief among voters that the economy was moving in the right direction was enough to ensure a government's re-election, no long holds. So, despite having convinced an extraordinarily high proportion of the public that the economy was on the upturn ... Gordon Brown could not muster the votes he needed".

The authors also point out that Gordon Brown's ratings as Prime Minister, whilst very low, followed a simple extrapolation of Tony Blair's declining figures though his time as Prime Minister. The problem wasn't that Brown worsened the long-term trend; it was far more that the decline had set in the moment Blair became Prime Minister and at election time Blair was up against unpopular Conservative Party leaders whilst Brown was up against David Cameron, far more popular than his three predecessors.

Indeed, the book points out that on their overall bundle of measures of leader image, Gordon Brown was in a slightly better position in May 2010 than Tony Blair had been even pre-Iraq was in April 2001. But William Hague was no David Cameron.

At times the authors skirt with over-playing the determinism of Labour's long-time decline in popularity during its term in office. After all, John Major - a Chancellor succeeding a three-times winning Prime Minister too - did pull off a slim victory against the odds. However, the authors do also point out that the final result was by such a fine margin (not many extra Labour seats would have transformed the possibilities of a non-Conservative coalition) that small events might have tipped the final outcome one way or the other.

As it was, Labour went down to defeat with for the first time in its history with, as the book points out, "more middle class voters than working class voters".

This book includes a useful introduction to how polls are conducted and how they are often misreported, with the warning to "think of polls as being like a barometer - barometers don't predict the weather; they measure something that is helpful to know if you want to predict the weather. But for that purpose, rather than relying purely on voting intentions the many other measurements that the polls regularly provide may be far more useful in developing an impression of what the future may bring". Wise words from a good book that ends with a very welcome appendix - a survey of the political cartoon during the 2010 election, an often overlooked form of commentary.