Lewis Baston is author of Reggie: The Life of Reginald Maudling and several books about British general elections. He is a consultant on politics, elections and constituencies.

For either main party to win outright in May, or even get into a position where an agreement with the (presumably depleted) Liberal Democrats would be the basis for a workable government, national voting intentions have to shift quite dramatically away from the level-pegging state of affairs that has been the broad pattern for the last few months – a Labour lead of five points, or a Conservative lead of 11 points, would probably be those at which we could take talk of overall majorities seriously.

They would also, after a parliament in which most movements in opinion have been slow have suddenly to become rapid. So, are we about to see the trench warfare of the battle for public opinion become a war of movement?

By many standards, the 2010 campaign was a pretty exciting one. It was the first to feature a series of televised, head-to-head debates between the leaders of the three main parties. Nick Clegg, given a platform of equality with Gordon Brown and David Cameron, used it skilfully and came across as reasonable, natural and sincere while the other two seemed tired and over-rehearsed.

The first televised debate transformed the campaign, with a surge in ‘Cleggmania’ and some polls showing the Liberal Democrats on the verge of breaking through and polling a record share of the vote. The dynamics of the election shifted, with the Conservatives and the press trying to beat back the Lib Dem surge. Labour seemed to flirt with collapse, with Gordon Brown providing another key moment of the campaign with his offhand description of a ‘bigoted woman’ in Rochdale.

The ever-accelerating spin cycle of the campaign, the sheer tonnage of opinion poll results, and the growing role of social media and internet campaigning made it a fast-moving and somewhat nauseous occasion, like being on a circus waltzer after having had too much candyfloss.

But when the votes were counted at the end, the result was pretty much what one would have expected had there been no campaign at all. Turnout was up a bit, but there was no surge. The Conservatives were in the lead by the same seven points over Labour they had been in the polls at the start of March, and the Liberal Democrats were in third place with their vote slightly up and their seats slightly down.

The result, a hung parliament, had looked likely since the Conservatives’ lead had faded at the turn of the year as the economy started to pick up. Two months of hectic campaigning, millions of pounds spent…and everyone might just as well have done something better with their time and money.

It is impossible to tell for sure whether campaigns matter, because we cannot run an experiment holding everything else constant and either going straight to election day, or running a campaign but having one or more actors in it making difference decisions.

It seems unlikely that a period of intense political campaigning and coverage in the media, directed at a general public who have not been paying politics much interest and towards a quarter of whom say they make up their mind at the last minute, has no real effect. It may be that there is a counterfactual version of 2010 where things did shift significantly: Andy Cooke has written one, The Fourth Lectern, in which UKIP gets into the televised debates and there is a different outcome to the election.

There is also the possibility that had the Conservatives (and their press allies) and Labour tried to ignore the Lib Dems, or mismanaged the counterattack even worse, that Cleggmania would have ended up reflected in the results. It is conceivable that a better Conservative campaign in 2005 could have deprived Labour of a working, or even overall, majority: Data from the 2005 British Election Study (BES) Rolling Campaign Panel Survey show that Labour were fortunate that the Conservatives performed particularly badly and, in effect, lost the 2005 campaign. (Clarke et al, 2006)

Most election campaigns produce results that are more or less what a reader of opinion polls and other electoral data might have anticipated a few months in advance, even though expectations have sometimes ebbed and flowed in the meantime, as in 1955, 1959, 1966, 1979, 1987, 1997 and 2001.

Even 1970 and February 1974, which both came as shock results given the expectations at the end of the campaigns, were pretty much in line with what the polls said in November 1973 and March 1970. Apparent changes in opinion during the campaigns in 1959 and 1987, like those in 2010, turned out not to be reflected in the result.

Elections in which the result is significantly different from expectations just before the start of the campaign are the exceptions, and usually the difference is not so much in the winning party as secondary features of the election such as the size of the majority and the performance of third parties. Labour had been expected to win October 1974 (see my piece on this site about that event), 1950 and – a few months before the poll – 1964, rather more comfortably than it did, and the same is true of the Conservatives in 1951.

Majorities were unexpectedly large in 1935, 1931 and 1924. In 1983, a big Conservative win was accurately anticipated, but the surprise was the collapse in Labour’s lead over the SDP/ Liberal Alliance from 15-20 points at the start of the campaign to two points in the final result.

Labour’s 1945 landslide was unexpected, but accurately predicted by the polls and the wartime by-elections; although there were Conservative missteps during the campaign itself, it seems extremely unlikely that they had a significant effect on the result, or for that matter whether a perfect Conservative campaign could have done much to erode Labour’s advantage.

And, because of the experience of 1992, few people could quite believe the suggestion that in 1997 Labour was indeed on course for the landslide majority consistently implied in the polls.

In thinking about election campaigns and how much they matter, one does come back to 1992. Much of the reason that result came as a surprise was because the opinion polls, which set expectations and the climate of the campaign, were systematically wrong. Rather than being level, the Conservatives were probably always somewhat in the lead, and benefited additionally from some degree of late swing at the end of the campaign.

However, there is no single thing that one can point to in the 1992 campaign which one can describe as being the cause of what late swing there was: Labour’s response to the Budget? John Major’s soapbox campaign and plucky underdog status? Labour’s overconfidence? Chris Patten and Shaun Woodward’s inelegant and negative but undoubtedly functional and effective campaign? Was it The Sun wot won it? I doubt whether the formula is replicable.

One can also make the case that, even though the 1970 and February 1974 elections confirmed the verdict implied before the campaign, the campaigns themselves did matter because, in each case, they overturned a swing back towards the government that seemed to be taking place in the pre-campaign period. The Tories in 1970, and Labour in 1974, adroitly turned the campaign agenda away from what the Government had intended at the start (complacency in the first case, mild panic in the second) and on to issues where they were stronger, in each case with the help of Enoch Powell who was working to his own agenda.

It is the lot of campaigns hailed as innovative to fail to change much. Labour’s brilliant media campaigns in 1959 and 1987 caused poll wobbles during the campaigns, but produced the same results one would have expected at the outset. Occasionally the side that was winning anyway gets to write the history: the Saatchi & Saatchi campaign for the Conservatives in 1979 was a breakthrough in advertising but, arguably, Labour’s campaign was more effective. Baldwin’s ‘Safety First’ advertising innovations in 1929 did not prevent his government from being on the losing side of the election. The televised debates, and the greater prominence of the internet, did not produce an unexpected result in 2010. The same is very likely to be the case whatever innovations, alarums and excursions take place between now and 7 May.

Further reading

Dennis Kavanagh and Philip Cowley The British General Election of 2010 (Palgrave, 2011)

Harold Clarke, David Sanders, Marianne C. Stewart and Paul Whiteley “Taking the Bloom Off New Labour’s Rose: Voting in Britain’s First Post 9/11 General Election.” Journal of Elections, Public Opinion and Parties, 16 (2006): 3-36.

Lord Ashcroft Smell the Coffee (2005) and Minority Verdict (2010).