Do General Election campaigns make a difference anyway?

Does the intense campaign over the last few weeks of an election make much of a difference to the outcome?

The country is about to submit itself to five weeks of election frenzy as thousands of candidates and their supporters take to the streets, party leaders are driven and flown around the country and millions of pounds are spent on campaign advertisements, leaflets, and broadcasts. In 2010, during the short campaign (the period after the dissolution of parliament to the election) the Conservative Party and its candidates spent nearly £5 million and the Labour Party just over £3.5 million. Campaign expenditure totals are expected to be even higher this year.

Despite all this effort and expense, does the intense campaign over the last few weeks of an election make much of a difference to the outcome? The psephologist, Lewis Baston, has made what might seem like a counterfactual claim that the intense campaigning during the last few weeks might make little difference to the outcome. The outcome of most elections is often the result that the polls suggested before the campaign started. For example, in 2010, despite the excitement of the campaign and the impact of the leaders’ debates the Conservative Party ended up with the same seven point poll lead that they started the campaign with in March. Despite the mid-campaign Lib Dem/Clegg mania surge they ended the campaign in much the same position that they started, their vote was slightly up, but seats down. His analysis highlights similar patterns over nearly every recent election where millions have been spend and seemed to make little difference to the final result.

The one election where he concedes that the outcome at the end of the campaign was different to what the polls were suggesting at the start was 1992. However, that was mostly due to a systematic inaccuracy across all the main pollsters which under reported Conservative support rather than a significant change in public opinion over the campaign, although he does concede that there was a swing to the Conservatives at the end of the campaign.

With recent polls suggesting that up to half of people who say that they will vote at the General Election undecided, one thing that can be guaranteed is that political parties won’t be prepared to test Baston’s theory. Parties and individual candidates will be investing as much time and money as they possibly can hoping that each hour of campaigning or pound spent will help to make a difference to the result.

The results

The table below shows the results of opinion polls in March or April, so before the start of the ‘short campaign’ the intense period of campaigning after the prorogation of Parliament and the General Election day.

Apart from significant difference between what the polls estimated as the proportion of Labour supporters in the electorate prior to the 1992 election and the eventual result, all of the final results were within the plus or minus 3% margin of error, which suggests that the intense and expensive short campaign might have made no difference at all!