ComRes Unveils Voter Turnout Model

As the British Polling Council commences its inquiry into how UK pollsters underestimated or completely missed the Conservative lead over Labour during the run-up to last month's general election, ComRes says it has developed a solution to predict who is more and less likely to vote.

Of the final eleven polls conducted before the election, ten suggested that Labour and the Tories were neck-and-neck, which would have resulted in a hung parliament. By contrast, the BBC exit poll predicted that the Tories would get 316 seats and Labour 239, and proved to be surprisingly accurate, with the final outcome delivering 331 seats to the Tories and 232 to the Labour Party - 36.9% and 30.5% of the total vote respectively. The inquiry, announced the day after the election, is currently looking into the possible causes of 'apparent bias', and will make recommendations for future polling.

ComRes, along with a number of other British pollsters, has traditionally asked respondents to rate their own likelihood to vote using a 1-10 scale, but says that while this has proved a useful indicator of likely turnout, some survey respondents overstate their likelihood to vote. In particular, based on constituency-level turnout data as well as ward-level turnout data from local elections held at the same time, ComRes' modelling suggests that less affluent voters are more likely to exaggerate their likelihood to get to the polling station. The firm's new Voter Turnout Model simulates turnout among different demographics, and provides 'more consistent' patterns from election to election than the demographics of party support.

Having looked back over its campaign polls to measure the impact of this model, ComRes found that had it used the new method during the recent general election, its final poll would have had a 5 point Conservative lead over Labour. The company has also applied the model to its first post-election voting intention poll, and says the new approach now produces a 12 point Conservative lead, as opposed to an 8 point lead if the old approach is used.

Head of Political Polling Tom Mludzinski (pictured) comments: 'As with all approaches, there is no one magic fix that is perfectly without flaws. However, we do believe this to be a significant improvement on the accuracy of identifying the voting public and separating them out from the wider, general public. This model of course also has wider applications for all elections, and we will therefore be applying it to other upcoming elections including the EU referendum as we continue to review, innovate and test approaches'.

Web site: www.comres.com .