Labour’s Vanishing Lead
Here are Leo Barasi’s new calculations. His rule for the polls a year before an election is: halve the lead (of whichever party) and then move the lead 3½ points in the governing party’s favour.
That means, if the average Labour lead is still four points in a month’s time, his model would give the Conservatives a 1½-point lead in the election on 7 May 2015. (Halve Labour’s lead from 4 to 2 points, and subtract 3½ points, giving a Labour lead of -1½ points.) That would mean that the Tories would win the most votes and Labour the most seats.*
Stephen Fisher of Trinity College, Oxford, meanwhile, calculates that in the average pre-election period since 1950 there has been a swing of 4½ points back to the governing party. (Barasi summarises the main differences in method in his post.) That would turn the present Labour lead into a Tory lead of 5 points – enough for the Tories to remain the largest party, but not enough for a majority.
The third estimate of average historical trend will be coming shortly from the psephologists behind the Polling Observatory, as trailed at the end of this post.
Naturally, the past does not determine the future. Elections and the year before them can be very different, and this one is complicated by having two parties in government. But, as you can see from Barasi’s upper-right quadrant in the diagram, there have been no cases of oppositions with a small lead a year before winning the subsequent election.
*This would be an odd situation, although it happened in 1951 (Labour most votes, Tories most seats) and February 1974 (other way round). But it is seats that matter in forming a government, and once it is formed the votes hardly matter, just as they ceased to matter as soon as the Supreme Court ruled for George Bush in 2000.Tagged in: 2015 election, leo barasi, opinion polls, stephen fisher
Latest from Independent journalists on Twitter