14 May 2015
Polls Taxing
by J.R.Thomas
As the spent ballot papers are fed through the shredders of town halls up and down the land, the reproaches begin. Some of the loudest come from the opinion pollsters – reproaching themselves. The practitioners of this mysterious and endlessly reweighted and rebalanced science have fallen into what seems to be a cacophony of guilt, accompanied by a rending of survey sheets. This acceptance of blame is eagerly encouraged by newspaper editors and politicians – especially newly out of office and out of seat politicians.
Yet, examination of the results suggest that the forecasts were not that far off, especially in relation to votes cast. The forecast for the main parties was consistent throughout the campaign at around 34% for the Conservatives, 33% for Labour, 13% for UKIP, and 8% for the Liberals. On the day, the forecasts for UKIP and the Liberals were spot on, but the Conservatives took 37% of the vote, and Labour 32%. And the SNP did precisely as the polls said they would.
It is well known (and the pollsters allow for it) that Conservative voters are less likely to declare their true intentions and that Labour voters have a lower propensity to turn out on the day. Because the surveys are weighted for this (and do not, as suggested in some press comment this week, just survey the unemployed in shopping malls, thus omitting Tories busy in their workplaces) it seems unlikely that these were serious factors. It does seem that Conservative supporters may have turned out in higher numbers than usual – maybe inadvertently encouraged by Ms Sturgeon and fear of bagpipes in Whitehall. The one set of polls that did pretty much pick up what was going on was Lord Ashcroft’s – which concentrated on marginal seats.
It seems that there was a fair degree of voter sophistication in those marginal seats on the day. Here several themes coalesced; in Tory/Labour marginals there was a determination to keep Labour out (so St Albans, a traditional southern marginal which normally has a strong Liberal vote, saw both the Labour and Liberal votes collapse and a sudden transformation into what now looks like a Tory safe seat). Then there were the Liberal seats in the West Country, where the Labour voters traditionally vote Liberal to keep the Conservatives out – but this time one can surmise that disillusioned Liberals switched allegiance or stayed at home, as did their former Labour supporters. And in some of the Midland and northern seats UKIP were picking up enough Labour votes to let the Conservatives in.
Most people who live in marginal seats know that they do so, and know that their votes are the ones that count. They will vote against those they want to keep out, almost more than for those they might want in. This behaviour was widespread enough to affect the results in those key seats, but without doing much to the national aggregated figures. And the effect was sealed on election day by an especially strong turn out by Tory voters.
So have some pity for those masters of the balanced sample and the weighted questionnaire. At least they will know where to concentrate their clipboards next time. And Mr Cameron should reflect that the apparent enthusiasm in the country may actually be disappointment with his opponents, and that he has not won the battle for hearts and minds quite as resolutely as he may think.
