Is the Bob Worcester analysis on the nail?
Sir Robert Worcester, founder of MORI, posted this on the overnight thread.
â€œWhatâ€™s going on, the polls seem all over the place? One says a lead of 6 and another a lead of 17. â€ Iâ€™ve been hearing that for many a year. So what is going on? Part of the answer is â€˜events, dear boy, eventsâ€ And the Pre-Budget Review (PRB) was not just an event on Wednesday the 9th of December, it was a big event.
Another part of the answer is the mediaâ€™s fixation on headlining the lead, both in print and in the broadcast media, rather than the shares, especially the share for the opposition. Nine out of the ten reported so far this month have had the Conservatives at or over the magic 40% level, where theyâ€™ve been hovering since the beginning of July.
Yet three polls had leads were of 17% including this monthâ€™s Ipsos MORI Political Monitor, and three, ICM (9%) and YouGov (9%) and Populus (8%), below 10%. Yet none, not one of the ten, have varied in their share of Tory vote by more than three percent, pollsâ€™ statistical â€˜margin of errorâ€™.
Third, they all â€˜weightâ€™ their â€˜rawâ€™ data in various ways, the way economists have different models on which they base their estimates. The best way to see the impact of this is to compare the â€˜rawâ€™ data, and then youâ€™d know the basis that they use to adjust their figures to balance the inherent anomalies (explained in full at the Ipsos-MORI site. All the pollsters (except BPIX) have similar website detailed explanations, although some others are not fully transparent in their methodology, not willing to reveal their â€˜rawâ€™ findings.
A fourth part is that what the media really want is the polls to be able to forecast the outcome of an election six months or more, yet none of us have crystal balls, read tea leaves or palms, at least not to my knowledge.
If we sent our interviewers out to a carefully selected sample of constituencies across the country and got each one to stand in the highest point in the constituency and precisely noon read the carefully calibrated thermometers we armed them with, and struck the average temperature of the country in that way, and then to do the same the next day, no one would expect exactly the same average temperature. Yet public opinion is more variable than the weather tomorrow, much less than six months away.
From a six point Tory lead before the Pre-Budget Review to this monthâ€™s 17, a swing of 5.5% in one week, which could see the Tories in with a landslide.
Currently the Conservatives are on 43%, 26% say they would be voting for Labour if the election were held tomorrow, and 20% say theyâ€™d vote Liberal Democrats. Thatâ€™s five or six people in a hundred changing their mind about who they think theyâ€™ll vote for before and after the PRB. UKIP gets four% and BNP two%, the Greens 3%, and others 2%.
The impact of the PBR moved the economic optimism in the country from last monthâ€™s 46% believing the economy was going to improve in the next 12 months falling precipitously to 32%. At the same time, the satisfaction level of Gordon Brown fell another 10 points, to just 28% and the government also by 10 points, to 21%.
The impact of the PRB is felt by everyone. Most people, 69%, say that which party wins the election is important to them personally. If they werenâ€™t paying attention before, they are now.
Overall, the percentage of people who expect the economy to improve has fallen from a remarkable 46% expecting the economy to improve in the next year, to 32%. What is interesting politically is who has fallen the furthest. Last month 41% of the youngest third of the electorate, those between 18 and 34, said they expected an upturn; this month itâ€™s still 41%.
But of those 55 and over, thereâ€™s been a 20 point drop, from 50% optimistic, to 30%. This is a vitally important finding to all the parties, for those 55 and older have four times the voting power of the young. There are twice as many of them and they are twice as likely to vote. Other big drops came among men (-18) rather than women (-11), among the working class and those in the Midlands, and especially with those on a mortgage, who fell nearly by half, from 50% optimists to 27%.
The turnout is the key to the outcome of the next election. No election saw turnout under 70% from 1945 to 1997. Yet in 2001, only 59% of electors in Great Britain voted; in 2005 61%. In 1959 and again in February 1974 watershed elections, the turnout was over 78%. To get to that level, the reluctant Labour supporters whoâ€™ve sat on their hands in the last two elections would have to turn out by the thousands, especially in the marginal seats.
Even to get a turnout of 60% would require a massive increase in a now relatively apathetic base of Labour abstainers. For while a turnout of just 50% would return a Tory majority over all other parties of over 100, a 78% turnout would see a Labour majority of about 25. At about 60%, a 40% Tory, 30% Labour, 20% LibDems, 10% others, would give the Conservatives a majority of around 20 seats over all other parties.
Is fear of the Cameron/Osborn team enough to energise the Labour reluctants?
Whatever, last monthâ€™s poll findings certainly put the frighteners into the Tories, hope into Labourâ€™s minds, and heart into the fond hope of the Liberal Democrats that there was a window of opportunity to hold the balance of power in a hung parliament. Last month the margin was closing, Post PBR it looks like itâ€™s Cameron for No. 10 without a massive and effective effort from the Labour campaign team. But a week, never mind a month, is a long time in politics.
Thanks Bob – it’s always great to welcome you to PB and the detail you provide is fascinating. The point you always make about focussing on the shares rather than the lead is one which we should all heed. It’s so easy for a headline to put the emphasis on that which can be misleading.
Looking back over the past fortnight the polling oddities seem to have been the last Sunday YouGov poll in the Sunday Times and Tuesday’s ICM in the Guardian. Both showed the Labour share increasing and it looked as though the PBR had had little impact.