Issues Don't Count (As Much As They Should)
Politicians and pundits alike decry that 'this election should be about the issues'. Yet personalities intrude. In the four-sided tetrahedron that drives my model of voting behaviour, 'values' count for most, as roughly 80% of the public have their mind made up before any election is called, and most of them don't change the habits of a lifetime, in the aggregate.
That is not to say that individuals don't switch their allegiance from one party to another over the course of a number of elections, it is to say that the 'core' vote for the Tories is about 30%, tested nearly to destruction in 1997, and about to be tested again; the 'core' vote for Labour is about 30%, tested beyond destruction in 1983, and the 'core' vote for the collective others, Lib Dems, Nats, etc., is about 20% or a bit more, as proven in the European Parliamentary elections in 1989 when the Greens soared to 15% when the Liberals and the Social democrats were falling out of bed. That leaves 20% who determine not only the outcome, but also the magnitude of every general election.
Further, it is only the 20% of the electorate in the 20% of the constituencies whose votes matter, so 4% really determine the result of this, and every other on the nine elections since I've been here closely monitoring British General Elections.
Issues don't count for as much as they should for four very good reasons:
- Salience: if people don't care about an issue, one party or another's argument is not going to sway that voter to move their allegiance.
- Discernment: if people don't discern differences between the parties on the issues they care about, then they are not going to change their allegiance either to support the party which they think has a sound policy on the issue they care about, or reject a party which they don't think has the best policy on that issue.
- Ability: if people don't think the party that has the best policy on the issue they care about has the ability (power) to do something about it, they won't be moved, and
- Will: if people don't think the party that has the best policy on the issue they care about has the will to something about it, they won't be moved.
There are precious few issues that can jump those four hurdles. About 40% of the determinates of the 'floating voters' are issues, and about 60% image, split more or less equally between party image ('good team of leaders', 'keeps its promises', 'represents all classes', etc.) and leader image ('capable', 'understands the problems facing Britain'), and the negatives, e.g., 'talks down to people'. These are derived in focus groups, and tested for saliency in quantitative surveys and replicated over many years using multivariate analysis techniques to track changes over time.
Latest Polls
Rasmussen Research, an American polling company, has entered the fray, but you won't know that they are polling for the Independent from the BBC, because they are not on the list of 'approved' pollsters, broadcast on the 'approved' programmes that have been licensed by the political advisor as suitable outlets for the scant coverage they give to the standing of the parties in this election, a holdover from the awful days of John Birt's dictatorial style of news coverage. Never mind news values, this is the BBC's Guideline.
Anyway, they are, and their record in the American elections was nothing to write abroad about. In the final event, they were 15th out of the 16 polling firms in terms of forecasting the outcome, with an average error of 2.4%.
While the outcome, famously, was c. 48% Bush and 48% Gore, Rasmussen called it 52% for Bush and 43% for Gore, so was over by 4% on the President, and under by 5% on the Vice President. Further, the local talent is anxiously awaiting some explanation of what a 'telephone computer' methodology can be. If it is what we think it is, it's a failed marketing research technique that's been tried and found wanting in this country.
For what it's worth, they too have the Conservatives at 32%, Labour a bit lower than most pollsters here, at 46%, and the Lib Dems at 13%. So, they must have the 'others' at 9%, a highly unlikely scenario.
ICM in the Guardian this morning reports 31% for their client the Conservative Party, which can't make for very happy meetings at Central Office, 46% for Labour, following Rasmussen, and the Liberal Democrats at 16%, which is up from the 14% average for the Lib Dems in most polls.
We'll see.
