The Stirrer

news that matters, campaigns that count

for Birmingham, the Black Country and beyond

Mick Temple’s Blog



Isn’t it great when opinion pollsters get it wrong? Mick Temple reckons we should take a lead from the Italians and simply lie to those who want to know who we’re going to vote for.

Two groups of people had egg on their faces this week – those American opinion pollsters who unanimously gave Obama a New Hampshire win, and the British papers whose early editions brought back memories (to those of us who are very old) of Dewey Defeats Truman from 1948.

For me, the pollsters’ shame brought back more personal memories from 1992. Picture the scene. I’ve just finished my award-winning PhD on coalitions. I’m watching TV on the night of the general election, flicking between BBC and ITV. Both channels’ exit polls predict a hung Parliament. I know they must be right as my very own PhD supervisor, Professor Colin Rallings, is responsible for ITV’s polling.

Despite my gammy leg, I dance around the room! My teaching contract is coming to an end and I am looking for a new academic post. The country is heading for coalition government and I am the nation’s leading expert - Oxford University here I come …

Two hours later and my horizons have shifted – I’ll be lucky to get a job at one of those American postal universities which peddle ten-dollar degrees. Thank you Neil Kinnock for opening your big Welsh gob at the Sheffield rally.

Now – the real point of this blog. Opinion polls now dominate all election coverage and the ‘horse-race’ – who’s winning, who’s coming up on the rails, who’s been left at the start - has become the news, rather than important issues of policy.

Polls are presented as ‘scientific’ – for example, a representative sample of 1000 people is accurate to within 3 per cent, 95 per cent of the time, the pollsters tell us. Up to a point, Lord Copper. What this overlooks is the ‘tweaking’ that has to take place to get them anywhere near approaching ‘reality’.

Over the years ‘allowances’ have been made for the reluctance of Conservative voters to reveal ‘private’ information or even take part; the unwillingness by supporters of unpopular positions (e.g. BNP voters) to admit this to a stranger; and telephone pollsters’ failure to question many professionals (who tend to be ex-directory) and those without phones.

In the past, Labour votes have also been ‘over-estimated’. Internet pollsters tie themselves in knots trying to balance the books.

And in the last few days of an election there is a well-observed trend of undecided voters switching to the party ‘in front’, because people like to back a winner. The polls also have to be tweaked to take account of this.

I have a solution to this problem. In Italy, lying to opinion pollsters is a national sport. Let’s introduce the sport over here. Next time you’re asked what you think of Nick Clegg, lie. Say he would make a fabulous prime minister.

If we all fibbed in this way, opinion polls would rapidly become so discredited that pre-election coverage would have to concentrate on discussing the issues. There would be no sense among the electorate that a particular party was not worth getting out of bed to vote for because it was either obviously going to win or had no chance.

Turn-out would increase – and more importantly, we’d all be taking part in the democratic process free from the instinct to follow the herd.

And it would put an end to those witless wonders who moan about never having been asked which party they support …

Have you ever been asked your opinion by a pollster? Did you tell them the truth? Leave a comment on the Message Board.

The Stirrer Forum

The Stirrer home

©2007 The Stirrer